2024-03-01 00:00:00 CAFIAC FIX

2024-03-11 00:00:00 CAFIAC FIX

2023-05-22 17:12:00 Once again, the American Anthropological Association will be voting on a (twice rejected) “Resolution to Boycott Israeli Academic Institutions”. I must share with you a website launched to clarify why it should be voted down decisively: https://t.co/YfAGoWAP1p https://t.co/ybPku7XZnI

2023-05-21 20:35:07 @AlfredoMorabia @DrRichBesser @LauraMagVall @JeromeAdamsMD @mfraserdc1 @AMJPublicHealth @AJPHThinkTank @PublicHealth @ASTHO @NACCHOalerts @CDCgov @ASPPHtweets @RWJF @JHUPress @JohnsHopkinsSPH @ColumbiaMSPH @QC_News Multi-congratulations, Alfredo. How about a link to the index?

2023-05-21 20:19:35 @kenkoonwong @anirbanb_007 @piyushm76777614 @mkashiouris Neural network is a model of the data, not of the world that generates the data. Its role in causal inference is to manage the estimation part of the process, going from finite samples to distributions

2023-05-21 16:41:07 @kenkoonwong @anirbanb_007 @pmorenofranco @MITCriticalData @judywawira @freemanwd @jhalamka @AbhiDuggalMD @shapatel13 @Dev_Sanghavi @piyushm76777614 @mkashiouris @VHerasevich I can't answer this question, since I haven't attended the meeting. But I agree with you that any talk about "bias" that does not leverage the transparency and testability offered by DAGs is way behind the times. And I say it from a totally unbiased perspective.

2023-05-21 16:25:55 'BY ANY MEANS': Israel-haters march in Toronto, call to annihilate Jewish state, globalize terror | World Israel News https://t.co/OyAF2eILS0

2023-05-21 01:04:47 @aynumazi @jacobin @TJAllon1 @ShMMor @blakeflayton What part of the model you have doubts about?

2023-05-21 01:02:07 @aynumazi @jacobin @TJAllon1 @ShMMor @blakeflayton I did not say "Oppressing" I said "harbor genocidal ambitions." Are you doubting the latter?

2023-05-20 23:40:17 @terrible_archer @peteratmsr @eliasbareinboim The points you are raising are crucial for understanding what ChatGPT can do for us in the near future. I am still trying to formalize how much causal knowledge it can safely fake

2023-05-20 23:33:39 @aynumazi @jacobin @TJAllon1 @ShMMor @blakeflayton I did not say "oppressing". I said "planning a genocide". They invariably admit (to the last man)to "planning the end of Israel." There remains only to prove that "end of Israel" amounts to a genocide. So, ask Israelis if they would relinquish freedom w/o bunker to bunker fight.

2023-05-20 23:20:03 @erikbiz "Hatred" is a slippery concept. You can't get a Zionophobe to admit that he/she hates Israelis, for it's not socially acceptable to hate people. But you can easily get a Zionophobe to say that he/she is planning the end of Israel, a genocide that IS socially acceptable.

2023-05-20 20:54:32 Come to think about it, of all racist groups currently roaming our social media, Zionophobes are the only ones who actually harbor genocidal ambitions. https://t.co/jcCnpOUelL

2023-05-20 20:35:03 To all Portuguese speaking readers, including Brazilians, appreciating your support of the Israeli miracle, its moral imperatives and its contribution to world's prosperity. https://t.co/69G8iTAHSk

2023-05-20 20:23:18 @jacobin When Zionophobes use the trauma of the Nakba to defend Palestinian eliminationism, they are betraying the spirit of “never again” that was supposed to ensure the world would never stand by as genocidal ambitions were publicly discussed &

2023-05-20 13:50:34 @DavidHirsh @_ZachFoster and other "Historians of Palestine" have not heard about Zionophobia

2023-05-20 12:50:56 Speaking about "evidence" for causal hypotheses, in this episode of the @ezraklein podcast: https://t.co/fiL0XSfa4w we hear about statistical evidence (Rung-1), but not about the causal assumptions (Rung-2) that turn the data into a definitive evidence. Can it be repaired?

2023-05-20 06:50:10 @soboleffspaces Mediation is good example, and so is attribution, e.g., finding "causes of effects," or finding probabilities of harm and benefit . Non-compliance, however, is a Rung-2 problem.

2023-05-20 06:28:43 So why am I so excited to hear that the oldest manuscript of the Hebrew Bible was auctioned yesterday (for $38,1M) and will soon be displayed in Tel Aviv Museum of the Jewish people: https://t.co/kYTpmDEOQ5 See https://t.co/g09NNetuGQ for a closer look. The reason is probably 1/2

2023-05-20 04:57:22 Many readers have expressed ambiguity regarding the distinction between Rung-2 and Rung-3 of the Ladder. I'm retweeting my answer with the hope clarifying this barrier. https://t.co/S5NIGwwOdn

2023-05-19 23:10:42 Sharing @bariweiss conversation with Daniel Gordis https://t.co/2RgIlYedOw, which I've found to be just the right thoughts with which to celebrate Israel's 75th anniversary.

2023-05-19 22:53:59 @jmugan If we have a causal model then, you are right, we can get an an answer to every counterfactual question. Now suppose you have only experimental studies, no model. Some counterfactual questions could still be answered, some could not. The former are Rung-2, the latter are Rung-3.

2023-05-19 19:49:03 Tlaib’s shameful weaponization of the Elie Wiesel Genocide Act against Israel https://t.co/mAQISUeOGF

2023-05-19 19:21:43 Made my day. Now I fee that: "We're back to this ancient city" (Chazarnu El Borot Hamayin ... Ba'Eer HaAtica" https://t.co/DLffqbCeKw

2023-05-19 19:00:00 CAFIAC FIX

2023-05-21 19:00:00 CAFIAC FIX

2023-04-21 00:00:01 CAFIAC FIX

2023-04-14 23:01:25 Mighty congratulations to Dr. Ang Li, who has accepted a faculty position at FSU (Florida State University, Tallahassee, FL). We welcome FSU as a new center of CI research, specializing in theory and applications of personalized decision making. See https://t.co/MlyGdEjQHD

2023-04-14 22:42:19 Grateful to @smueller for clarifying the relationships between the PoC bounds obtained from multiple experiments vis-a-vis those obtained from observational studies with a known graph structure. Both will revolutionize personalized decision making: https://t.co/rGHIc93DYO https://t.co/yDlVrFvsAD

2023-04-14 16:25:09 @LuisaBorrell I think the emphasis here is on the words "rightfully confused." It's been 30 years now that I am trying to understand what g-formula-g-computation is

2023-04-14 04:28:07 For North African Jews, it’s a night when we turn over our homes from Passover to Mimouna, a joyous family and community-oriented cultural celebration that’s all about blessings, smiles and sweets. https://t.co/0GGQ7o305D

2023-04-13 22:25:56 Important announcement from the Journal of Causal Inference (JCI): We are looking for an energetic new talent to fill the position of Editor in Chief: https://t.co/08Im4m0iI7  If you are qualified and enthusiastic for undertaking this exciting job, please apply before April 30.

2023-04-13 10:13:20 @fabianbgng's paper describes clearly why causal mechanisms need be taken into account in defining fairness criteria, let alone estimating degrees of fairness. A word of caution, however, on the use of "matching", which is not a causal notion. #Bookofwhy shows how misleading 1/2 https://t.co/M7wtfj8W3R

2023-04-13 09:20:25 @DavidSKrueger @lastpositivist @fabianbgng @csilviavr Causality does not "solve" fairness, but it is necessary for solving problems associate with fairness, simply because "fairness" is a causal notion

2023-04-13 07:03:25 It's still not the root cause. https://t.co/DCukhp312B

2023-04-13 03:21:09 @suzannTee @erichorvitz We had a few intimate sessions.

2023-04-13 03:04:49 I've seen people in deep prayer attacked by less moral armies. I'm sure you have seen some too. I leave it to your judgment. https://t.co/P2BhV5oz9R

2023-04-13 02:51:33 You can't fight antisemitism without addressing Zionophobia, squarely and explicitly. Priorities have shifted! https://t.co/DngmEhVzzX

2023-04-13 02:47:45 RT @vardi: :-) https://t.co/wygSuKWn0y

2023-04-13 02:37:02 Whether or not IDF is "the most moral army in the world" is told by the faces of these worshipers, none of whom shows any fear whatsoever when asked and forced to move back a few steps. They know they are perfectly safe, in the hands of this moral yet dehumanized army. https://t.co/sAsulBGVVM

2023-04-13 02:13:20 Going by the title alone, "Council on Academic Freedom" may be either an extreme right-wing organization or a radical left-wing movement. However, the authors: Steven Pinker and Bertha Madras render it safe for ordinary Sapiens. https://t.co/Kg4bWirQRm

2023-04-13 01:31:53 Do two men walk together Unless they have planned? (Amos, 3.3) https://t.co/lGR9zo8Lr3

2023-04-12 23:05:03 @noahdgoodman @akatzzzzz @johnjnay @Stanford I think when you work out the theory the difference between probabilistic and causal dependencies will surface automatically. Recall, conditional independencies are not axiomatizable, whereas counterfactuals ARE.

2023-04-12 20:48:18 You made my day! Now I only wish that one of the authors of LLM would read #Bookofwhy. I just received a proof of the Hebrew translation, in case they need an truly approachable version. https://t.co/EnjPNs8jii

2023-04-12 20:42:02 @noahdgoodman @akatzzzzz @johnjnay @Stanford The question you ask in this work goes beyond the causal/probability dichotomy. It deals with going from local judgments to global inference. In what sense do conditional probabilities chain? I think they don't. Can you elaborate?

2023-04-12 20:36:01 @ThomSeaton @EinatWilf @Adi_Schwartz There are 120 professors on my "strong Zionist" list, who are not too happy about the "Bundist Occupation" of Jewish institutions at UCLA.

2023-04-12 20:30:31 @dggoldst @erichorvitz I bet you never forgave Ward Edwards for that. He was a straight shooter, and strongly opinionated. It was through him that I met the Tversky-Kahneman team.

2023-04-12 20:24:17 @erichorvitz Did you figure out why gpt4 ordained you and Jack as the authors of the QMR paper? Do you understand why we are spending precious time asking why GPT4 did this or that?

2023-04-12 19:58:38 @erichorvitz The guy is changing his mind. Look what he tells me: The term "Bayesian network" was first published in the paper "Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference" by Judea Pearl in 1988. Doubly wrong: (1) It's a book, (2) 3 yrs too late.

2023-04-12 12:12:29 @peteratmsr How do you ask "via Bing"?

2023-04-12 10:32:23 @EKochba It takes counterfactuals to stay strong and young at any age, but when you have only 34 yrs left, it comes natural.

2023-04-12 08:11:06 @eleroymason Hard core Bayesian statisticians claim there is nothing Bayesian about Bayesian Networks. Why, because they do not assign priors to parameters. Go argue with statisticians.

2023-04-12 08:02:47 @BillBainbridge5 I don't know much about Control Loops, but why shouldn't LLM access Google Scholar, validate its sources, double check its answers, and save me many routine key strokes.

2023-04-12 07:51:26 Our slightly belated List of new papers in Causal Inference https://t.co/Pf2UtbXbU8. Having spent many sessions on personalized decisions, our interest would naturally be drawn to the paper on Bounding Probabilities of Causation https://t.co/rfzIpFseRZ, which I am still 1/2

2023-04-12 06:49:01 @taravanand @AJAveritt @eliasbareinboim Intriguing !! What is the simplest C-DAG you can draw to demonstrate the problem, its solution and, most importantly, the input information required of the user.??

2023-04-12 06:41:49 @entropypromoter I'm not trying to use it as a replacement to Google scholar. I am trying to uncover its chain of associations, and what type of text makes the chain stronger or weaker.

2023-04-12 06:34:55 I remember my parents reading the newspaper in the kitchen, trying to hide it from me, 11 yrs old, inquiring "who did it?" https://t.co/hXcj8KBo0Y

2023-04-12 06:28:30 One of my most beloved authors. Sad to know he will not surprise us again with a new color of beautiful Israel. https://t.co/QPe6UQScnA

2023-04-12 06:23:09 Have you ever tried the word "Zionophobe" to neutralize racists of this breed? It has worked miracles for me. https://t.co/H0bQPcs6pE

2023-04-12 06:13:54 RT @TheLeadCNN: Anti-Netanyahu protesters flood a Tel Aviv highway as opposition grows over Israel's judicial reforms, an uptick in violenc…

2023-04-12 06:09:44 RT @TheLeadCNN: “Not on my watch in Massachusetts.” Democratic Governor @maura_healey takes executive action to protect women’s access to a…

2023-04-12 06:05:04 @AvivaKlompas @FranceskAlbs You are asking @FranceskAlbs to answer a moral question, as if she is capable of understanding such questions, as if she herself has not emboldened those killings by bestowing automatic impunity to the killers.

2023-04-12 05:49:05 RT @DannyAyalon: 20 years ago today, I participated as a member of the Israeli delegation to the Camp David Summit. Holding the foreign pol…

2023-04-12 05:47:49 RT @DavidLBernstein: The goal should be helping people achieve the American Dream not declaring it racist

2023-04-12 05:25:50 published in 1988. The paper discussed the use of Bayesian networks as a representation for capturing the probabilistic relationships between variables in the context of medical diagnosis." This answer has 3 errors: (1) The INTERNIST paper was NOT authored by Breese and Horv. 2/3

2023-04-12 05:25:49 My question was: "When was the term "Bayesian network" first published? DPT-4 answered: The term "Bayesian network" was first introduced in the paper "Probabilistic diagnosis using a reformulation of the INTERNIST-1/QMR knowledge base" by John S. Breese and Eric J. Horvitz, 1/3 https://t.co/oRjtdxKC5D

2023-04-12 04:58:52 Lyrics: @FranceskAlbs Music: @KenRoth Choreography: Supreme Ayatollah https://t.co/cJSXledyxP

2023-04-12 04:29:30 @neuro_data networks to medical diagnosis, might have been more frequently mentioned or discussed in the training data, leading me to associate it with the first appearance of the term." And btw, the 1988 paper mentioned was NOT authored by Breese and Horvitz. 2/2

2023-04-12 04:24:42 @neuro_data The answer was generic: While I cannot pinpoint the exact related works that led to the incorrect association, it is possible that the prominence of the 1988 paper by Breese and Horvitz in the context of Bayesian networks played a role. This paper, which applied Bayesian 1/2

2023-04-12 00:53:35 @AJAveritt @taravanand Any link to @taravanand work?

2023-04-11 23:24:41 @Philosofrenzy @kevinmcld Of course it can. You and I are perfect proof that it can.

2023-04-11 23:03:48 @Philosofrenzy @kevinmcld I took it tongue in cheek, knowing that it meant something else. I am still exploring what it really means, so I can properly interpret what next to expect of it once it says "thank you". Any idea?

2023-04-11 22:47:35 @kevinmcld @Philosofrenzy Pearl once wrote that "thank you" is the most common counterfactual expression in the English language.

2023-04-11 13:42:25 @mcxfrank I was set out to test if GPT-4 can pass our definition of "understanding" and ended up surprised that it doesn't double check its answers, even when it is available in its data base. https://t.co/xY1c7qKkaS

2023-04-11 13:30:25 They claim millions of things but their key, conflict prolonging claim is: Tel Aviv is a stolen land. https://t.co/XR4sXWo94f

2023-04-11 13:22:47 RT @Ostrov_A: At last, we hear from British PM Sunak: "The killing of British-Israeli citizens Maia, Rina and Lucy Dee is abhorrent. The UK…

2023-04-11 13:15:26 Many pundits enjoy drawing analogies between North Ireland and the Israeli-Palestinian conflict. Most fail to note one essential difference: The IRA never claimed London was on stolen land -- the Palestinians do.

2023-04-11 13:15:25 Opinion | What can Israelis and Palestinians learn from Northern Ireland’s 25 years of peace? https://t.co/GqwPwF0Cp2

2023-04-11 12:49:12 RT @mcxfrank: What does it mean for a large language model (LLM) to "have" a particular ability? Developmental psychologists argue about th…

2023-04-11 10:04:25 transitivity, but failed to give an example of such violation. It thanked me profusely for producing such example (two-coins and a bell), and we both agreed that the reason intransitivity seems unassailable in this example is that it invokes causal, not statistical relations. 4/4

2023-04-11 10:04:24 I had a few intimate sessions with GPT-4, which I intend to summarize systematically on my blog. In a nut shell, after GPT failed to find where Bayesian Networks where first published it confessed: I gave you a paper that does not exist, and doesnt mention "Bayesian Networks" 1/4

2023-04-11 07:59:38 @ThomSeaton @EinatWilf @Adi_Schwartz I've read "War of Return" and I advice all readers to read every word of it, if they wish to understand why the conflict has been going on for 100 years with no resolution in sight.

2023-04-11 07:55:22 @JohanDH2O @noah_greifer @RWJE_BA It may seem like another case of tomahto/tomeito until you try to articulate what the condition is for the validity the g-formula. Try it without the back-door and see how far you can go.

2023-04-11 03:58:34 Retracking, sorry. The murdering ring is being chased in Nablus, not Jericho. Is there a person in the civilized world who is not praying for their quick capture? https://t.co/InAM8YXfrx

2023-04-11 01:30:48 Just watched PM Netanyahu on Israeli TV: (1) Yoav Gallant is reinstated as defense minister. (2) All Israel's problems, terror, discontent, etc. are blamed on the opposition and the previous government. Barring the latter, he could have reclaimed his Mr. Smart legacy.

2023-04-11 01:13:36 Whenever a University hires an Ex-CAIR executive, it is doomed to lose some of its most creative faculty to identity politics. CAIR is a hotbed of hatred and intolerance. https://t.co/SHL0K1GGbR

2023-04-10 20:06:49 Readers have asked many times: How can you be sure that, deep inside, Palestinians are not disposed to some sort of peace with Israel? This new post https://t.co/8cIlv8p0be summarizes the key evidence: Denial of any Jewish connection to the land. Nothing could be more telling 1/2

2023-04-10 19:25:34 I welcome introductory courses in causal inference -- the world needs it badly. I havn't read the content but, seeing the abstract, I can tell that it is not one of those courses titled "causal inference" that ignore the problems of transparency, identifiability and testability. https://t.co/2BYcqQ6iae

2023-04-10 19:12:53 Request for readers who are proposing expansions to the ladder: Please provide a typical question answerable at layer i that is not answerable by layers j <

2023-04-10 10:06:03 @soboleffspaces Nice graph. I remember how Imbens tried once to convinced me that he can select the right covariates by thinking ignorability -- the envy of us, mortals.

2023-04-10 09:57:42 What move diplomats to make statements and say nothing. My gardener can do it at 1/10 the salary. https://t.co/YIpIz5lT4k

2023-04-10 09:53:33 Algerian Radio is adapting a Palestinian mentality. https://t.co/ON3DzDR2uc

2023-04-10 09:46:35 I assume they are searching for the murderer of the Dee sisters. ?. https://t.co/On3PwrXmTx

2023-04-10 09:33:05 The setup in this paper is the same as in @soboleffspaces work on hip-fractures, https://t.co/UWo5Anu64d, where the proper ignorability condition was derived from the model, rather than assumed a priori (bad habit of PO thinking). @smueller, @EliBenMichael, https://t.co/TWRRwPqLZX

2023-04-10 09:30:33 @haaretzcom So why does @haaretz display the idiotic poster of the first group?

2023-04-10 08:41:34 I join @lauder_ronald in applauding the decision by Columbia University to build an academic center in Tel Aviv, thus assuring students and faculty that BDS mentality and Zionophobic racism have no place on US campuses. President Bollinger should make it explicit. https://t.co/8jUgZkv1k5

2023-04-10 08:15:55 @rogerra43393439 Professor Fuchs is a long-time pollster. I heard him on Israeli TV explaining how they overcame biases due to Holiday vacation etc. Evidently they group voters by the party they last voted got, then reweigh the results knowing the group size from the last election.

2023-04-10 05:47:01 The reason is that "bound-narrowing" occurs only when observational data are confounded. It is only then that additional information about individual behavior can be extracted beyond RCT data. We show it in https://t.co/rGHIc93DYO https://t.co/iKkxvu37nX

2023-04-10 05:09:53 This paper assume strong conditional ignorability in an observational study which, essentially, gives us both experimental and unconfounded observational data. The bounds on P(benefit) and P(harm) are same as in Causality Eq. (9.9), and do not improve by including the latter. https://t.co/TRk9Pr3Gds

2023-04-10 03:18:45 Significant upset in Israel polls: Netanyahu Crashing, Gantz Far Ahead, Lapid Second. Just 20 percent of respondents in the Channel 13 poll say that Netanyahu is doing well as a prime minister, compared to the 71 percent who say that he is doing badly. Opposition gets 64 mandates

2023-04-10 03:05:35 RT @smueller: Great paper by @EliBenMichael, Imai, and Jiang on learning optimal policies when the gain from beneficial treatment ≠ cost of…

2023-04-10 02:42:39 I love it! If we are destined to be controlled by someone, its better be Jews

2023-04-09 20:50:26 "Ani Maamin" ad sung at the funeral https://t.co/th15vRzkDr https://t.co/K0eZoFWdtB

2023-04-09 20:45:24 Today I sing "Ani Maamin" (I believe), in the funeral of Rina and Maya Dee (15 and 20), who were murdered on their way home with their mother. (The mother is still fighting for her life.) It means "good will prevail over evil" -- as my grandparents sang in Auschwitz -- it will! https://t.co/DppnNvvqIW

2023-04-09 19:53:05 @AviBachsh Are you serious? And I thought she was changing her mind due to my unique power of persuasion. I feel cheated.

2023-04-09 19:43:49 @DKedmey Oh, NO! Robert Malley again? How many times should @USEnvoyIran be proven wrong, before being replaced by someone who understands the Middle East? @omriceren

2023-04-09 17:52:58 'Elimination of Israel' on the horizon, IRGC commander says | World Israel News https://t.co/dawZdT6Juq

2023-04-09 16:45:10 A truly illuminating time-line of AI history, from Turing (1950) to GPT-4 (2023). Grateful to @mishadavinci for posting this gem of history and for squeezing a seat for me on Mt. Olympus. https://t.co/PanmmvHGvo

2023-04-09 08:56:37 No! I am telling myself. You can't let this picture of 6 brothers and sisters, victims of the latest attacks on Israel remain on your retina without sharing it with others who perhaps haven't seen their smiles, their hopes, their humanity, and their mourning families. https://t.co/RGibIDug78

2023-04-09 08:42:58 @kenroth has a point. The purpose of the IHRA definition of antisemitism is, indeed, to distinguish Zionophobic racism from legitimate criticism of Israeli policies, to expose the genocidal agenda of the former, and to limit the outreach of its top ideologues, @kenroth being one. https://t.co/aU6jQUnzmr

2023-04-09 08:01:50 Sharing an informal interview I did last week with Adel Daoud's, to Social Science audience. https://t.co/XmbRTHgt4o I believe it contains some new anecdotes, absent from my formal lectures, which you may find relevant to the history and logic of causal inference and AI. Thanks! https://t.co/5odUeP3aUb

2023-04-09 06:50:55 To all my Christian followers on this educational channel: Happy Easter! The day of Jesus resurrection! Well matching the Jewish Passover - which commemorates our people resurrection. https://t.co/X41LirXtoq

2023-04-09 03:05:56 Sick of seeing Al Jazeera cesspool overflowing social media, I am tuning to @IDF for real-time information on what's happening in Israel. It's amazing how refreshing dry facts can feel when fake News hide the day. https://t.co/fWxeRe6kjA

2023-04-09 02:26:24 RT @ReportingLabs: SRL alum @BrumerDelilah talks about how Student Reporting Labs helped her expand her journalism skills as she heads off…

2023-04-09 02:25:21 Britain 2023! https://t.co/5al8mhkDBF

2023-04-09 02:21:13 RT @SachaDratwa: Two pictures that explain the Israeli-Palestinian conflict: 1⃣ Israelis mourning the loss of an Italian tourist killed y…

2023-04-09 00:07:34 Hats off to @FleurHassanN for the courage to be interviewed by Al Jazeera on the latest attacks in Israel. Last they interviewed me (in Doha) they vowed never to do it again. They now quote the UN who quoted Al Jazeera on Israel's wrong doing. Evidence-Based Deceit -- EBD. https://t.co/OJUjmLxb49

2023-04-08 23:40:44 We can argue till dawn whether Israel has done its share for 2SS, but one thing is beyond dispute: Israel teaches its children that 2SS is an ideal long-term goal, Palestinians teach their children that 2SS is unacceptable. I have personally experienced the former. https://t.co/rXLdud6e0e

2023-04-08 23:00:16 @robshookphoto @hahussain The state that Israel want to be is no different from Canada or Swizerland, Multi-ethnic but still proud of its history and distinct from its neighbors.

2023-04-08 21:10:04 @_gamalan Sure. GPT-4 started quoting me (albeit sloppily) when trying to convince me that it knows about "confounding", but it didn't realize that Soldier-2 deconfounds the effect of Soldier-1 in the firing-squad. It needed help in getting it, then argued forcibly for it.

2023-04-08 20:47:33 @Burning12196087 @L4tkeMuncher @hagadol2011 @hahussain Sorry, but when I read what my Palestinian academic colleagues write about Israel's future, this is exactly what the Palestinians are planning to do, given the chance.

2023-04-08 20:32:48 @FranceskAlbs's logic is missing one tiny ingredient: The so called "people it oppresses" are also those who call for Israel's demise once liberated. @FranceskAlbs is yet to endorse a 2SS under: "Equally legitimate and equally indigenous". We are waiting. https://t.co/h64d9gH3f1

2023-04-08 20:16:57 @academic_la The guy is as dangerous as he is nut.

2023-04-08 20:07:00 @MandE_NEWS path blocking, colliders, etc. etc. Why did you switch to this language instead of remaining in the potential-outcome framework and convincing me that choosing W = S2 satisfies the independence condition required by the defining equation? " That's when GPT-4 confessed. 2/2

2023-04-08 20:04:22 @MandE_NEWS I've asked: "We started with a definition of non-confoundedness in the form of an equation about conditional independence of potential outcomes. Yet when I asked you to check non-confoundedness on a specific problem (the firing squad) you gave me arguments based on diagrams, 1/2

2023-04-08 19:49:16 Especially surprised by the ability of a Rung-1 extrapolator (GTP-4) to take text containing instructions on how to climb from Rung-1 to Rungs 2-3, and sometimes (not always) follow those instructions when given text with Rungs 2-3 information. https://t.co/gCnAdTZb0P

2023-04-08 18:39:01 @robshookphoto @hahussain Trouble is, Westerners expect to hear this from Hamas, but when told that this is also the agenda of EVERY Palestinian in the West Bank, they refuse to accept. "It can't be" they say, "surely, some just want freedom and independence." Asked to name ONE, and they can't.

2023-04-08 12:27:53 A confession I couldn't get from any humanoid trained in the potential outcome culture. Victory! 4/4

2023-04-08 12:24:35 framework and the graphical causal model framework are both formal approaches to understanding and estimating causal effects. My previous statement about one being "more formal and direct" was not accurate, and I apologize for the confusion." A confession I couldn't get from 3/4

2023-04-08 12:24:34 Today I scored my first victory trying to convert GPT-4 into a commonsense Reasoner. To my question: Isn't possible that the two frameworks [potential outcome and DAGs] are equally formal, and saying that one is "a more formal and direct way to assess confoundedness" is 1/4

2023-04-08 09:09:35 @hahussain A tiny correction. The sentence "Now get me a Palestinian commitment that under 2SS an Arab Palestine will live next to a Jewish Israel and I’ll get you 2SS executed tomorrow" should read: "Now get me ONE Palestinian leader, journalist, TV-anchor, educator or cleric ...[same]"

2023-04-08 08:48:51 I dream of a new Rashida. One who has the courage to tell Al-Jazeera: "Come on, guys, we all know that Israel has no interest in wreaking violence at Mosques - it makes her look very bad on TV. So why don't you come up with more credible stories for what's happening in Al-Aqsa?" https://t.co/Z9z16hL8st

2023-04-08 08:18:43 @ansarikashif7 @FranceskAlbs A new causal model for current wave of terror: It is all fabricated by PM Netanyahu. The notion of terror, especially against innocent civilians and Italian tourists is so foreign to Palestinian mind that it had to be fabricated by PM Netanyahu. Readers RT-ing it = believing it.

2023-04-07 22:14:36 @AsraNomani @YasMohammedxx @DrZuhdiJasser @CLARITyCoal @AHAFoundation @Raheelraza @SoooMD @StateDept @mtaibbi @ConceptualJames I would label Al-Jazeera "deceit-affiliated" network, you don't need a state to be committed to lies anti-Western propaganda.

2023-04-07 17:31:28 Guess who is heading this flock of blood- smelling vultures -- of course, it's @FranceskAlbs ! As soon as the number of rockets launched at Israel exceeds 20, she knows exactly who to blame: https://t.co/5oxt5xP39z

2023-04-07 09:48:51 For a split second I thought @PatrickKingsley of the @NYT is at it again, trying to convince us that Israel has something to gain from violence at the Al-Aksa Mosque. Luckily, this time the picture does not serve his "storming" agenda. https://t.co/SGzknASsxC

2023-04-07 09:23:21 RT @hahussain: Picture of rulers of #Lebanon. What investor would look at this and not think: I must invest in this lovely country? What t…

2023-04-07 09:13:51 Wishing @Cotler a broad and curious audience on her important Northeast speaking tour. https://t.co/vjNCwy3xpv

2023-04-07 09:07:36 Just as Israel's enemies began to celebrate a crack they thought to have found in Israel-US relations, President Biden surprises them with an ironclad rebuff: Stay away, vultures! Go find yourself another prey! https://t.co/OBTZ9Orzy4

2023-04-07 08:43:51 @stephensenn @Raamana_ @f2harrell No comment.

2023-04-07 08:40:00 @stephensenn @Raamana_ @f2harrell This necessary assumption, that "relative bioavailability is transportable from the trial to target population", is what SCM undertakes to enable/verify/test, based on more ascertainable assumptions, like those encoded in the model. We are closer to clarifying the two approaches.

2023-04-07 04:48:25 by “next year in Jerusalem.” Really? Lander's voting base must be so uneducated and gullible that he finds it necessary to apologize for world Jewry, as if Israel takes pleasure in preventing Muslims from praying. Lander, btw, is on Sarsour's band wagon: https://t.co/2nCKQwTWFX

2023-04-07 04:48:24 Evidently, one of those horrified by the idiotic implications of @PatrickKingsley is Brad Lander @bradlander, who rush-rushed to tweet: "Horrific. As we prepare for Passover, Jews around the world have the responsibility to make clear that this is not what we mean 1/2 https://t.co/Nbu6CuBymW

2023-04-07 04:18:03 Tbh, I too didn't expect such a coherent and thoughtful answers from GPT-4. How did you prompt it? Note however that abilities 6 and 8 have already been captured in do-calculus, and other layers too (eg social intelligence) are arguably within reach. Thanks for sharing. https://t.co/cxXVVNeBJe

2023-04-07 04:04:00 The closest I get to the "Humean vs. Kantian" debate is in my article on "radical empiricism and Machine Learning Research" https://t.co/dxqdoqvpOm I try to avoid invoking names of famous philosophers, simply because they did not have computers on which to base theories -- we do. https://t.co/NGM5btxXOz

2023-04-06 22:42:36 @stephensenn @Raamana_ @f2harrell I feel uncomfortable seeing the word "Bioequivalence" used to name "studies". I was under the impression that "Bioequivalence" is a judgmental assumption (say that men and women are bioequivalent in some hormonal treatments), to be tested by scientific studies.

2023-04-06 22:21:09 The most idiotic words are those implying that Israel has any interest in preventing Muslims from praying at the Al-Aqsa Mosque. Words coming from major media outlets, including @nytimes correspondents in Jerusalem. @PatrickKingsley https://t.co/D29SEQDBpp

2023-04-06 21:57:06 @DKedmey This is indeed the question I am trying to answer by playing with GPT4. Whatever the answer, it shouldn't require a "major" AI breakthrough, since counterfactuals have already been algorithmicized.

2023-04-06 18:17:28 IDF says Hamas behind rocket barrage from Lebanon

2023-04-06 18:03:35 Another empty tweet from a Norwegian politician. Can they ever tweet content? https://t.co/jcEtro7gJg

2023-04-06 17:49:47 @dggoldst @MattDeakos I would like to believe that many statistics instructors do rebel seeing how stat textbooks describe Simpson's paradox, but choke the rebellion inside rather than echo Novick's "This is ridiculous!" The stat literature is glaringly deficient of rebels for causal intuition.

2023-04-06 17:10:48 @stephensenn @Raamana_ @f2harrell I am not representing any argument, just sitting quietly waiting to hear what "the purpose" of RCT is, hoping it is somehow related to the clinical questions asked.

2023-04-06 17:02:09 @ArziARStudio Is this my latest photo?

2023-04-06 17:00:44 @f2harrell @Raamana_ I would gladly switch to another idea, once we adapt another definition of what the purpose of RCT IS, rather than what it is NOT. Hopefully, the new purpose would be related to the clinical research question asked.

2023-04-05 20:15:00 @DKedmey @DavidSuissaJJ @bandlersbanter @RabbiWolpe We use the @StandWithUs Haggadah, it has the picture of Herzl there, and a time-line summary of 3,000 yrs of Jewish history.

2023-04-05 11:52:17 @blakeflayton Beautiful answer - symmetry can only go so far. Nakba's victims seek to destroy Israel. Jewish expulsion victims do not seek to destroy Iraq.

2023-04-05 08:55:13 @taha_crypto1 Songs do not prove that someone ancestors were living somewhere (no one tries to prove that), they prove a vivid, authentic and unfakable historical attachment of today's claimants themselves. Try it on China, it won't work.

2023-04-05 08:21:12 on their watch, so that their families can celebrate the Passover holiday the way they did when they were toddlers. Join me (Rabbi Yehuda), if you feel likewise

2023-04-05 08:21:11 Two special features will be added to our Seder tomorrow night: (1) A prayer for the State of Israel (especially "bestow Your light and truth upon its leaders, ministers, and advisors") and (2) A moment of silence in gratitude to the IDF soldiers who are freezing cold 1/2 https://t.co/AvBWtlOtZ2

2023-04-05 07:42:26 @taha_crypto1 @FranceskAlbs @josemercado_ar @BalfourProject @TwitterSupport @TwitterNews I gave you a test: Please point to one historical event prior to 1900 that you are celebrating in songs or a holiday. It's just amazing!

2023-04-05 07:27:41 @taha_crypto1 Fairy tale or not, tomorrow night we will be singing a Hebrew song written in the holy land prior to 500 BC, and we will sing it in Hebrew. Fairy tale or not, we will sing it, atheists included, as if it was us who sang it 2,500 years ago, not "some ancestors". Hard to swallow?

2023-04-05 07:14:50 @daoudkuttab May he live to splash in the Mediterranean together with my grandchildren, when peace descends on us. Amen! Inshallah!

2023-04-05 07:04:07 Is it merely ignorance for anti-Zionists to dis-believe that tomorrow night Jews from Prague, Moscow, Geneva, Berlin, New York, London, Warsaw, Athens, Baghdad and Rabat will be narrating the Exodus story as if they themselves came out of Egypt, the house of bondage? Ignorance? https://t.co/YyDON8WfJF

2023-04-05 06:43:12 @taha_crypto1 @FranceskAlbs @josemercado_ar @BalfourProject @TwitterSupport @TwitterNews Not only is it believable, it is testable as well. I gave you a test: Tell us a story of a Mamluk hero, sing an Ottoman song, celebrate a Canaanite holiday. Go to Biblical times and every Jewish child will pass the test half asleep. Tomorrow night we'll pass it blind-folded.

2023-04-05 01:53:20 @SayedMhmdAli @bentriderdeb I'll walk with you the Straight Path from here to Mt. Olympus, but do me a favor and drop that "Hell Fire" you were using on my friends who love Israel.

2023-04-05 01:22:03 Ingenious. With a few minor fixes, I would be willing to sign this letter. https://t.co/l33UNomUhZ

2023-04-05 01:17:30 @SayedMhmdAli @bentriderdeb "Fire of Hell"?? Watch out! People will quickly recognize the mentality of the culture you came from.

2023-04-05 00:55:01 @taha_crypto1 @FranceskAlbs @josemercado_ar @BalfourProject @TwitterSupport @TwitterNews Claims based on strong current historical connections are at least as valid as those based on physical presence void of history. Tell to us a story of a Mamluk hero, sing to us an Ottoman song, before you dismiss a nation of home-comers as having an "illusion".

2023-04-05 00:37:35 @hartificialman what was missing?

2023-04-05 00:29:59 @MattDeakos do not use the treatment, but if the gender is unknown we should use the treatment! Obviously that conclusion is ridiculous. (Novick 1983, p. 45). If a statistician calls a conclusion "ridiculous" wouldn't anyone else? 2/2

2023-04-05 00:26:19 @MattDeakos Why? I thought every thinking organism would rebel hearing of a drug that is good for me good for women and is bad for a person. Hear what Novick says: "The apparent answer is, that when we know that the gender of the patient is male or when we know that it is female we 1/2

2023-04-05 00:05:58 @abiylfoyp Nice, the key argument is #3, which a normal person would use without the fancy words of "risk benefit analysis". But, as GPT-4 admits, it's not too good in logic as it is in collecting texts and rumors. Both are needed.

2023-04-04 22:54:40 @JoshuaParkhurst @Ostrov_A He didn't even say "maybe". All he said was that some foreign diplomats are concerned about such possibility, which is exactly what Israel's Govt and media have been saying for years, i.e., "some outside sources say." Time to accuse and time to heal!

2023-04-04 22:38:20 @academic_la I just ordered 3. One for Passover, One because it's really good and one for BDS's sake. How lucky we are that one key stroke can make us feel like we have made the world a better place.

2023-04-04 18:25:29 @taha_crypto1 @FranceskAlbs @josemercado_ar @BalfourProject @TwitterSupport @TwitterNews The Palestinian tragedy has been, and still is the illusion that Israel was created by some "colonial settlers", rather than a home-coming people, equally indigenous to the land. I hope your aim is to end that tragedy, not to prolong it.

2023-04-04 12:29:50 @FranceskAlbs @josemercado_ar @BalfourProject @TwitterSupport @TwitterNews I can't hardly think of anything more racist than stripping one people from its one state.

2023-04-04 11:34:37 Thanks for bringing the #BalfourProject to my attention. And, just in case you missed the ubiquitous significance of the Balfour Declaration, here is my reminder: https://t.co/9XOCOs9hGT Please read it at the Conference. https://t.co/r1ubI3XaDw

2023-04-04 08:06:47 @PrzemekPWr @EinatWilf @Ostrov_A @DKedmey @academic_la Glad you asked. I had the same click. Social Intelligence and Political Intelligence should be two AI branches to emerge from causal inference.

2023-04-04 07:46:42 Fascinating new idea: Israel's democracy has its roots in the semi-autonomous political life of Jewish communities in the diaspora -- and that's the secret of its resilience. Illuminating!! @EinatWilf @Ostrov_A @DKedmey @academic_la https://t.co/tEcH8glFA6

2023-04-04 05:57:45 @drstoyanpavlov @TeddysMom8 I find it very useful as a representative of "a man in the street" (which I very rarely meet). I am disappointed ocassionally when it doesn't argue like "a man in the street." It's still very useful for me to know what shallow internet conversations have been like.

2023-04-04 05:49:58 Monday brings us another bunch of causal inference publications https://t.co/exe8UpN4xi. My old fascination with philosophy was reignited by the paper of Vasse, defending Causal Decision Theory. I was surprised though that the debate with Evidential Decision Theory is still on.

2023-04-04 04:28:52 @TeddysMom8 That's why I wished GPT-4 would argue: "Risk should be considered, fine, but what about all those who were saved by the vaccine? And all those who would die if we ban the vaccine, even temporarily. " (as we do in #Bookofwhy page 44)

2023-04-04 02:51:19 Moving GPT-4 from Simpson to Smallpox. Me: If more people died from smallpox inoculations than from smallpox itself, what's wrong with the argument that inoculation should be banned until it is made safe. GPT-4: You raise a valid concern, and it's important to consider the 1/2

2023-04-04 02:21:31 @julioalexo It is essential to separate the two because, invariably, university administrators take antisemitism to mean "a license for inaction". They designate a "task force" with four philosophers to decide what it is, and that's where it gets buried. See USC https://t.co/vXxPyUokOW

2023-04-04 02:02:14 @RobotLaunch I believe 1Billion is very conservative figure, counting educational budgets, Gothic Cathedrals and industrial training.

2023-04-04 00:34:41 ‘Historic Victory’: US Gov’t Rules That University of Vermont Failed to Address Campus Antisemitism https://t.co/pC7aL6pIeK

2023-04-03 21:54:40 @Drombolombo Why the weighted average argument works in causal setting and not in Simpson's Paradox: See The Sure-Thing Principle: https://t.co/ODjYZuaAFZ

2023-04-03 21:47:13 @Drombolombo I meant grade in school, but it could also be pay grade. The answer is "definitely NO" because drug that is good for me good for women and bad for person does not exist. Sure Thing Principle . A theorem in do-calculus.

2023-04-03 18:52:36 @dimgrr Nice, except I've never heard of the "Principle of consistency", especially when in stochastic settings, where by "X causes Y" we mean "X raises the probability of Y". Worth asking for reference to "consistency".

2023-04-03 11:49:51 Why the weighted average argument works in causal setting and not in Simpson's Paradox: See The Sure-Thing Principle: https://t.co/ODjYZuaAFZ https://t.co/U3QfNoXO79

2023-04-03 11:42:05 @mcdonalds_tim The answer is correct, but the reasoning is not, because the rule of total probability applies as well to Simpson's paradox, where the issue is association, not causation and, Lo and Behold, there we do have sign reversal. The impossibility is a theorem in do-calculus.

2023-04-03 09:39:59 I am curious about your conclusion: "Adding more controls need not bring you closer to identification! " So, how does one know if the controls available are sufficient, redundant, or bad? https://t.co/nIy7nyffFr

2023-04-03 08:59:49 @KarlPettersso10 I am not sure about the meaning of "over a life time". Can you prove it from do-calculus?

2023-04-03 07:48:59 @davidboxenhorn We should not conflate AI with GPT-4. The latter is indeed very useful for reflecting the current wisdom of the crowd which is often not reflected in the literature of in what your politically correct colleagues tell you.

2023-04-03 07:43:14 @bytebiscuit Every stat instructor can parrot this slogan. I'm trying to test causal understanding.

2023-04-03 07:35:38 Unfortunately, it didn't get it. Me: Shouldn't the same argument works in Simpson's Paradox? GPT-4: If smoking is consistently associated with a grade decrease within every age group, then it should be associated with a grade decrease overall, as you correctly pointed out. Wrong https://t.co/45aRenlUjh

2023-04-03 07:26:15 Luckily, it's teachable: Me: But if smoking causes a grade decrease within every age group, shouldn't it also cause an average grade decrease overall? Isn't the latter just a weighted average of the former, which makes it impossible to reverse sign? GPT-4: "You are correct. 1/2 https://t.co/TCmSQRO0ps

2023-04-03 07:14:36 @rootsaandwings No. By "causal" we mean the Probability of grade increase if you decide to smoke is higher than if you decide not to smoke.

2023-04-03 06:46:56 not associational (" Is it possible that smoking CAUSES ...")

2023-04-03 06:46:55 for smoking to cause an average grade increase overall while causing a grade decrease within every age group. However, this scenario would likely involve some form of Simpson's paradox, where the overall relationship between two variables reverses when accounting for 2/4

2023-04-03 06:46:54 Got my first session with GPT-4, amazing! Though it failed its first causal understanding test. Me: "Is it possible that smoking causes grade increase on the average and, simultaneously, smoking causes grade decrease in every age group?" GPT-4: It is theoretically possible 1/4

2023-04-03 04:04:34 @noatishby Sad to see you go, Noa, but the war is getting nastier, and your comrades in the trenches need you. You do not need a title to be part of the War Cabinet -- your seat will be waiting for you.

2023-04-03 03:47:06 A re-discovered hard truth: Israelis love their country. "Now that they’ve seen who they voted for, many of the coalition’s voters have buyer’s remorse". @DavidSuissaJJ

2023-04-03 03:47:05 The ocean of Israeli flags that have dotted the landscape in 2023 speaks to a deep, visceral patriotism that cuts across much of Israeli society. https://t.co/Urb3hRX4Jy

2023-04-03 00:29:25 @forde_nathaniel @stephensenn @elmir1omerovic @f2harrell @kaulcsmc Enders's analysis is still stuck in Rubin's marshland. It warns readers: "These assumptions ... are inherently unverifiable. Your analysis will stand or fall depending on how plausible each assumption is..", yet makes them as incomprehensible as possible using stat notation.

2023-04-03 00:16:19 @setspace @ralfduschef This is precisely why we need to understand humans using new tools, new languages and new logic.

2023-04-03 00:13:10 Refreshing to see missing data analysis elevated from the statistical marshland in which it was initially hatched, towards causal modeling, where it obtains transparency, identification and testability. https://t.co/XXvdNxs00p

2023-04-02 23:57:00 @f2harrell @elmir1omerovic @stephensenn @kaulcsmc @rlmcelreath It's not so much the difference between Bayesian and frequentist approaches, as it is between statistical and causal thinking. The latter identifies when the former makes sense, and adds to it transparency plus testability.

2023-04-02 22:16:59 RT @forde_nathaniel: @stephensenn @elmir1omerovic @f2harrell @kaulcsmc @yudapearl Not sure which bucket I fall into there, but I found the…

2023-04-02 21:10:04 @f2harrell @matloff @elmir1omerovic @stephensenn @kaulcsmc @rlmcelreath Beg to doubly differ. Missing Data is a causal, not statistical problem. Causal analysis is indispensable for showing when statistical heuristics (e.g., multiple imputations) work and when they don't. See examples: https://t.co/1BNZSxTYQE

2023-04-02 18:45:13 Statisticians' wars over p-values and Bayes' subjectivity are merely a prelude to the gigantic wars still awaiting them when the number of samples increases and at issue is whether statistics is the language of science. See "Why I am only a half-Bayesian" https://t.co/gqeQb2lNBh https://t.co/5kBDbIndkM

2023-04-02 09:42:40 RT @NTarnopolsky: *The Biden Effect* Tel Aviv right now https://t.co/nZb6ZA1pGb

2023-04-02 07:55:21 @3scorciav @AlexTensor @berkeley_ai NSF and top-funding agencies do fund some small-scale research in Rungs-2&

2023-04-01 11:06:01 @ralfduschef I can't buy this dictum: "The social building of trust is beyond engineering," perhaps because I remember how people argued (before 1947) that "playing chess is beyond engineering".

2023-04-01 10:29:30 @Alberto_Begue_ @HannesMalmberg1 I should get it. Thanks.

2023-04-01 10:17:16 @3scorciav @AlexTensor @berkeley_ai Asking "what cues, like the logos on uniforms, it used to figure out who the Boston Red Sox are" is a Rung-1 task. Explainability is a Rung-3 notion. NSF's projects have 2 more rungs to go before they can explain why the Red Sox coach should be fired. Rung-1 gulps all resources.

2023-04-01 09:40:06 @HannesMalmberg1 Beautiful! What Chatbot is it?

2023-04-01 08:23:06 Call me sentimental, @jazchaz, but don't ask me why tears start sliding down my cheeks when I hear this song, and why it makes me so absolutely sure Israel will survive this crisis. Here is the original, sung by Gali Atari in 1986: https://t.co/OBsCBMvJhf https://t.co/909mNtEUSF

2023-04-01 02:26:07 RT @realbassemeid: The BDS Movement's campaign against #Yale is an attempt to ostracize the Jewish community and promote hatred against Isr…

2023-04-01 01:50:55 That's another reason why I like Twitter. https://t.co/knxx0tx9DR

2023-04-01 01:47:58 RT @dylanarmbruste3: @nick_krontiris Bayesed

2023-04-01 01:39:05 I'd strongly support the idea of a Manhattan Project of intense research to make machines more trustworthy and interpretable (regardless of, or in parallel with a moratorium.) The premature super-investment in non-interpretable technologies is the core of our problems. https://t.co/CJ9Q4e3vlw

2023-04-01 00:37:06 RT @DrMichaelOren: After a week that left many Israelis wondering, “Where do we go from here?” I take a look at the existential consequence…

2023-03-31 18:31:26 @fhuszar Counterfactuals help us reason the world in which we live. I find it hard to imagine anyone interested in the process by which GPT-4 chooses its words

2023-03-31 10:08:32 RT @_B___S: This is how they sleep every night. https://t.co/lPkTgc4PCl

2023-03-31 10:03:16 @MichaelofAlbans Agree! But if we let government to get there too, we will not witness freedom reclaimed, but freedom doubly oppressed.

2023-03-31 09:09:24 I am on the waiting-list for GPT-4 access. Once I get it, my first task would be to teach it counterfactual reasoning on the firing-squad story (GPT-3 had hard time grasping that I have zero interest in whether shooting rifles is legal in California), and then move to an 1/2 https://t.co/1lBQLsEqvU

2023-03-31 08:35:34 @CotlerWunsh Hear, Hear, in its entirety, and add to it the Prayer: "Barech et Medinat Israel, Resheet Tsimat Geulatenu... and send your light and truth to her leaders and experts and bless them with your good advice." Amen!

2023-03-31 08:20:59 @fhuszar The theorem says nothing about the causal graph

2023-03-31 07:38:28 @CotlerWunsh Agree on the potentials that this crisis holds for Israel: The backbone of Israel's society has discovered its true identity

2023-03-31 07:01:28 I strongly condemn this vicious, defamatory and baseless post by @WounterDeVriendt against @GeraldNGOM. I hope decent members of the Belgian Parliament will examine the facts and join me in this condemnation. https://t.co/ijolBUDv9h

2023-03-31 06:35:56 Prof. Yuval Noah Harari: Judicial coup in Israel ‘will destroy academic freedom’ https://t.co/TyuSnXdVaj

2023-03-31 04:57:09 First good news from CUNY (The City University of New York). Congratulations @SAFECUNY! https://t.co/X1fZlZrJYv

2023-03-31 04:45:35 @agarret7 Apologizing - My acronym-knowledge is limply.

2023-03-30 21:49:54 @ATaher_ @julioalexo @_ZachFoster The Mufti had a stronger grip on the Palestinian people in 1930's than any "president" I know. Are you disavowing his teachings vis-a-vis the need to crush Zionism, and to prevent European Jews from escaping Europe?

2023-03-30 21:43:49 @skassabr No. I do not know of any Palestinian capable of uttering the words: "Equally legitimate, and equally indigenous" when the objects of discourse are the Jewish and Palestinian PEOPLES (forget "religions"). And I have challenged them on at least 100 occasions

2023-03-30 19:53:33 @KillianForde When was the last time you told your son that his real home is Bergen, from which you were expelled by force, that you never signed off your property rights to anyone, and that he should expect to return there soon, to sing again the poems written there, which he knows by heart.

2023-03-30 19:41:57 @skassabr Your key words are "As well," which I wholeheartedly accept. But, have you ever seen a Palestinian acknowledging Israel's historical roots? Or capable of uttering the words: "Equally legitimate and equally indigenous"?

2023-03-30 18:25:01 Thank for posting. I wasn't aware of the fact that the Peel's Commission (1937) stated explicitly that the Arabs charges had no basis. The Arabs refused, at that critical moment of history, to consider European Jews as part of the deal, thus sealing their fate in the ovens. https://t.co/RgtgEobx6H

2023-03-30 12:42:27 @julioalexo @ATaher_ @_ZachFoster I can understand why Hajj Amin al-Hasayni befriended Hitler and mistook Zionists to be crusaders. What I cannot understand is why Palestinians today, especially their intellectuals would perpetuate the Mufti's mistakes to prolong and amplify their people sufferings.

2023-03-30 12:04:25 @FJnyc According to Rashid Khalidi, Palestinian national identity was forged in the 1920s, but it was too late "to crush the Zionist movement". It's the only national identity in the world forged on "crushing" rather than collective memories.

2023-03-30 10:49:13 @MSkywalker @NewwAmericanna @_ZachFoster this yearning as legends of the past

2023-03-30 10:40:41 @MSkywalker @NewwAmericanna @_ZachFoster The idea that "We never really left that land" is not a Zionist invention, but a thread governing governs 80 generations of Jewish thoughts. My grandfather said 3 times a day "The merciful one will soon walk us, in sovereignty, back to our country." So, please, do not treat 1/2

2023-03-30 10:25:57 and stories about causal assumptions as assumptions. The conversion, from stories to substance, would be very useful as a user- interface in many causal inference tasks. We need to make sure, though, that it can handle the substance properly, assuming a proper conversion. 3/3

2023-03-30 10:25:56 GPT-4 should know that my statements about DL never rising above "curve fitting" is not an opinion, nor a "common concern" but a mathematical theorem. You cannot answer causal questions from passive data alone, unless you enrich the data with: (1) causal assumptions or 1/3 https://t.co/EcRzBWHzpm

2023-03-30 09:41:19 @MSkywalker @NewwAmericanna @_ZachFoster You make it sounds like an afterthought

2023-03-30 09:20:44 @ATaher_ @julioalexo @_ZachFoster The tragedy of the Palestinians started with the Mufti of Jerusalem, Haj Amin Al Husseini, who could not see the difference between crusaders and home-coming neighbors. It continues today.

2023-03-30 05:37:02 *One case of settlers whose holidays commemorated historical events in the land to which they moved, not in the lands from which they came. • One case of settlers who did not name towns "New XXX", like New York or New Wales, but by names those towns enjoyed in ancient times 2/3

2023-03-30 05:37:01 Sounds like our "historian" has no history. * Show me one case of colonialists moving into a country they perceived to be the birthplace of their history • One case of settlers speaking a language spoken in the old land before the one spoken by its contemporary residents 1/3 https://t.co/LICoM8Rg04

2023-03-29 21:51:35 When a Palestinian intellectual says "It is amazing ..." you can expect a shallow regurgitation of some populist slogan, combined with a profound emasculation of Palestinian agency. @dianabuttu @Ostrov_A @EinatWilf https://t.co/6QJDYX78RF

2023-03-29 21:30:03 Note also: The name was "Palestine Liberation Organization", not "Palestinian" -- an essential distinction, when it comes to aims and agenda. https://t.co/6pBM837w2M

2023-03-29 21:13:17 Can some pro-Palestinian activist explain to their "historians", intellectuals and spokespersons that denying Jewish roots in the holy land does more to prolong the occupation than all terror attacks put together? They don't seem to get it! https://t.co/um64P3YWaS

2023-03-29 20:49:22 @desai_pratik I have the same feeling but I can't put my finger on any scholarly work on this theory

2023-03-29 19:39:09 @Anatomi54244944 @PrzemekPWr Zionists covenant with G-d was: We will support your plan about the end of days, if you support ours about tomorrow -- he signed it.

2023-03-29 19:28:43 @hearnimator And there should always be a discussion of Zionophobia in the context of Islamophobia

2023-03-29 11:24:18 @blakeflayton I witnessed an earlier peace rally -- the "Land for Peace" euphoria of August 1967, before any settlement was in sight. It was rebuffed by Khartoum's "Three No's" which proved the occupation inevitable. The emasculation of Palestinians agency is fundamental to their history.

2023-03-29 05:58:57 @chaimlevinson @bezalelsm Amnon Shapira was my High School mate at Ironi-Datti (Zeitlin). We were both in Bnei-Akiva when it was a Torah V'Avoda movement, before it was highjacked by messianic ghosts. Shalom Amnon, Shalom Chaver, Yehi Zichro Baruch.

2023-03-29 02:43:44 @AleksanderMolak @yunliwoo Unfortunately many stat-ML authors are now adding the key word "causal inference" to their papers, though they have nothing to do with causal inference. I am using my own formula, it's not perfect, but captures the low lying fruits.

2023-03-28 11:42:25 Like other Mondays before this, we are welcoming 5 pages of new publications in the many aspects of causal inference: https://t.co/w2UUQEuM45 Please search for and share breakthroughs.

2023-03-28 10:06:26 @jeremy_b12345 @DuaneJRich If you are stuck on RCT and looking to action something useful from #Bookofwhy, consider generalizing RCT findings. Can you do it without do-calculus? I doubt it. Now watch this: https://t.co/dEPwcv3l40, or this: https://t.co/vdYIoZ3wSP, or this: https://t.co/w5fRMmcBkg.

2023-03-27 18:34:26 If I were not a disbeliever, I would swear that it was God who listened to my prayer and implanted a seed of responsibility into Bibi's head, to halt the Judiciary overhaul and allow time for reaching wider consensus. But given who I am, I'm waiting for him to reinstate Galant. https://t.co/AvBWtlOtZ2

2023-03-27 13:03:59 McGill Tribune Rejects Pro-Israel Column Submitted By Jewish Student Saying Zionism Doesn’t “Align With Our Values” https://t.co/1jkCj7jqGZ via @honestrepcanada

2023-03-27 11:01:07 My anxiety over the events in Israel has risen to the point where all my secular convictions have given way to a prayer: Oh Father in Heavens, Bless the State of Israel, Shield it with your love, and spread over it the canopy of your peace. Please join me: https://t.co/y4evnxXNov

2023-03-27 10:39:55 Ben Gurion departures halted as Histadrut declares ‘historic’ strike over overhaul https://t.co/IEgXlWYIZV via @timesofisrael

2023-03-27 01:23:29 Nice test. I was suggesting generalizing from story to story, but time is a much better filter against bluffing. https://t.co/kalDvqk63O

2023-03-27 01:17:03 @Michael_Wgd The photos are distasteful, but ignoring half of the country is a disaster.

2023-03-26 23:43:54 @ddkirsch @PHuenermund @jhurwitz @GaryMarcus @CausalInferBot I can't brush it off so easy. I've defined "understanding" as the ability to answer questions on all three levels of the Ladder. Sticking with this definition, GPT-4 has not failed yet, especially if it can generalize from story to story. I haven't tested it on this dimension.

2023-03-26 23:37:45 @michelnivard @PHuenermund @ddkirsch @jhurwitz Very nice! From now on, if anyone has a question about #Bookofwhy, don't ask me, ask GPT-4.

2023-03-26 22:07:07 Will Bibi lead Israel to suicide? Tamut Nafshi? Cry, the beloved country! Za'aki Erets Ahuva! Livshi Bigdei Tifartech, Ami! https://t.co/OSGEYw5Xkw

2023-03-26 19:54:14 For this reason, GPT's understanding should be tested on questions whose answer is "NO, it is not possible under any circumstances". Such questions are the core of causal logic. e.g., can a drug be good for men, good for women and bad for people? https://t.co/OoViF1fBx0

2023-03-26 19:46:49 Netanyahu fires defense minister Gallant for calling to pause judicial overhaul https://t.co/KXNCbxmsvd via @timesofisrael

2023-03-26 19:36:23 @ATabarrok @PHuenermund @ddkirsch @jhurwitz Fairly amazing. GPT seems to have read #Bookofwhy. The answer is wrong though. The question did not state that K and U are initially uncorrelated. If they are correlated then it is feasible that they would turn uncorrelated after conditioning on X.

2023-03-26 19:27:25 @itamarcaspi @PHuenermund @ddkirsch @jhurwitz Good suggestion, but #Bookofwhy was written for people, i.e., a species with inborn crave to represent its understanding of the world in a road-map called "causal model". Unless CHAT is endowed with such craving, I doubt he could understand #Bookofwhy.

2023-03-26 19:17:46 @PHuenermund @ddkirsch @jhurwitz CHAT can always get away by saying: "it all depends" . And HE is right, it does depend. To test his understanding, we should specify the scenario unambiguously, eg. two coins and a bell, oxygen-match and fire, two soldiers and a prisoner, etc. #Bookofwhy has dozens. @GaryMarcus

2023-03-26 13:24:53 @JerryJjjjjjjjjj Asking a "Palestine Historian": "Why do you lie?" is like asking a Holywood producer: "When did you shoot your last movie?"

2023-03-26 13:16:06 @GeraldNGOM @KenRoth It sure is an expression of hate, even racism, but why give the racist an instrument to deflect his racism, and to cover it with populist slogans that exonerate him from antisemitism. Why not pinpoint precisely, and undeniably, the kind of racist he is -- an ugly Zionophobe.

2023-03-26 13:00:03 RT @j_temain: Whatever one’s political leanings or opinions of the current Israeli government, the protestors, or the judicial review, can…

2023-03-26 12:50:36 "Jewish study major" - Another glimpse at the sophisticated educational level reached by the "Jewish Studies Association". @StudiesZionist https://t.co/KXLGeFzTq9

2023-03-26 12:38:13 I knew he would take that route, and that is why I was careful NEVER to call @KenRoth an antisemite, always a Zionophobe -- a stronger indictment which he cannot deflect nor deny. https://t.co/1DN6DzIU2a

2023-03-26 12:18:16 @gilled34 @avrahamkorea That's the time to say: We erred!

2023-03-26 12:16:56 @malagadatos Could not get the question.

2023-03-26 12:15:49 RT @RabbiWolpe: Is Israel heading toward disaster? (10 min) https://t.co/7YLhfvv46x

2023-03-26 11:43:21 @gilled34 @avrahamkorea Agree. All eyes are now looking at Likkud ministers to take a stand against the reform, but perhaps more effective would be for the staff of Kohelet to come up and say: "We erred! We underestimated this or that.. We erred." They will be forgiven, and commended for their courage.

2023-03-26 08:03:56 @avrahamkorea @gilled34 Stupid and extremist calls come from both sides, what does not come is a clear analysis like this one: https://t.co/JNCrX2lj0z

2023-03-26 07:45:45 @avrahamkorea @gilled34 No one expected that MANNER, the disrespect, the bulldozers and the 4 extortionists tearing their pound of flesh from the rest of society.

2023-03-26 06:51:19 @jiaxx213umn I'm stuck on the questions and the input information.

2023-03-26 06:49:48 @avrahamkorea I heard them too. But as King Solomon the wise said: There is time for everything under the heavens. A time to give birth, and a time to die

2023-03-26 06:43:12 Palestinians have created a new logic for cause-and-effect, called "triggered by Zionists". It has already given rise to a great number of scientific innovations, startups, and research universities. Watch @AhmikAser for additional progress. https://t.co/GT2M5FfgIq

2023-03-26 06:16:29 Israel refuses to surrender, 300,000 protesters in Tel Aviv alone. Cry, the beloved country! Za'aki Erets Ahuva! Livshi Bigdei Tifartech, Ami! https://t.co/6Dpsw8zLnM

2023-03-26 01:26:30 Israeli defense minister, Yoav Gallant, calls for halt to judicial overhaul. The ball is now in the hands of other sane ministers in the government

2023-03-25 16:28:44 RT @vardi: https://t.co/eb9NGi9ZnW

2023-03-25 15:54:08 @blakeflayton Same goes for 'Jewish Voice for "peace"', 'New Israel Fund', perhaps also schizophrenic J-Street, regardless of whether they're overtly or covertly pro-BDS.

2023-03-25 08:59:07 For our Hebrew speaking readers, Prof. Yuval N. Harari (author of Sapiens) is here giving the best analysis I've found of the Judicial crisis in Israel: https://t.co/JEcc5hfWUJ He mentions a pending decision by the academic senate to close the Hebrew University in Jerusalem.

2023-03-25 00:41:00 Jerusalem security forces on alert as tens of thousands at 1st Friday Ramadan prayers https://t.co/22kdytHJH2 via @timesofisrael

2023-03-24 23:37:02 @marciesilvers1 The lullaby "Palestinians want peace" has long been replaced by "From the river...". They now get offended by the former. "Peace with whom?" they ask, "with an entity that does not exist?" Don't offend them, please.

2023-03-24 22:55:35 When @jeremycorbyn calls for justice, equality and peace we should all freeze in attention

2023-03-24 22:49:51 40 years of Zionism brings us to 1936, when I was born, and we were 400,000 Jews in Eretz Israel. (By records of the Mandate Authorities) https://t.co/52FEgla5Qh

2023-03-24 22:37:49 When a reader asked: "Do Palestinians love Allah more than they hate Israel?" other jumped: "How dare you offend a whole culture?". I think a culture born on negation needs to be understood that way. https://t.co/eknRO9TNR7

2023-03-24 22:25:59 @DavidHirsh I don't think Zionophobes care about France

2023-03-24 22:13:33 RT @HenMazzig: After the protests in London today I went to the BBC News studio to explain why Israelis, who love their country, are protes…

2023-03-24 22:08:08 @jiaxx213umn I cannot tell from your paper what questions can be answered by your do-DAGs that cannot be answered by ordinary DAGs. Also, if there are such questions, does a do-DAG convey more information from users?

2023-03-24 06:12:35 Do you have a GPS in car? Note that, in addition to those sweet "turn right, turn left" instructions, you also get a road map. Why? To enable you to pick a detour, in case you hit a road block or a construction site. So, next time you hit a perturbation over what your LLM 1/2 https://t.co/qNahmGDqJX

2023-03-24 05:21:40 RT @academic_la: Brilliant.

2023-03-24 05:06:20 @GeraldNGOM @yoavgallant @netanyahu I was hoping

2023-03-24 05:03:05 @DavidDeutschOxf @michael_nielsen Is there a test to distinguish between the two?

2023-03-24 04:57:24 RT @SAFECUNY: We can't emphasize this enough: CUNY PAID for this antisemitism. This wasn't students expressing hateful free speech. It wasn…

2023-03-24 04:53:35 @AlexanderNaume2 Fine, but this does not imply sufficiency. We still need to differential results of actions from results of observations, and more.

2023-03-24 04:43:23 @AlexanderNaume2 I fail to see why it is sufficient.

2023-03-24 04:19:39 This demonstration takes place a few feet from the house where I was born. But in my days we didn't have such demonstrations, we played together Haredim, Chofshiim and Masortiim. It was before the Dark Ages, then the city renamed Herzl Steet Rabbi Schach St. --dark onto darkness. https://t.co/5uNmUFZmvZ

2023-03-24 03:43:42 Peoples with no historical records are yearning to go back to the Ottoman Empire, arguably the most repressive, corrupt and dis-functioning regime in human history, to comfort their need for victimhood and to negate their neighbor's history. https://t.co/zewEqBVugk

2023-03-24 03:22:54 The opening plenary set the tone by urging attendees to think big, to dream big, to see far down the road. https://t.co/rHe58SiM6o

2023-03-24 03:05:57 @AngeloDalli @soboleffspaces @stephensenn @JohanDH2O @RWJE_BA I can resonate to (1), (4) and possibly (5), because I've tried them on the logic of causation and proven them valid. I'm hesitating about 2 &

2023-03-24 02:16:07 make up for missing models. Still, model-free RL cannot fully climb to Rung-2 of the ladder, as shown by @eliasbareinboim https://t.co/0GfXdwdTkp, not to mention climbing to Rung-3. 2/2

2023-03-24 02:16:06 Commenting on: "DeepMind successfully demonstrated “a recurrent network with model-free reinforcement learning to solve ..[causal problems] (Dasgupta et al., 2019). RL uses actions, to 1/2 https://t.co/3JB6jVhjUf

2023-03-24 01:58:19 @Dude62261752 I don't know any of my students, past and present, who believes in either: (1)Model-free deep learning, or (2) "AGI is impossible".

2023-03-23 20:14:04 I havn't read this essay before - thanks. It nicely compiles the arguments why AI needs causation, but I do not find its arguments against AGI to be compelling

2023-03-23 17:54:29 What happened to CUNY? - A fairly incriminating report!!! https://t.co/xZWp9LSfiL

2023-03-23 05:02:01 @prem_k @soboleffspaces @GaryMarcus Thanks for posting this article. It mentions causation 5 times, but I am not sure I would agree with the authors on how they phrase it.

2023-03-23 03:54:33 Refreshing to see Students Supporting Israel carry the torch of sanity to the heart of the "2nd Annual Palestinian Mentality Week", which started today at UC Berkeley! (They call it "apartheid week", but the idea is the same: From the River to the Sea, no one else but We &

2023-03-23 01:45:34 @soboleffspaces I wouldn't call it a 'framework' until I see at least: (1) A new question answered by the integration. (2) A recognized CI researcher in the References, (3) An equation describing what one hopes to get out of the integration.

2023-03-23 01:25:52 @soboleffspaces @stephensenn @JohanDH2O @RWJE_BA I agree that the integration problem should better be phrased as "what questions can be answered by {RCT+RWD} combinations, that cannot be answered by each component alone?" The ability to answer questions entails logical deduction, and rules out hand-waving.

2023-03-23 01:16:16 @soboleffspaces @rtaira18 @stephensenn The authors are not known as CI contributors, so one would expect them to be unaware of the Ladder of Causation and of what questions can be answered by what data+assumptions combination.

2023-03-23 00:13:18 The reason I asked "Can they do it?" is that I do not believe it can be done without understanding the principles of integrating RCT and RWD data, and those principles will get buried in noise unless we can demonstrate them on toy examples. Moreover, though everyone agrees 1/2 https://t.co/Vq19icLj6T

2023-03-22 20:48:40 @soboleffspaces @dylanarmbruste3 @stephensenn @BD_Zumbo Some philosophers (I among them) would take it as evidence that humans carry DAGs in their heads. But I am sure others would object.

2023-03-22 18:53:55 @rtaira18 @soboleffspaces @stephensenn This is an extremely useful article, calling for "(insightful) integration of RCT data with 'real-world-data (RWD)". First, I'm delighted to see RWD gaining respectability. Second, I'm unable to spot any CI researchers among the references, which worries me - can they do it?

2023-03-22 11:52:45 @_J_Pe Would you still resist the temptation if Putin's mentality was driving force behind Russian national identity?

2023-03-22 11:21:17 As readers know, "attribution science" deals with the "probability of necessity" (PN), a counterfactual term that is analyzed here https://t.co/Zy7uTB30gV and https://t.co/v3KQZbMZ4k. Note that, in climate context, we do not have experimental data, but monotonicity holds. https://t.co/aZSpqWeMJM

2023-03-22 10:33:28 @_J_Pe The wisdom you wish to be distilled from history teaches us that not all "national identities" are equally benign, some can even be dangerous, especially those who are bent on denying national identity to their neighbors.

2023-03-22 09:18:36 @Dude62261752 As I read this paper, it assumes a (triangular) structure of IV type, and further imposes shape restrictions (eg monotonicity) on some of the functions. In our "Personalized Decision" https://t.co/EIzi5wSxW4 we make no structural or functional assumptions, hence the bounds.

2023-03-22 01:31:02 whether the success observed can be improved, or whether it is not due to some temporary condition, etc. We then seek conditions under which success is guaranteed. "Guarantee" is a logical notion, demanding deduction. A good example is the practice of "taking the average" of 2/4

2023-03-22 01:31:01 "Proven by practice" may be a good guideline for repetitive routines for which we have a valid criterion of success (e.g., using seatbelts). But when we seek justification for a totally new practice (Fisher's RCT), we can't rely on practice alone. Also, we may wish to ask 1/4 https://t.co/af7ABIvl40

2023-03-21 21:44:49 Forgot to tag @stephensenn @soboleffspaces

2023-03-21 21:32:02 @_StephenOlivier Great question, which has been floating in the air throughout these discussions. "justify" means proving deductively that the result of a given exercise (OS or RCT) gives the correct answer to our research question: "What's the Causal Effect of treatment T on outcome Y."

2023-03-21 21:11:29 Which confirms my theory: "Only he who can justify OS can also justify RCTs." https://t.co/Vmsu9JWMLe

2023-03-21 21:04:57 The 5th question for the Passover Haggadah of Ex-Jews: "Why the hell were we ever Jewish? We don't even live "there" (the holy land)

2023-03-21 20:22:33 Which proves that causal diagrams can turn even physics into a colonial enterprise. What a powerful tool! https://t.co/uVVi4sH7Dl

2023-03-21 20:04:33 @AviMayer @Jerusalem_Post Congratulations, Avi. B'Hatslacha

2023-03-21 19:54:39 @soboleffspaces @stephensenn In other words, only he who can justify OS can justify RCTs.

2023-03-21 19:10:54 @stephensenn @soboleffspaces Your point is well taken. My point: People forget that "randomization" as used in RCTs is not random sampling, but "randomized intervention". In other words, we need to mention "intervention" explicitly, because it is not automatically understood from the word "randomization".

2023-03-21 19:03:49 @dylanarmbruste3 @soboleffspaces @stephensenn @BD_Zumbo inferences are only as credible as the causal (judgmental) assumptions that enters into our model. The latter became articulable only with the advent of DAGs. 2/2

2023-03-21 18:57:21 @dylanarmbruste3 @soboleffspaces @stephensenn @BD_Zumbo I didn't find this part illuminating, perhaps because today we know so much more (about OS justification) than anything that could be articulated in 1990. Why? Because in 1990 we could not articulate the assumptions needed for valid causal inference, and we know that all 1/2

2023-03-21 18:38:59 @stephensenn @soboleffspaces in addition to randomization. The idea that patients are forced (or advised) to act in accordance to a coin's outcome, possibly contrary to their natural inclination, elevates RCTs to level-2 of the Ladder, and allows us to infer causal effects

2023-03-21 18:28:58 @stephensenn @soboleffspaces I've found that "randomization" is well understood, because it is easy to imagine a process of selecting samples based on coin's outcomes, and because randomization is the basis of statistical surveys. What is harder for people to realize is that RCTs invokes "interventions", 1/2

2023-03-21 17:51:42 @soboleffspaces @dylanarmbruste3 @stephensenn @BD_Zumbo It's truly interesting to read Greenland's article of 1990. DAGs and d-separation were already in existence, but not causal DAGs (1991), for this was before I met Sander and Jamie(1993), a meeting that led to the causal revolution in Epidemiology, 1999: https://t.co/OFS5Cu6PH2

2023-03-21 09:48:06 Palestinians national identity, because it is only 102 years old, has shallow historical roots. It's sad, because roots make nations secure and pragmatic. Instead, it revolves around negating their neighbor - a futile battle, given the neighbor's stubborn insistence on life. https://t.co/w54Y7urg36

2023-03-21 08:10:08 I love it: "Ignorance is a choice". https://t.co/Bp10GD7AZh

2023-03-21 07:52:40 Palestinians national identity, because it is only 102 years old, has no historical roots - an ingredient that makes other nations secure and pragmatic. Instead, it revolves around negating their neighbor - a futile battle, given the neighbor's stubborn insistence on life. https://t.co/fFp7jqjPZW

2023-03-21 07:33:11 Westerners will never dig Palestinian mentality. https://t.co/KoGkbIWUGx

2023-03-21 07:04:06 RT @academic_la: 32-year-old Or Eshkar died today after sustaining injuries in a terror attack in Tel Aviv 11 days ago. His doctor said, "A…

2023-03-21 06:50:01 @stephensenn @soboleffspaces The hazarad of being contrarian. "Most observational studies have all the problems of RCTs and some in addition." Really? Here are 4 problems RCTs have and OS don't: expense, small sample, needing consent, placebo effect.

2023-03-21 06:32:37 "probabilities of benefit and harm through sensitivity parameters and proxies" by Jose M. Pena https://t.co/YWx8k5Glxa Perhaps because of its relation to our "Personalized Decision Making" paper https://t.co/rGHIc93DYO, which was discussed at great length on this channel. 2/2

2023-03-21 06:32:36 It's Monday again, bringing us another stream of papers in causal inference: https://t.co/75yAUagFdw. As I read through the abstracts, I can spot some that are wrongly headed, and some that I put on my "next to read" list. The one that caught my attention is: "Bounding the.. 1/2

2023-03-21 05:27:29 @stephensenn @soboleffspaces I beg to differ. The fact that they are taken under RCT regimes DOES make them causal

2023-03-21 05:13:05 @AndrewPGrieve @soboleffspaces @stephensenn The fact that the nurse does not know if the tablet in her hand is sugar or treatment does not absolves her from the responsibility to give patient i the tablet prescribed by the randomized assignment, not the one preferred by the patient. This amounts to "intervention".

2023-03-21 05:02:55 @OstlundOllie @AndrewDGarrett @soboleffspaces @stephensenn I assume ICH E10 is basing its restrictions on solid evidence and in-depth understanding of the issues involved. If, however, they are open to hear an opposite view, I'll be happy to share my understanding of the tradeoffs involved.

2023-03-20 23:44:48 @stephensenn @soboleffspaces Of course you need to consent, and of course this makes you different (this is called "selection bias") But, once you consent and your parents consent etc. etc, if you are assigned to treatment a nurse asks you to swallow the drug even if it tastes bitter -- "intervention".

2023-03-20 21:08:24 @stephensenn @soboleffspaces In the non-parametric analysis with which I am familiar, no assumption is made about relative magnitude of effects. In particular, using OS DOES NOT assume that "interactive effects (effect modification) are more important than main effects (confounding)."

2023-03-20 18:04:04 @soboleffspaces @stephensenn Why not call in "intervention"? I have the hunch that people do not fully appreciate the fact that RCT involves two components: (1) Randomization and (2) Intervention. E.g., once you are assigned to treatment arm, you are given the treatment whether you like it or not.

2023-03-20 13:55:29 @stephensenn @soboleffspaces I label P(y_t) &

2023-03-20 13:48:27 @revprez @fallingmrf Well put. @fallingmrf would like us to be generous to these ex-Jews, carrying "Humans Against Breathing" posters on the chests, as if they merely wish to alert us to the dangers of breathing foul air. I label them ex-Jews because breathing is an essential tenets in Judaism.

2023-03-20 12:22:58 @stephensenn @soboleffspaces Of course CC does not prevent the estimation of group means. We state it very clearly: "these two quantities are rarely reported separately

2023-03-20 07:56:44 @fallingmrf Posing in that picture, with those posters on chest is more than just "opinion". It's a virtue-signaling statement saying: There is something morally wrong with Zionism and we, the morally endowed below, have found its cure. All they've found is temporary relief from bullies.

2023-03-20 07:38:56 @mapto History teaches me that endowing someone with the title "identity" while deprecating your identity to the level of "political agenda" is a careless, if not dangerous act of inviting more bullying from those who are just waiting to find Jews lacking coherent identity.

2023-03-20 07:23:56 My inborn modesty would not allow me to say that you made the right choice, that your students are fortunate to start their critical thinking voyage on the right boat, and that my biased opinion is normally biased in the right direction. https://t.co/55wGL35jyE

2023-03-20 07:14:33 And I, personally, will not forgive these ex-Jews in the picture for giving our bullies the impression that we, Zionist Jews, lack spine, historical perspective and intellectual resources to tell them: "Go find your weakling prey elsewhere - we can fight your racism hands down. https://t.co/J659PAx0p4

2023-03-20 02:59:13 @kareem_carr In statistics, a "mental model of the world" is a tacit background assumption. In AI, an explicit mathematical specification of a "mental model" is the starting point of the analysis, which allows us to decide what assumptions are needed for any given task, eg, generalization.

2023-03-20 01:30:27 @kareem_carr I don't want to limit us to causal inference alone. Anything which makes "Data Science" a "science," beyond just "data", requires a new logic which cannot be found in statistics alone. Take the simple example of generalizing from LA to NYC

2023-03-20 01:13:35 @kareem_carr The logic needed is one that looks at a "data analysis procedure" and tells you whether it does what it's supposed to do. For example, whether what you measure in LA can be used in NYC, whether a drug can be good for men, good for women and bad for a person, etc

2023-03-19 21:36:43 RT @AsraNomani: As women are KILLED in Iran, @DNC Rep. @GerryConnolly grins, holding a picture of Khomeini, who unleashed gender apartheid…

2023-03-19 21:03:22 A constitutional opportunity for the bedrock of Israeli society to reclaim the rock. https://t.co/mohK4KN2Jy

2023-03-19 20:55:02 RT @academic_la: When the Knesset passes judicial reform, the Supreme Court will strike it down. A constitutional crisis will follow. 40% b…

2023-03-19 19:26:40 @kareem_carr It's not the automation in itself that matters but the logic of going from "data + assumptions" to "conclusions". The AI requirement of automation motivated the development of the logic, without which Data Science would reduce to pre-2000 statistics.

2023-03-19 18:20:12 @kareem_carr So, is GPT an example of what AI has contributed to Data Science? Or the ONLY example of what AI has contributed to Data Science? In other words, can one conceive of Data Science without AI?

2023-03-19 11:55:52 Painful words that must be said. https://t.co/AMpwQdivoo

2023-03-19 11:44:28 Not so harsh, @MichalSabra

2023-03-19 11:31:22 Strongly disagree. Jews will never be forgiven for allowing their sociopaths to pose as "Jews", deceiving the whole world about who they are, and handing anti-Semites an excuse to say: "We thought you are all as degenerate as those in the picture!" https://t.co/s9hPI5dC01

2023-03-19 11:10:47 Not to be forgotten. https://t.co/o5S83Qjc7j

2023-03-19 11:01:39 @kareem_carr What's the other sense? Is there anything worth learning from AI? Personally, I've learned a lot, and I'm so grateful for what I learned, that I'm anxious to share it with others.

2023-03-19 10:50:00 Important to note: The "concurrent control" restriction applies to RCTs studies in healthcare applications, not to Observational Studies, where we can estimate both P(y_c) and P(y_t) from the data + DAG combination, as in @soboleffspaces work on hip-fractures. https://t.co/vQcapB5yzm

2023-03-19 10:40:37 Where did you find such wondering souls? Last I checked, they all disappeared. https://t.co/uNgPXX5ZKQ

2023-03-19 08:51:03 Congratulations to @ThePearlPost staff for another mark of excellence. https://t.co/GKcByrn6Gk

2023-03-18 19:05:38 @DavidDeutschOxf Important to note that the Book of Esther does not take a religious stance against "The dark logic of moral nihilism" but a rather pragmatic one: "And you and your father's house will perish". Forget "good and evil", we shall all perish if we follow Chomsky and Mearsheimer.

2023-03-18 18:45:33 @NowakPosadzy @BrianCAlbrecht me a similar question: "If I feel more comfortable with equations, any reason to do graphs?" Heckman and Pinto tried, see how far they got without "backdoor" of d-separation: https://t.co/UGqywWLheF. It's like describing how computers work by appealing to electrons and holes 2/2

2023-03-18 18:35:27 @NowakPosadzy @BrianCAlbrecht Hard for me to accept the idea that economists "do equations," hence are mentally incapable of lifting themselves from the equations to see and work with certain qualitative relations (among the equations), other than the equations themselves. Hard, but real

2023-03-18 09:21:58 @gottfriedmath @artistexyz @soboleffspaces Twin networks are good for verifying independencies among counterfactuals (i.e., ignorability conditions). But the tree-step method is better for actually estimating counterfactual expressions. See https://t.co/nDpUzRoEKU

2023-03-18 09:09:54 For the many readers who were waiting for a video of Garry Kasparov delivering the Daniel Pearl Memorial Lecture, it is available now via this link: https://t.co/ACg5ZOTw0M @AsraNomani @ThePearlPost @MissDiagnosis @Ostrov_A @StandWithUs @JGreenblattADL

2023-03-18 08:38:28 @RonKenett "Looking at the data generation set up" - who doesn't?

2023-03-18 08:14:26 RT @bethsphillips: On this day, Golda Meir was sworn in as the first female Prime Minister Of Israel. Her words still resonate today, “Peac…

2023-03-18 08:13:17 RT @Martin_Kramer: “Judicial reform is better thought of as a project of a small minority that, because of the all-too-typical weakness of…

2023-03-18 07:51:40 @RonKenett I am familiar with one very general theory of generalizing empirical findings, written here: https://t.co/E1ap1MNWAh and here: https://t.co/w5fRMmcBkg. Is BVC an extension of data-fusion? a variant thereof? Any other relation?

2023-03-17 23:29:05 @NowakPosadzy @BrianCAlbrecht Why equations. We know that the answer is not in the form of the equations, only in the identity of the argument sets. In other words, the answer is in the graphical abstraction of the eqs. Moreover, we know the answer is in the backdoor criterion - why sweat?

2023-03-17 22:28:13 @NowakPosadzy @BrianCAlbrecht But if only Polish students rebel, what will happen to other econ students? Are they going to swallow the Ceteris Paribus confusion and pretend their professors know what they are saying?

2023-03-17 22:22:32 @RonKenett Trying, but can't parse abstract notions such as "data perspective". It sounds like pre-revolution apple pie.

2023-03-17 22:12:02 @BrianCAlbrecht No. To the best of my knowledge the notion of "ceteris paribus" is as confused today as it was in 2021, see https://t.co/1oOmVyYlQx The confusion between "stays unchanged" and "held constant" and the extent of "all else" should drive every thinking econ student to rebel.

2023-03-17 19:34:30 In the beginning, "pro-Palestinian" was used as a cover for anti-Israel eliminationists to voice their racist ideology in respectable circles. In time, the correlation between "pro-Palestinian" and racism became so obvious that even innocent pro-Palestinians now bear the stigma. https://t.co/RUhnm7W0jz

2023-03-17 19:11:50 RT @EinatWilf: A gorgeous day in Tel Aviv, the first Hebrew city that modern Zionism built. Shabbat Shalom! https://t.co/3p9kys0Mc4

2023-03-17 18:21:34 @JamesGrace111 Economists do rely on causal knowledge, since knowledge consists of assumptions and they do make assumptions. However, their assumptions are cast in such a screwy language (ignorability) that one is rarely sure if they reflect what one knows about the world.

2023-03-17 11:25:49 Belated thanks to @StandWithUs for giving me the honor of addressing their international conference, and sharing with students from all over the world some of my experience in fighting Zionophobia on campus, as summarized here: https://t.co/7jn0po3BDX https://t.co/qDT6sMFGE9

2023-03-17 09:12:39 @DrTomasAragon The intentions of @unbiasedscipod are pure, but they talk as if we are still in some pre-revolution hand-waving century, as if we still do not know how to address confounding. I hope they take you seriously and read #Bookofwhy.

2023-03-17 08:12:32 studies are restricted by concurrent control to assess only ATE = P(y_t)-P(y_c), not the individual probabilities, P(y_t) and P(y_c). This may not stop the infinite debates but it surely opens up new opportunities for personalized decision making. 2/2

2023-03-17 08:12:31 Just in, new breakthrough! Remember the infinite debates concerning "Personalized Decision" and "Concurrent Control"? https://t.co/rGHIc93DYO. Well, new results posted here: https://t.co/EIzi5wSxW4 show how observational data inform P(benefit) and P(harm) when experimental 1/2

2023-03-17 05:10:57 of assumptions with another, more defensible set. The two examples given in https://t.co/B0TKO64IzR demonstrate why this replacement is useful. In general, however, I'm uncomfortable speaking about "triangulation" in the abstract, without being specific about the replacement. 2/2

2023-03-17 05:10:56 "Epistemic diversity" and "triangulating across methodologies" are desirable hopefuls but, thus far, I haven't seen them formalized and demonstrated even in the simple case of fusing RCT with IV data. The best we can do is (see https://t.co/B0TKO64IzR) "Replace one set 1/2 https://t.co/OEoAUgxOnW

2023-03-17 04:22:23 A people with no history would spare no lie, to steel one from its neighbor. https://t.co/nUM7w8005m

2023-03-16 20:22:23 @bentriderdeb Your leadership is still felt at every corner of this school.

2023-03-16 19:34:21 Regarding "going beyond quasi-experiments", the best I can recommend would be my comments on "Pearl before Economists": https://t.co/5ZiBJjlhJn, especially Section 3 and "Conclusions". My rebuttal of Imbens should also be of interest: https://t.co/sZPpL7rckg https://t.co/dWEbLOfGlU

2023-03-16 12:45:13 I believe it is important that we RT such cases, to alert and inform universities like Leeds that their academic reputation has been badly damaged irrespective of whether they admit guilt or not. I trust the story of Danielle Greyman over that of Leeds. https://t.co/Bzd2gEWDGp

2023-03-16 09:27:03 I hope the Jewish community does not accept this empty statement by the Principal as an apology - the words Israel &

2023-03-16 09:04:32 The thing to notice is who is NOT on the list of signatories. (1) Berkeley Dean Chemerinsky, (2) Anyone from the Law School of my University, UCLA. Deeply disappointed, and hoping they will join later. https://t.co/9kBWDVxhZm

2023-03-16 07:31:03 @GadSaad Are you talking @USC ? The University whose President has appointed a (don't laugh) "task force", sorry an "Advisory Committee on Jewish Life at USC," in January 2022? See https://t.co/vXxPyUokOW I wouldn't be surprise if the Chair of that Committee was dancing at your talk.

2023-03-16 07:21:26 A people with no history can't respect others'. https://t.co/BJSGVqsQz8

2023-03-16 07:19:46 RT @Martin_Kramer: TE Lawrence (of Arabia): “Merit is no qualification for freedom. Bulgars, Afghans, and Tahitians have it. Freedom is enj…

2023-03-16 07:18:25 RT @peterhimmelman: Tomorrow morning's Substack is about a fisherman's transcendent journey. It features an evocative underscore you can li…

2023-03-16 07:16:27 A people with no history, must fabricate one. A culture born in lies, cannot not change its habits. https://t.co/pIQcKGciJE

2023-03-16 07:02:39 I'm not surprised to see Zionophobes such as Conell West, Ilan Pappe and Jewish Voice for "peace" signing this letter, but Peter Gabriel is a disappointment and I have been still hoping for Noam Chomsky to come to reason. After all, he is a scientist. https://t.co/fS8s2doW4u

2023-03-16 06:42:09 RT @academic_la: Netanyahu made a major mistake. Rejecting the Herzog compromise before the opposition (and they would have), he comes off…

2023-03-16 05:17:54 RT @KirkDBorne: Causal #AI and Bayesian Networks — download 680-page PDF eBook: https://t.co/JXONbVAyHG See @yudapearl BOOK OF WHY: https…

2023-03-16 05:08:22 RT @adrianachavira: It was an honor to have @yudapearl &

2023-03-16 05:06:32 @sbuhai @eliasbareinboim Are you implying editors of Econ journals are blessed with higher than average curiosity about relevant innovations in neighboring fields?

2023-03-16 04:19:05 Grateful for an exciting family visit to the Daniel Pearl Magnet High School, and for the opportunity to talk with students and teachers about the unique educational projects initiated in this school. From strength to strength! (Yes, the guy in the Bolshevik hat is me). https://t.co/Elt76oWJI5

2023-03-16 03:50:34 @bentriderdeb @ThePearlPost @maggie09032527 In the secrets of my heart I was hoping you @bentriderdeb will appear, suddenly, as you usually do, and join us on this exciting visit to the DPMHS today.

2023-03-16 03:17:17 No way! It's Brandeis EDI officers that should be sent to remedial training. Have they designated an ombudsman for Jewish/Israeli students? https://t.co/VGP061iVoS

2023-03-16 03:10:49 A good start, but will remain mute, ineffectual and actionless until the word "Zionism" is spelled out in all its six letters, and recognized as an integral part of Jewish identity. https://t.co/YPVJC4QtXq

2023-03-16 02:58:59 I love his Purim Hat, if only his words weren't taken seriously by his parishioners. https://t.co/2rsNCjvsqN

2023-03-16 02:49:37 What does @IlhanMN say to my friend Yasmine? https://t.co/FWtmrFXovs

2023-03-16 02:35:54 RT @emilykschrader: After 4 years and 7 months in prison, political activist Sepideh Gholian was released from prison today where she was t…

2023-03-16 02:35:14 RT @realbassemeid: The Islamic Fatwa Council issued a fatwa against Hamas last week criticizing the group's "reign of corruption and terror…

2023-03-16 01:52:25 The great Abba Evan: "All nations determine their policies in the light of self-interest and then explain them in terms of self-sacrificial altruistic morality.” Humble me: "Are people fundamentally different?" https://t.co/KNNz0EV4i6

2023-03-15 12:37:09 Last I checked, it is logically impossible to be a Zionist and hate Palestine. Look it up in your encyclopedia: Zionism stands for two equally legitimate states. I must see TLOU. https://t.co/s75J5yGvYP

2023-03-15 11:59:29 @KosherRedPill Where did they find so many Bundists? The dust-bin of history? Birobidzhan?https://t.co/16lYLqBiYk

2023-03-15 08:29:30 @sbuhai @eliasbareinboim I used "impossible" figuratively, to mean "very very hard". Indeed, it took Paul and Elias's paper 4 years to get published, which, in this fast moving field means "eternity". Most importantly, mission accomplished

2023-03-15 02:59:12 The New York Times’ blatant antisemitism https://t.co/QyBGX9ORKv via @JNS_org

2023-03-15 00:12:39 This week we are blessed with a swelling 6.5 pages of new papers in causal inference https://t.co/PX4rUv6jso Some of the titles are extremely inviting, for example: Causal Deep Learning https://t.co/uRAt6Nw1hH which refines the Ladder of Causation. Enjoy, and share breakthroughs.

2023-03-14 23:42:55 @nbarrowman Quantifying the strength evidence is difficult, similar to the difficulty of quantifying faithfulness, or incidental cancellation. See Section 6.4 https://t.co/7lUwc9SHN0

2023-03-14 20:47:05 @selkirkragdoll @eliasbareinboim Big data and very high dimensional time series do not produce causal relations.

2023-03-14 20:43:18 Virginia seems to be a cesspool of deceit, but if @IlhanMN joins the party, it will spread beyond Virginia. Thanks for fighting the Pagans. https://t.co/VKaT86hwKj

2023-03-14 20:24:57 You were too charitable to @guido_imbens by citing his 2020 "PO and DAGs" without my rebuttal https://t.co/sZPpL7rckg. Some readers might not see where he errs (most will). https://t.co/9eQEi2VTcD

2023-03-14 19:39:36 This is the first "video review" of #Bookofwhy that I see. Unique media that allows you to see when a reviewer is excited and when he hesitates. Very refreshing to see how causation shines in the eyes of a reader who doesn't pretend to already know everything about causation. https://t.co/5YvqOiO5mg

2023-03-14 19:25:47 @selkirkragdoll @eliasbareinboim What weakness you find in what theory?

2023-03-14 13:13:59 Sometimes even self-evident truth needs a trumpet to be heard. When the Flat Earth Society kicked off its last "The Earth is Flat Week", there wasn't even a single dissenting voice. No more! https://t.co/cMm6weKOs9

2023-03-14 12:55:05 RT @TheHistoryOfTh5: The coin with the earliest known representation of the temple menorah is now on display in Jerusalem. It was dated to…

2023-03-14 11:56:58 @academic_la @BondWolfie Very moderate comment indeed, which would only acquire Israelis trust if accompanied by the litmus test: An "Equally Indigenous" pronouncement by the caller.

2023-03-14 08:58:01 Multi-Congratulations to Paul and @eliasbareinboim for accomplishing the impossible - getting a paper on modern causal inference published in The Econometrics Journal

2023-03-14 07:39:08 @elderofziyon Well said. This brave young man would not be as brave facing a Palestinian Authority security man, not to mention Hamas.

2023-03-14 07:12:24 @HuelleAlexander You are right -- a glaring omission. I'll inquire.

2023-03-14 06:57:18 @TBroekel @PHuenermund @RobertHassink @m_fritsch @ArashMV @rodriguez_pose @HuiwenGong @causalinf I do not think the problem is in the idea of consulting history as it is in consulting history correctly, using methods that economists developed to consult other non-experimental data. For a glimpse, see https://t.co/Jf1SkhdcQS,

2023-03-14 06:41:14 When I hear the title " Presidential Task Force to Combat Antisemitism" I immediately recognize a Presidential Instrument to License Inaction. Things my be different of course at GWO, depending on whether students acquire two key weapons: https://t.co/7jn0po3BDX @sabrinasoffer https://t.co/bjdmjdCZ16

2023-03-14 05:47:42 @PHuenermund @TBroekel @RobertHassink @m_fritsch @ArashMV @rodriguez_pose @HuiwenGong This paper is a product of an echo-chamber bubble that hasn't been touched by the causal revolution, not even by the eco-causal revolution of Haavelmo, Marschack and the Cowless commission. It's true though that History is a causation-seeking discipline. #EconTwitter @causalinf

2023-03-14 02:45:19 For readers who asked about foreign publications of #Bookofwhy, here is what our publisher tells us: 1. US Basic Books, 2. UK Penguin Books , 3. SPANISH Pasado y Presente, 4. HOLLAND Maven, 5. FINLAND Terra Cognita, 6. PORTUGAL Temas y Debates, 7. JAPAN Bungei Shunju, (more) 1/2

2023-03-13 23:04:20 At long last! A non-boring, in-sight-ful news item on CNN. https://t.co/IdJYubP9yz

2023-03-13 11:21:14 Seemingly an anecdotal piece of data, to be added to my list of arguments that Jews are a people bonded by collective memories, more than by any other force. https://t.co/2q8VYuROPb

2023-03-13 10:46:17 RT @Martin_Kramer: “IDF reservists and the technology sector, widely regarded as among the best and brightest Israeli society has to offer,…

2023-03-13 10:40:41 RT @robsatloff: Thank you!

2023-03-13 10:23:16 RT @hahussain: Palestinians are not a race. They are a politically constructed identity. Their ethnicity is Semitic Arab, like yours truly,…

2023-03-13 05:21:53 @JeremyBenAmi I see no reason to apologize. Jimmy Carter was the first to label Israel "apartheid", the first to give terror an ideological legitimacy (see https://t.co/K9NMZGhJSC) and the only US President to be remembered as a Lover of Hamas. [reposted: https://t.co/hVRj5T50KJ]

2023-03-13 03:03:30 @viraldogenes Indeed, some of these people are protesting, not the judicial reform itself, but the disrespectful way it is bulldozed over the country and, more so, the grand sell-out to anti-liberal extortion groups who hope to benefit from their support of the reform.

2023-03-13 02:50:39 We should be overjoyed that protesters everywhere have made the Israeli flag the symbol of their disapproval with the government. We should reject protesting hypocrites like @jvplive and @NewIsraelFund for whom the Israeli flag has always been a symbol of ultimate evil. https://t.co/LDM7yeFgtA

2023-03-13 01:06:19 @lal_shiban Why? What does he have against Mexico existence?

2023-03-13 01:04:27 @tdietterich Deliberation overriding intuition is exemplified nicely in game playing, the former emerges from search and the former from static evaluation function.

2023-03-13 00:20:15 I would add three more -- Playful manipulations -- The laws of physics -- Inborn templates in which to store causal knowledge once it is acquired https://t.co/yqoOtOfZy3

2023-03-12 23:53:51 RT @GeraldNGOM: @OptimisticCon @washingtonpost @elderofziyon @EvanHill did pseudo-forensic CSI-style propaganda on IDF counter-terror opera…

2023-03-12 23:52:08 RT @HenMazzig: Heartbreaking story in the Albany Times Union today from two sexual violence survivors

2023-03-12 23:47:48 A record half a million Israelis protest judicial reform, announce 'Day of Escalating Resistance' | World Israel News https://t.co/9SZZPbDcsj

2023-03-12 23:14:49 Last I heard, America has not denied its neighbors', Canada and Mexico, right to exist. Palestinians do, Putin does, and so do many gullible readers who side with the latter two

2023-03-12 18:37:16 @EricJGreenberg1 These British-minted coins were labeled "beautiful old Palestinian coins" for external deception. I don't believe any Palestinian has nostalgic memories of the 1936-1942 riots, when thousands were brutally killed by British troops. "Government" was not even in one's dream.

2023-03-12 08:01:45 @EricJGreenberg1 These coins were minted by the British Mandate Authority, the colonial power that ruled Palestine after WWI, and didn't know how to reconcile its international commitment to establish a Jewish Homeland with the Arabs violent commitment to fight any such homeland at all costs.

2023-03-12 01:03:31 Lapid proposes constitution for Israel to extract itself from ‘terrible crisis’ https://t.co/T1JDQV4yGL via @timesofisrael

2023-03-12 00:56:54 @partygrrrl65 @baroquenearth @maradonasattva "Separate us from them" is what we, Americans, hear from our Canadian neighbors, and they don't expect us to label them "apartheid", for a good reason: separation helps emphasize our uniqueness and rejoice our diversity and mutual respect.

2023-03-12 00:41:33 I hope you get a chance to convey my humble message to them: "The way you treat Israel is the litmus test of your interfaith objectives." https://t.co/DmTC6P8uYU

2023-03-11 21:36:32 Some readers were "shocked" at my attempt to explain why 1930-40 coins minted under British colonial rule are used by Palestinians to claim exclusive ownership of the holy land. I am retweeting my answer with the hope recruiting those readers to the "equally indigenous" camp. https://t.co/zfLoUtkivW

2023-03-11 20:49:16 @maradonasattva Supremacists are those who deny their neighbor's right to existence, not those who uphold this formula for peace: "Two states for two people, equally legitimate and equally indigenous" (Pearl, 2018) https://t.co/Px86GK01Jz

2023-03-11 12:25:29 @aayhaq When people are trying so desperately to fake history, see https://t.co/GHnFjFHaIB, the question arises as to why they do it. What's your explanation?

2023-03-11 11:46:19 @AlexanderNaume2 Sorry, I have a prior obligation to my students, and I am familiar with their ideas.

2023-03-11 11:42:49 @Dam_Nuwen Except those who fake it.

2023-03-11 11:39:47 @mangefort When people are trying so desperately to fake history, see https://t.co/GHnFjFHaIB, the question arises as to why they do it. What's your explanation?

2023-03-11 10:40:21 @aayhaq A people secured in its history need not dig 1936 coins to prove their neighbor's non existence.

2023-03-11 10:19:33 A beautiful example of cause-effect reversal for the next edition of #Bookofwhy. One can pick up dozens such examples by searching for the phrase "it should be clear" or "everyone knows" in the writings of @MiddleEastInst analysts. https://t.co/WLUPSqGJQt

2023-03-11 09:29:22 A people with no history, in frantic search for non-existent roots. https://t.co/d8aRdnTxi2

2023-03-11 08:18:09 We are informed that the submission deadline to the Nineteenth Conference on Theoretical Aspects of Rationality and Knowledge (TARK 2023) has been extended. Website:https://t.co/hw5clvAfRl

2023-03-11 06:33:13 Yes, sure: https://t.co/4k3wv7q2HG https://t.co/i2EqjIgc2u

2023-03-11 03:53:43 RT @AvivaKlompas: Two Israelis - Rotem Mansano and Or Ashkar - were shot and critically injured in yesterday’s Palestinian terror attack.…

2023-03-11 03:51:52 RT @elderofziyon: The last time that @Amnesty called on Lebanon to stop discriminating against Palestinian residents was 16 YEARS AGO, in 2…

2023-03-10 20:42:46 RT @soboleffspaces: Check out this report on estimating the bounds of #CAUSAL mortality reduction resulting from early surgery using RWD no…

2023-03-10 20:41:14 The Book of Why is translated into 14 languages. The last one, Hebrew, will be published Sept. 2023. Looking forward to it. https://t.co/hBUYgWXOUn

2023-03-10 19:38:29 @mapto @JosepBorrellF @ProSyn No, I am talking about a much deeper rejectionism, one that negates the very notions of "borders" and "negotiation", one that denies the existence of a sovereign entity with which to have "borders" and with which to "negotiate". Settlements are but a pebble on this rejection wall

2023-03-10 11:59:43 @mapto @JosepBorrellF @ProSyn Yes. But the word "rejects" is misleading in its softness. We are facing a side that fights a two-state solution tooth and nail, to the last person, at all costs.

2023-03-10 11:55:02 @PeterS80148175 If I didn't believe this is possible, I wouldn't be working in AI. I would probably join some deep-learning group and do curve-fitting for living.

2023-03-10 00:26:55 They don't celebrate murder! Far from it. They celebrate seeing stupid Europeans like @DTzantchev speaking about "cycle of violence" again, knighting them with legitimacy of reciprocity. · 3h https://t.co/P2f6cVzjNX

2023-03-09 23:47:49 "I'm proud of being programmed at Deep Mind" said one Robot to the other. "Why?" said the other. "Because chances are they seeded into me some neat programming tricks I saw in AlphaGo

2023-03-09 21:52:38 This country is splitting too. https://t.co/0bO8RrO1OQ

2023-03-09 21:41:17 @Inframethod @EnglishOER Blessed are the hunches, on their march to wisdom.

2023-03-09 21:30:15 And most Europeans fall for it. https://t.co/bs5RHcjpxQ

2023-03-09 21:23:01 Still recovering from an inspiring evening with Chess grandmaster and political activist Garry Kasparov : https://t.co/pnDPjUcMEi

2023-03-09 20:54:33 @Inframethod @EnglishOER The aim of causal inference is to liberate that line from the tyranny of hunches and make it a beneficiary of mathematical analysis.

2023-03-09 20:48:34 RT @GeraldNGOM: Palestinian terror attack (shooting) in center of Tel Aviv, exploiting internal Israeli political conflict. Another warning…

2023-03-09 20:35:15 @JosepBorrellF @ProSyn The paradigm shift must start with EU realizing that one side would fight tooth and nail against your stated goal: to see a safe, secure, globally recognized state of Israel live in peace alongside ...freedom, prosperity, and peaceful relations with their neighbors. Tooth+nail.

2023-03-09 20:14:26 @Inframethod @EnglishOER The key is not how overwhelmingly likely we make something happen but that "we make" it happen, rather than let it happen on its own. See "probability raising" trap in philosophy, #Bookofwhy pp. 47-51.

2023-03-09 12:07:47 @Ostrov_A Better.

2023-03-09 09:31:46 Sad news. Chaim Topol, star of "Fiddler on the Roof" is dead at 87 https://t.co/HZSetAQhjO On a personal note, Topol served in the IDF with me (basic training, 1953). He was able to make the whole platoon laugh to the ground, including the toughest commanders.

2023-03-09 08:51:56 Please see (or re-write) the History section of "Confounding" on Wikipedia: https://t.co/hDSR4u0F4i https://t.co/oODaNTDE66

2023-03-09 08:47:33 @stephensenn Please see (or re-write) the History section of "Confounding" on Wikipedia: https://t.co/hDSR4u0F4i

2023-03-09 08:37:50 @stephensenn Nice opening. But I lose you here: "His approach starts from the recognition that it's impossible to eliminate the effect of such nuisance factors in practice". I don't see why these are "nuisance factors" if we haven't defined the research question: "causal effect". What is it?

2023-03-09 08:19:39 Asked if I would recommend this paper https://t.co/MTA5PgUVZh Of course! It's one of my clearest! BTW, the statisticians who insisted that "confounding is solidly founded in standard, frequentist statistics" talked about "effect-biased" type of confounding, not Fisher 1926 type. https://t.co/jVdkzi7jhw

2023-03-09 08:05:57 @stephensenn What is your interpretation of the confusion that Fisher labeled "confounding" in 1926. ?

2023-03-09 07:17:19 @Corey_Yanofsky the role of SCM in this tectonic shift, one need only go to Imbens &

2023-03-09 07:04:34 @Corey_Yanofsky Agree. And that was in 1974, while mathematical statistics has been in existence since 1713 (Bernoulli). So, there has been some tectonic paradigmatic shift 50 years ago, before which no one could say "causal effect", nor express what "confounding" is. Now, to appreciate 1/2

2023-03-09 04:49:26 @stephensenn It wouldn't be surprising if statisticians, lacking a language to define "causal effect", used the word "confounding" to name just ANY confusion, not necessarily the disparity between "causal effects" and association-based predictions. Today, "confounding" stands for the latter.

2023-03-09 04:40:54 @Corey_Yanofsky @analisereal @AdanZBecerra1 @camjpatrick @RWJE_BA @beenwrekt The way to get convinced is to ask: When was it the first time that the concept of "causal effect" was formally defined in the framework of "mathematical statistics broadly construed". BTW: https://t.co/oxFe9eYWnO

2023-03-09 04:33:34 It's easier to convince the Emperor of China than Mathematical Statisticians that a Collider is not just "Jargon of Structural Causal Models" but a piece of reality, deserving a name. Greeting to the Emperor. @analisereal https://t.co/oD6KBpoodl

2023-03-09 01:16:59 @QRDL @tdietterich @sinanaral @GaryMarcus @nytimes Beg to differ. Causal reasoning is precisely what it advertised to be.

2023-03-09 00:56:47 @omaclaren @analisereal @learnfromerror Even if I agree that your framework unifies SCMs and other approaches to causal inference, I would call it "Oliver's framework," not "Statistics provides framework". Because statistics had no framework at all till Rubin, and no operational framework till SCM. History matters.

2023-03-08 10:56:02 The historical account here is as complete and accurate as the act of discarding the works of Wright, Duncan, Haavelmo, as "clutter", not to mention Structural Causal Models, which give Potential Outcomes their scientific standing.

2023-03-08 10:56:01 This is Rubin's famous "clutter paper": "I find much of what is written about causal inference to be mathematically inapposite in one of these senses because the descriptions either include irrelevant clutter or omit conditions required for the correctness of the assertions" 1/2 https://t.co/UbsWW2yOBv

2023-03-08 08:52:44 Delighted to share another stream of CI articles published this week: https://t.co/no5Aaiy7SC. I choose to highlight "Combining Graphical and Algebraic Approaches for Parameter Identification in Latent Variable Structural Equation Models" by Ankan, Wortel, Bollen and Textor: 1/2

2023-03-08 04:54:10 A powerful video by Arnold @Schwarzenegger on hatred and haters. Readers write that haters like Kanye West and Louis Farrakhan should listen. True, but it's more important that @HRW-clocked haters like @KenRoth and @FranceskaAlbs listen, who consider themselves angelic. https://t.co/0h0VdhmCbO

2023-03-08 02:33:06 @omaclaren To elaborate, statisticians who remained outside Rubin's camp declared causality a "subject-matter" issue, lying outside statistics. Those who joined Rubin's camp had to define confounding using ignorability assumptions, i.e., beyond comprehension (See Imbens &

2023-03-08 02:19:43 @omaclaren Of course statisticians have been concerned with more than just prediction and association, But until Rubin 1974 they were using a language built strictly on prediction and associations. This disparity is what drove them crazy up to Rubin, and what drives them crazy today. 1/2

2023-03-07 23:45:08 Illuminating discussion, not to be missed. My takes: (1) Expressional power matters. (2) Explainability, transparency and having world-models are more important than we thought. https://t.co/zVb0e8IXoH

2023-03-07 23:35:10 And in case you hesitate, and we all do sometimes, let's remember what Mordechai told his niece, Queen Esther: "And who knows, perhaps you have attained to royal position for just such a moment.” https://t.co/90OrLEfE1z https://t.co/p4VH07Xtqa

2023-03-07 23:13:29 RT @academic_la: The Israeli military raided Jenin today, and killed 6 terrorists. Amongst them was this sack of crap, Abdel Fattah Hussein…

2023-03-07 22:23:22 RT @AvivaKlompas: 6 Palestinians were killed today, including the terrorist who murdered two Israeli brothers, following an IDF operation i…

2023-03-07 22:22:01 RT @EinatWilf: To the Anti-Zionists who seem to think that essays by Bloomberg or @MattiFriedman mean they have joined your camp, a clarifi…

2023-03-07 22:16:22 @ewenharrison @omaclaren No, a causal interpretation is all you need. My point was that the discussion we are having on "confounding" is not a "turf war" but an attempt to clarify notations and vocabularies.

2023-03-07 22:13:33 @ewenharrison @omaclaren No, a causal interpretation is all you need.

2023-03-07 22:11:27 @omaclaren @learnfromerror No need to link this again, just an answer: in what way your definition of statistics is not 'everything' (it already embraces the entire Ladder of Knowledge) and whether you don't find there elements that most 20th century statisticians tried to exclude from "statistics."

2023-03-07 21:32:48 @omaclaren @learnfromerror My characterization may differ from yours, but it is at least crisp, well defined and same as Fisher's ("summarization of data") and most statisticians of 20th century. You are now claiming it's too narrow, fine, let's hear your definition, and in what way it's not 'everything'.

2023-03-07 21:25:20 @omaclaren @learnfromerror In the transition between my definition of statistics (i.e., properties of the distribution that governs the data) and yours, there must be at least one primitive relation that is part of the latter, but not the former. Can you identify such relation?

2023-03-07 21:13:53 @omaclaren @learnfromerror I would be equally curious to find out how @learnfromerror defined "confounding" last time a student asked. Did she send the student to read the "accounts developed along these lines"? Or used "prediction" and "association" as the basic primitives in the definition?

2023-03-07 20:58:23 @omaclaren @learnfromerror No contention! If statistics is defined as a language for expressing every scientific thought, past and future, real and imaginary, then obviously it can define "confounding". I'm still curious though, how did you define "confounding" last chance a student asked.

2023-03-07 20:43:36 @ewenharrison @omaclaren This is not a "turf war", it is an attempt to define languages, assess their limits and enrich them when necessary. The importance of this enrichment surfaces when a stat professor tries to define "confounding", and can't find the vocabulary needed to do so. Can you try?

2023-03-07 20:20:22 define "confounding" in your stat class (rephrased from https://t.co/XFZYzzUIFy). The conversation immediately turns realistic

2023-03-07 20:20:21 Statisticians would go to great acrobatics rather than admit neglect. When an academic colleague asks me “Aren’t you being a little unfair to our discipline, considering the work of so and so?”, I answer “Must we speculate on what ‘so and so’ did? Can we discuss how YOU 1/2 https://t.co/NLOlw9q3el

2023-03-07 19:56:47 Free translation: The key struggle in Israel focuses on the question whether the ultra-orthodox and the messianic sects will be allowed to shape the entire State in their image. These sects undermine the ideological foundation of Israel -- secular Zionism. https://t.co/U56Xs14M40

2023-03-07 19:31:20 @omaclaren Let's hear your definition of "the language of statistics", so as not to "mis-define" it the way some crooks do, for the own selfish purposes.

2023-03-07 16:32:56 @adrianachavira @bentriderdeb @ThePearlPost I was telling @bentriderdeb that we missed the students reporters from @ThePearlPost who normally meet with the speaker before the Lecture. Let's plan for next year.

2023-03-07 16:25:02 Thank you @bentriderdeb for posting these photos from last night's Lecture of Garry Kasparov. It surpassed all expectations, and we were delighted to see you there, with so many supporters of freedom and democracy. I hope they post a video later on. https://t.co/SjVbTHoEt2

2023-03-07 16:18:06 You are right! #Bookofwhy is a glaring omission from this pile of mighty good books. Unbiased observer. https://t.co/rmZ2O0TVPA

2023-03-07 16:05:36 @omaclaren Doubly curious to see confounding defined in the language of "prediction" or association. Take 50 data distributions of 10,000 observables and non-observables, but show me how confounding is defined in the language of "prediction" or association.

2023-03-07 15:56:47 @leonpalafox @GaryMarcus "Making stuff up" -- This has been the pride of Al Jazeera and affiliated news agencies, from which we could protect ourselves. No more. It's now in our bedrooms, guised as "ChatGPT Lectures".

2023-03-07 15:32:59 @Martin_Kramer This is close to what Sheik Yusuf Qaradawi told his Al Jazeera listeners: "Take from the West all its science and technology, but you must reject Western philosophy, because it is corrupt at its roots by the pagan philosophies of Greece and Rome"

2023-03-07 06:52:22 @omaclaren "Statistics" in some super-broad sense of the word may be compatible with everything under the sun. Still, i'd be curious to see confounding defined in the language of "prediction" or association. See why it's impossible: https://t.co/7lUwc9SHN0 (section 6.2)

2023-03-06 17:43:23 @noah_greifer @JohanDH2O @stephensenn @AdanZBecerra1 Recall: "Equally biased" is true even under unmeasured confounding. See https://t.co/BYnjoEKwmL

2023-03-06 03:41:38 @daoudkuttab Marwan Muasher's understanding of Israeli intentions is pathetic. Israel's withdrawal from the West Bank depends not on Palestinian demography but on Palestinians mentality

2023-03-06 01:20:21 Happy Purim, sister! We didn't have an army when I was your age, my costume was a donkey-head. You made a better choice, treasure! Purim Sameach! https://t.co/K1amUWXAUx

2023-03-05 14:30:23 @HuelleAlexander They are trying, but the spoiled youth are on my side, so they can't succeed.

2023-03-05 12:18:39 Asking @PaleNews not to lie is like asking an eagle not to fly, or a fish to stay dry. https://t.co/VbTFknCEBf

2023-03-05 11:52:22 My brothers and sisters are there, demonstrating for democracy, but what they are really demonstrating for is decency, a scare commodity in the current coalition government. https://t.co/Jj477UmX6s

2023-03-05 11:32:59 Divisiveness is not a crime, and producing antisemitism is something they would surely deny. They should be shut down for spreading Zionophobia, a form of racism they can't possibly deny. https://t.co/8xiF7FWpne

2023-03-05 11:17:12 @academic_la @FaithQuintero7 @rookringer @RenzoGraciani It is also dangerous to frame the Palestinians as a people "with ambitions we need to respect" without mentioning their principal, top-top ambition which we cannot possibly respect. Such framing encourages them to use the former as a cover to advance the latter.

2023-03-05 10:00:00 CAFIAC FIX

2023-03-02 22:00:00 CAFIAC FIX

2023-02-28 05:35:52 May your memory forever be a blessing, Elan. I cry for you, and I feel for your family. https://t.co/GZ5xdBVKBS

2023-02-27 23:27:41 Bipartisan Congressional Letter Calls for UN to Disband Commission of Inquiry, Fire Palestine Rapporteur https://t.co/WwXjJ0BGfE

2023-02-27 21:44:32 RT @Sabine_Messner: Anyone who: - demands "peace talks" with Putin - thinks we’re doing „charity“ to Listen to this discussion! @Kaspa…

2023-02-27 10:56:12 @gypsybeepie @AsraNomani @CAIRNational An enormous waist for offices that only exacerbate hateful atmosphere on campus. Why? Here is an article to read: https://t.co/3CwwpkCk3Y

2023-02-27 10:24:14 Goodness highness! Nature has discovered causality https://t.co/WmiX2Rk2kL We should soon expect a flood of machine learning folks asking: What is it? BTW, I didn't tell them, but the book I'm holding in the picture is the first printing (1763) of Bayes paper on Bayes Rule.

2023-02-27 08:28:32 RT @HillelNeuer: No comment.

2023-02-27 08:11:11 And I can't resist the temptation of endorsing your temptation. These are good papers. https://t.co/YQWRtXAJbG

2023-02-27 02:43:09 Wow, now they (CUNY Administration) are attacking those who request to find out what their "highly qualified chief diversity officers" thinks about Zionist professors at CUNY. This is called intellectual humility. https://t.co/kJsqvkBppF

2023-02-27 02:32:56 RT @jewinthecity: Two brothers, ages 22 and 20, were murdered today, hours after National Day of Jew hate ended. They were shot in their ca…

2023-02-27 02:31:05 RT @StandWithUs: The victims of today's deadly terror attack in Hawara have been identified as #Israeli brothers Hillel and Yagel Yaniv. Th…

2023-02-27 02:30:05 RT @SAFECUNY: Has anybody noticed that CUNY's HEAD of Diversity, overseeing 25 campuses at the largest urban university in the U.S. is NOT…

2023-02-27 02:20:53 RT @Franckalbert201: Fantastic news! Papua New Guinea to open embassy in Jerusalem, Israel FM Eli Cohen confirms! https://t.co/LMRMgTFTVJ

2023-02-27 02:17:31 RT @AJCGlobal: There is no room for vigilantism in Israeli society. We condemn this violence in the strongest terms.

2023-02-27 02:16:45 Condemning the inexcusable! https://t.co/nMbWiFozBI

2023-02-27 01:00:00 CAFIAC FIX

2023-02-20 11:49:06 @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ What's wrong with it? It saves some people and kills others, the former and the latter are each 10% of the population and, obviously, non-intersecting. Do we really need to say it at such length? Let's get to the meet of things and ask: Can we devise a test for the two models?

2023-02-20 10:04:49 @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ If you are referring to the example in our paper, then my conclusion is somewhat different: The FDA should license the drug for all females and lounch a study to explore the existence of features E and F that produce benefit in some males and harm in others.

2023-02-20 07:56:27 @stephensenn @HL327 @f2harrell @AngeloDalli @Raamana_ None of the above. It's an innocent question for readers with intellectual curiosity: Devise a test to determine which model is behind our data. Model-1 – The drug has no effect whatsoever on any individual Model-2 – The drug saves 10% of the population and kills another 10%.

2023-02-20 07:31:12 @sim0ngates @f2harrell @HL327 @stephensenn @AngeloDalli @Raamana_ It is a wider comment, suggesting that RCTs as well as all other available sources of data (eg, Observational Studies) be leveraged coherently to answer the research question at hand. By "coherently" I mean using the logic of causal inference.

2023-02-19 20:07:20 @heckerma1 @PSpirtes I got a broken link, can you check?

2023-02-19 07:24:07 @aydinonat @paolosilvest @Catemarchionni I still could not find direct access.

2023-02-19 07:20:56 @voidmstr I'll figure it out tomorrow. Sorry

2023-02-19 07:15:11 Happy to share the final version of my paper: "Comments on Nick Huntington–Klein's review ‘Pearl before economists: The Book of Why and empirical economics’" : https://t.co/hurMBmYgak https://t.co/6rCKHjpNED

2023-02-19 04:57:49 @soboleffspaces @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ Sure. But as soon as you put down ATE as a difference ATE = P(Y_1 = 1)-P(Y_0=1), some RCT folks will say: We don't care about ATE

2023-02-19 03:34:09 @soboleffspaces @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ To make the contrast even more general I would write: RCT delivers [P(Y_1=1)

2023-02-19 00:57:27 @HL327 @smueller No matter how sophisticated and clever the RCT, adding an Obs. Study may always add something about the individual harm, depending on the bounds produced.

2023-02-19 00:19:47 @DavidHirsh @EliSennesh There is no question that there are antisemitic roots to Zionophobia. But these roots can be explored once we clarify the dangerous consequences of Zionophobia and its moral deformity. Note how successful Muslims were in turning Islamophobia into a moral taboo in the West.

2023-02-18 23:53:17 @DavidHirsh We can say and say, and we have been saying and saying

2023-02-18 23:36:20 @aydinonat @paolosilvest @Catemarchionni Elon Musk should outlaw all journals that announce online availability of new issues, without providing linked access to the content of those issues.

2023-02-18 22:26:24 of these perpetrators for spreading the stench of Zionophobia - a genocidal moral deformity that calls for the destruction of a people's homeland and for stripping eight million refugees of their history, safety, creativity and dignity. @noatishby @JGreenblattADL @EinatWilf 2/2

2023-02-18 22:26:23 This campaign, to fire Zionists from their jobs, is a result of our failed strategy of naming the perpetrators "antisemites" and quit. It's time to move the fight from the hopeless legal arena to the moral arena, where we can win hands down. It's time to call for the firing 1/2 https://t.co/VEc2wZroJr

2023-02-18 21:19:17 @DoctorNazarian Are you trying to give antisemitism a bad name? They might sue you for liable.

2023-02-18 20:58:40 @HL327 All 'harms", whether expressed counterfactually or in English, are due to "separate causal mechanisms". So, I see no reason to speak about "two types of 'harm'". The underlying mechanisms could be very complex, but always manifest themselves as P(y'_t, y_c) >

2023-02-18 13:04:40 @f2harrell @HL327 @AngeloDalli @stephensenn @Raamana_ If no confounding exists, and we do not suspect any, we do not need to run the RCT

2023-02-18 13:01:05 @HL327 Model-1 from Model-2 even under Odds Ratios or any other comparison of P(Y_t) and P(Y_c). Our paper demonstrates that the pair [P(Y_t),P(Y_c)] can be the same in men and women, yet women face P(harm)=0 and men P(harm)=21%. Odds Ratios can't save RCT from its blindness. 2/2

2023-02-18 12:30:18 @HL327 The importance of estimating P(individual harm) &

2023-02-18 00:36:43 Good, now that we are all exhausted, and everyone sees the importance of estimating P(individual harm), here is a revised version of our paper: https://t.co/rGHIc93DYO to be published in the Journal of Causal Inference. Enjoy, and try to assess your P(individual benefit). https://t.co/EMCexO3DNI

2023-02-18 00:03:24 @HL327 @AngeloDalli @stephensenn @f2harrell @Raamana_ Recall that, if we were not worried about such "hidden covariant that acts as an effect modifier" we wouldn't run RCT in the first place. If you believe that "Obs Study added info abt the RCT," please convince @stephensenn &

2023-02-17 23:37:15 @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ I couldn't relate to your analysis, b/c it was too wedded to the idea of varying dose, which takes us away from the main point: P(harm) vs. ATE I love your concluding sentence: "Such observational studies are of course well established practice." Please send one to @stephensenn

2023-02-17 23:22:28 Gee, his mentality is closer to Kanye West than to Rashida Tlaib -- the pendulum is swinging back.

2023-02-17 23:22:27 The suspect, Jaime Tran, has admitted to the shootings, adding that he searched Yelp for a kosher supermarket and was homeless for the past 12-14 months. https://t.co/XV2mlYB7s8

2023-02-17 22:52:38 @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ Miscommunication again. 10% of 'IndivHarm' &

2023-02-17 22:40:59 @stephensenn @AngeloDalli @RWJE_BA @f2harrell @HL327 @Raamana_ There is NO ASSUMPTION whatsoever in the M&

2023-02-17 22:24:34 @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ Everything hinges on understanding that, for both a patient and a policy maker, taking a sugar tablet is NOT the same as taking a drug that saves 10% and kills 10%. I believe even FDA will agree on the importance of the difference, and our paper convincingly explains why.

2023-02-17 21:22:35 @f2harrell @HL327 @stephensenn @AngeloDalli @Raamana_ Two reasons why I prefer Twitter: (1) It forces you to convey ideas succinctly and precisely, rather than bury yourself in details. (2) I have an obligation to 65k (silent) followers who are taking this educational channel seriously.

2023-02-17 15:52:28 @jrosell @stephensenn @HL327 @f2harrell @AngeloDalli @Raamana_ I don't see why. I can run very replicable study on rats, but if I would try to generalize the conclusions to humans, many would scold me for using "bad evidence".

2023-02-17 14:33:01 @f2harrell @stephensenn @HL327 @AngeloDalli @Raamana_ This is one of the options we can take: Totally discard epidemiology as invalid, not to mention economics, political science and more, every science that has not purged OS from its midst, and legitimize only RCTs, though they are excused from external validation by definition.

2023-02-17 14:19:15 A culture seeded in deceit must remain deceitful to preserve its identity. https://t.co/Tuq3JE25i4

2023-02-17 11:06:10 @stephensenn @soboleffspaces @f2harrell @PavlosMsaouel @jon_mellon @Raamana_ @smueller Do you accept the hundreds of Observational Studies in epidemiology as an example of " the sort of information M&

2023-02-17 10:18:02 @stephensenn @HL327 @f2harrell @AngeloDalli @Raamana_ I have just given you hundreds real life, real data studies in which you can estimate P(ind.Benefit). You can either dismiss the whole field of epi as BS, because they are not using RCTs, or you should accept those studies as a legitimate real-life example. Your choice.

2023-02-17 10:07:44 @stephensenn @HL327 @f2harrell @AngeloDalli @Raamana_ I do not have a list of studies on diseases and t reatments and populations. But take any OS in epidemiology. It contains OS data, and a CI model, so it contains OS data and (implied) RCT data. For any such combination we can find bounds on P(ind. benefit). What else is needed?

2023-02-17 10:00:08 @stephensenn @HL327 @f2harrell @AngeloDalli @Raamana_ It is impossible to search for evidence not knowing what you would consider to be an "evidence", when we can't even agree on what "evidence" there is that RCT works.

2023-02-17 09:51:10 RT @SAFECUNY: BREAKING NEWS: Moments ago, CUNY announced a university-wide diversity/discrimination event that will address antisemitism…

2023-02-17 09:14:44 @stephensenn @HL327 @f2harrell @AngeloDalli @Raamana_ Another suggestion: Instead of seeking "problems" in the M&

2023-02-17 09:05:23 @stephensenn @HL327 @f2harrell @AngeloDalli @Raamana_ Forget about C. Why should A be different than B? Finite sample effect? Different populations surveyed? Nothing is a problem if it is physically feasible. The only problematic data combinations are those identified as "infeasible" by the system.

2023-02-17 08:54:52 @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ Here is a challenge to all who march by testability, or imply that CI needs testing: Devise a test to determine which model is behind our data. Model-1 – The drug has no effect whatsoever on any individual Model-2 – The drug saves 10% of the population and kills another 10%.

2023-02-17 08:38:51 @HL327 @stephensenn @f2harrell @AngeloDalli @Raamana_ To find out is something can be tested you need CI, which tells you the testable implications of your assumptions. I'm not clear what test you are proposing for P(ind. Ben). According to RCT experts, RCT can't even test ACE, do you expect it to test P(ind. Benefit)?

2023-02-17 07:43:45 @stephensenn @f2harrell @HL327 @AngeloDalli @Raamana_ Notwithstanding the shelves upon shelves of exp. design papers written by FDA people, I'm unaware of anyone leveraging two new developments: (1) the mathematics of causal inference, its tools &

2023-02-17 07:23:17 @stephensenn @f2harrell @HL327 @AngeloDalli @Raamana_ True! I always tell myself that things on the ground are harder than they look on paper. But I also tell myself that new promises must show up on paper first, before realized. I therefore explore new possibilities on paper, hoping for more than "this is not how RCTs are run"

2023-02-17 05:07:06 paper of ours that does just that. See Section 2 of: https://t.co/LMyQ84zEil

2023-02-17 05:07:05 Having seen how trialists exploit the additivity property of the Odd Ratio (OR) to escape selection bias, readers may be interested in seeing how the symmetry of OR is leveraged in non-parametric CI to remove selection bias. @eliasbareinboim has reminded me of an old 1/2

2023-02-17 04:11:44 RT @StudiesZionist: https://t.co/Im9d0E9ATQ Incisive piece as always from @GilTroy, with a shoutout to me and the JSZN regarding the tatt…

2023-02-17 03:56:15 A must read for all Jewish faculty, students, parents, donors, and community leaders who still believe that Jewish Studies Centers in their universities are not molesting their values as Jews, or betraying their expectations as a community. My UCLA experience needs a story too.

2023-02-17 03:56:14 Numerous critics have focused on the fact that the "art" of the “Gaza Ghetto” tattoos trivializes the Holocaust. https://t.co/grvwW4HNvv

2023-02-17 01:47:52 @edelmann_domi @f2harrell Your post made me go back and read everything I've tweeted the past two weeks -- Sorry, couldn't find any **** around. I've found however some good ideas, that should be taken as seriously as I was when I wrote them. Please try.

2023-02-16 23:45:41 @HL327 @stephensenn @AngeloDalli @f2harrell @Raamana_ You seem to be familiar with both (1) causal inference (CI) and (2) Multiple parallel feedback control loops (MPFCL). If you judge (2) to do the work of (1), there is no need for you to use (1), just improve RCT to the extent permitted by (2). I am limited by what I see in (1).

2023-02-16 23:28:43 @f2harrell @HL327 @stephensenn @AngeloDalli @Raamana_ I heard from reliable sources that FDA's people are scientifically minded and would surely not accept "RCTs are not run that way" as an argument for not exploring how RCTs SHOULD be run.

2023-02-16 22:48:05 @HL327 @stephensenn @AngeloDalli @f2harrell @Raamana_ We are well aware by now of the many reasons that medical RCT are done they way they are. But current practices should not be brought up as evidence that they should continue. We are examining the scientific basis of those practices with an eye towards improvements. It's 2023.

2023-02-16 22:40:31 @stephensenn @soboleffspaces @f2harrell @PavlosMsaouel @jon_mellon @Raamana_ @smueller I don't think you would want your students to quote your last Tweet-- it looks like written by someone who does not understand the difference between "explicating assumptions" and "assumptions must be true", or, between "correcting for effects" and "ignoring effects". Reflect.

2023-02-16 22:10:49 @stephensenn @soboleffspaces @f2harrell @PavlosMsaouel @jon_mellon @Raamana_ @smueller All the king's horses, and all the King's men couldn't change the fact that "so far, the only method that conceptualizes, addresses, treats, defines, or estimates Pr(benefit | z) is PNS by @yudapearl &

2023-02-16 20:42:55 mathematically in the literature called "causal inference" (CI), which derived both: (1) warning signals, and (2) tricks for circumventing violations that led to such signals. Another literature exists, called "trialists," for whom all this is "impractical" or worse

2023-02-16 20:42:54 I'll try, though I am not sure in what terms clinicians speak nowadays. Falsifiability means a warning signal, coming out of the study data, telling us that something went wrong in the study, namely, the assumptions we made about the people in the study, or 1/n https://t.co/8D7LhLrIIB

2023-02-16 09:46:42 @f2harrell @jon_mellon @Raamana_ @stephensenn To see whether population ATE is computable from RCT requires mathematical representations of both ATE and RCT. "Transportability Folks" do have such representations (eg https://t.co/dEPwcv3l40). Q. In deciding "non computable", have you used their representations? Alternatives?

2023-02-16 07:53:15 @stephensenn @f2harrell @jon_mellon @Raamana_ Can I understand this to mean: If you are interested in a representative sample, free of bias-producing interactions, you do not sample at random, but make sure your sample has 50% males and 50% females. Fine. But isn't random sampling a good protection from unknown factors?

2023-02-16 04:03:12 AI critics are warning us of future robots deployed in battle fields. They havn't warned us of organic robots deployed in our restaurants and bus stations. Watch them below: https://t.co/5julT5u30V

2023-02-16 03:53:31 @tdietterich @aperianez @erichorvitz I don't mind calling the marginal "foundational" .But I mind calling them "models" not knowing what in the world they model, if any.

2023-02-16 03:24:16 @f2harrell @jon_mellon @Raamana_ @stephensenn Can I at least choose cancer cells at random? See https://t.co/56iKoTSiTB

2023-02-16 03:18:45 RT @PHuenermund: "The information extracted from these documents is encoded in knowledge graphs—a way to organize data that captures links…

2023-02-16 03:08:40 I once had a student like that but, as far as I know, he hasn't finished yet. https://t.co/knV3UFCd9F

2023-02-16 03:04:48 Great laughs @SarahKSilverman https://t.co/jbdh91mHNh

2023-02-16 02:12:11 @jon_mellon @f2harrell @Raamana_ @stephensenn We have seen two kinds of "theoretically motivated transportability claims". (1) "No interaction" assumptions used by current RCT practitioners. (2) non-parametric assumptions used by transportability folks. The latter claim ATE for a target population. The former claim =???

2023-02-16 01:58:53 @soboleffspaces @JohanDH2O @stephensenn @analisereal @ildiazm @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars If a statistician wants to call her survey decisions (eg who to send a questionnaire to) a "design", I'd honor her choice of words. But I can't honor the word "design" where there are NO decisions to be made, as in Lord's story. It's a bad imposition of RCT jargon on OS.

2023-02-16 01:45:44 @f2harrell @jon_mellon @Raamana_ @stephensenn The fact we can't choose any plot of land on earth does not mean that we can't sample from those I can choose, or that we can't estimate the ATE for exactly those we can choose. @stephensenn explains what "fine RCTs" are doing today, not what RCTs are capable of doing.

2023-02-16 00:53:08 @f2harrell @jon_mellon @Raamana_ @stephensenn Though "no RCT does that" on humans, with our current state of knowledge, @jon_mellon asked whether RCT can do it in principle, say on internet traffic, e-commerse, or say using our current understanding of non-parametric causal inference methods.

2023-02-16 00:42:42 ‘Not Just Israel’: Biden Admin Says Human Rights Candidate Axed for Range of Issues https://t.co/lMLCWZ70Gx

2023-02-15 23:40:06 @jon_mellon @f2harrell @Raamana_ @stephensenn If you were to ask me, I would say: Of course! If an RCT is incapable of estimating the ATE of the population from which samples are selected why would we run RCTs ?? No matter how deep and wide RCTs can be, if they do not deliver ATE of an actionable population, they'r a waste.

2023-02-15 23:28:55 @stephensenn @analisereal @ildiazm @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars 1) two halls do not make two schools. Lord spoke about "a university", namely ONE university. Do you think anyone might interpreted "two Halls" to mean "two schools". I don't get it.

2023-02-15 21:37:02 @f2harrell @jon_mellon @Raamana_ @stephensenn Regardless of the history, my point is that there is no legal or moral impediments today to choose plot at random if we really want to do an RCT.

2023-02-15 17:20:32 @stephensenn @analisereal @ildiazm @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars AHA !! Eureka! What I call "story", you call "design". We are truly approaching a convergence. But if we are analyzing data already taken, with no option of change the way they were taken, can we safely say we have a "story" and not a "design"?

2023-02-15 17:15:36 @stephensenn @analisereal @ildiazm @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars Lord's story had one center, one university and data already taken -- no choices left. Is Nelder approach needed for multi-center RCTs ???

2023-02-15 17:12:33 @stephensenn @analisereal @ildiazm @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars Lord's story begins with data already taken.

2023-02-15 17:10:43 @stephensenn @ildiazm @analisereal @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars Perhaps the assumptions are different, but the aims were stated by Lord: "Determine which statistician is right? " I'm still anxious to see Nelder's assumptions stated without RCT metaphors

2023-02-15 17:05:27 @f2harrell @jon_mellon @Raamana_ @stephensenn Luckily it works for some RCT's - agricultural experiments, marketing research, and more.

2023-02-15 10:24:14 @BenMaimon4 @AizenbergSalo @TimesofIsrael Thanks for posting this article on J. Street, it summarizes Ben Ami's lecture very vividly, as if I were in the audience. https://t.co/xCMcJvo0Dl

2023-02-15 09:35:59 What I don't like about J. Street is their slogan: "Pro-Israel, Pro-Peace." Implying that being just Pro-Israel is not enough

2023-02-15 09:20:54 @stephensenn @analisereal @ildiazm @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars I hope I won't offend you if I make a confession: The word "design" throws me off to confusion whenever I see it in the context of Lord. Here we have two statisticians who make no design decisions whatsoever

2023-02-15 09:08:41 @stephensenn @analisereal @ildiazm @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars I still can't tell what words in BoW give you the impression that the story is different from the one in PGJ? The two look identical to me.

2023-02-15 07:01:48 This Tweet, and the photos of Asher (8) and Yaakov (6) (HY'D) jolted a reader to send me pictures of 12 Palestinian children killed in Gazza, and ask: How about? My response: He who counts bodies and discounts intentions emboldens the aggressor and indicts its victims. https://t.co/HdDZdLwQhj

2023-02-15 02:59:31 @Vivify705 We tried those other clues too, they are called "2-state solution" and "equally indigenous". Unfortunately, when you dont believe in logic, you don't believe in anything but "Me, Me, Me".

2023-02-15 02:21:45 In formal logic this is called "reasoning by cases". There is only one problem with it, which surfaces when your audience doesn't believe in logic. https://t.co/edF47XXvLU

2023-02-15 02:03:36 Today I was yearning for some good news, and it came. I only question how he got nominated with a Zionophobic record like his. Sadly, he taught at UCLA and thus stained my university with his apartheid bigotry. Harvard can afford such embarrassments, not UCLA.

2023-02-15 02:03:35 US Withdraws Nomination of Activist Who Accused Politician of Being ‘Purchased’ by Pro-Israel Groups https://t.co/NJjicb3gS0

2023-02-15 01:14:07 @analisereal @ildiazm @stephensenn @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars Moreover, if we assume that we have only ONE Hall, serving two diets, everything simplifies, no more diversionary issues, and the paradox is still paradoxical. What a relief. It took us 100 tweets to filter out orthogonal diversions

2023-02-14 23:21:08 @stephensenn @ildiazm @analisereal @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars 1)If one hall is further from the nearest bus stop, then obese students would prefer the closer one, so W1-->

2023-02-14 22:40:43 @stephensenn @ildiazm @analisereal @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars I can't recognize Lord's story in any of the "group structured" DAGs. The one on the right has "school" as a variables, but we have only one school. The one on the left shows an arrow from "Hall" to "W1", saying that the Hall affects students' initial weight -- hard to believe.

2023-02-14 21:06:42 @HL327 @f2harrell @StableMarkets @stephensenn @ildiazm @analisereal @RWJE_BA @Raamana_ @vandy_biostat @EdgeforScholars Sure. We are on a common ground even w/o the term "Decision Analysis". But you would need to explicate your understanding of P(harm) using the patient's utility function, as it was done here: https://t.co/MlyGdEjQHD

2023-02-14 21:00:56 @f2harrell @HL327 @StableMarkets @stephensenn @ildiazm @analisereal @RWJE_BA @Raamana_ @vandy_biostat @EdgeforScholars The A in CATE stands for "average", honoring the "Expectation" operator in E(Y(1,u)-Y(0,u)). Note that it does NOT equal E(Y|treat, C)-Y|cont,C), since the latter is a statistical quantity

2023-02-14 20:46:00 @HL327 @f2harrell @StableMarkets @stephensenn @ildiazm @analisereal @RWJE_BA @Raamana_ @vandy_biostat @EdgeforScholars In our paper, we gave two examples showing that two persons, having two different utilities (or beliefs), would prefer different treatments, depending on their P(IH)'s, even though their ATE's are the same. https://t.co/rGHIc93DYO page 3, paragraph 3.

2023-02-14 20:32:41 @HL327 @f2harrell @StableMarkets @stephensenn @ildiazm @analisereal @RWJE_BA @Raamana_ @vandy_biostat @EdgeforScholars Though we are from different disciplines, we both use counterfactuals when it comes to explicating concepts and assumptions. So, how about expressing formally the way you see "benefit and harm"

2023-02-14 19:35:48 @f2harrell @StableMarkets @stephensenn @ildiazm @analisereal @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars Two comments: (1) The estimand you are describing is known as CATE, where C stands for "Conditioned on patient characteristics". (2) CATE hides information about P(benefit) and P(harm) which are "pertinent for individual patient decision making", see https://t.co/rGHIc93DYO

2023-02-14 15:21:13 Every time someone introduces a new twist into Lord's paradox, I go back to my first paper on the topic, and I get a fresh air of clarity, coupled with an urge to share it with readers. Here it is: https://t.co/4LDOsGAur7 Enjoy the beauty and simplicity of the solution. 1/2 https://t.co/5G3TCH37Wm

2023-02-14 14:19:12 @stephensenn @ildiazm @analisereal @f2harrell @RWJE_BA @Raamana_ @HL327 @vandy_biostat @EdgeforScholars And my claim, having looked at all the arguments, is: (1) You do not need DAGs to present arguments, but you need some mathematical object to explicate assumptions. (2) The #Bookofwhy's resolution of Lord's Paradox is complete and beautiful, enjoy: https://t.co/MSJJViHxbM

2023-02-14 09:07:15 @MichiganWater2 @stephensenn @f2harrell Your interesting introduction of "the campus cookie shop," changes the story a bit, and is discussed in Fig. 2 of https://t.co/MSJJViHxbM . But do you consider the paradox resolved in its original version, assuming WI is the "sole confounder" between Diet and Outcome?

2023-02-14 08:38:40 Is there no responsible adult to tell President Abbas that such statements make a Palestinian state 10 times less realizable than 10 new Israeli settlements? Israelis have been hearing the "We Alone" slogan since 1929, and they know where it leads: https://t.co/sdkafXHCh5 https://t.co/Bz1ETV7o1Q

2023-02-14 07:19:39 RT @HOS_ASA: #OTD 1766 Thomas Malthus b (d 29 Dec 1834) FRS Fun fact: Cofounder 1834 w Babbage &

2023-02-14 06:59:05 RT @Israel: These are the incredible women who dropped everything they were doing and flew to #Turkiye to save lives as a part of Operation…

2023-02-14 05:52:30 @MichiganWater2 @stephensenn @f2harrell I do not understand the difficulty. I said: there is no need to confuse (confound) schools with diets or with halls, just think of diets, nothing else. See a clear discussion: https://t.co/MSJJViHxbM If you look at the graph, you see "confounding" created by the initial weight.

2023-02-13 12:03:38 When I tell my American friends that the entire Palestinian narrative can be summed up in just two words, "We Alone", they refuse to believe me. Surely, they say, some Palestinians have heard about their neighbors having had some history...etc. No, they haven't -Nada. "We alone" https://t.co/nQHkfMLEOH

2023-02-13 07:26:28 Responding to @KenRoth renewed attack on Israel, I am changing the title to read: "Three Slanderers and two victims" No cause-effect implied, just moral deformity. https://t.co/fyj17VkYrq… https://t.co/mTwWyQpJQo

2023-02-12 23:03:16 @analisereal @f2harrell @Raamana_ @stephensenn Almost. Except that P is undefined, hence, must be replaced by Freq. Also, the assumption OR(X) = OR(X, S=1) holds (after scaling) in linear models, which explains why selection bias is solved there https://t.co/xXl4OMzmIy

2023-02-12 22:52:12 @RajaChemayel @Ostrov_A @AgnesCallamard @amnesty @KenRoth @hrw @FranceskAlbs @UN_HRC @IsraelinUN @ShaharAzani @benshapiro @NGOmonitor @noatishby @SSI_Movement @COLRICHARDKEMP @The_ILF God forbid! That would make you anti-Palestinian, whose ideological DNA is to grab and possess their neighbors' country.

2023-02-12 22:45:16 @analisereal @f2harrell @Raamana_ @stephensenn Someone asked "the population is finite". I answered: Show me any statistical estimation exercise where the population is not finite. The population of Basel (Bernoulli, Artis Conjectandi, 1710) was also finite, yet by assuming it to be infinite, Bernoulli created statistics.

2023-02-12 20:51:07 Must read. https://t.co/1pONPy2Nyn

2023-02-12 20:45:08 @leonpalafox Let me know if you find out.

2023-02-12 20:36:43 @analisereal @f2harrell @Raamana_ @stephensenn I am not sure they would compromise to accept the average effect on the selected, ATE(S)= E[Y(1)-Y(0)|X, S=1] as a goal of RCT. Why? Because ATE(S) invokes expectation relative to some distribution, the existence of which they consider "fictitious".

2023-02-12 15:42:47 My title for this photo: "Slander and its victims". https://t.co/HdDZdLwQhj

2023-02-12 15:33:50 @Raamana_ @f2harrell God forbid! I am arguing against the current practice of some RCT experts, who refuse to admit that RCT has a purpose: To estimate the ATE in the clinical population.

2023-02-12 12:15:58 @HL327 @f2harrell @stephensenn @elmir1omerovic @OstlundOllie @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone I dont see any contradiction. The OS alone makes me infer NOTHING. I turned to a RCT to estimate something (ATE) and still found no contradiction, b/c OS does not compete with RCT on ATE, it merely adds to it information about individual behavior, which ATE lacks.

2023-02-12 12:04:27 @LGHemkens @artistexyz @soboleffspaces @smueller Ideal RCT's do not need DAGs. DAGs are needed to mend RCT's imperfections.

2023-02-12 12:00:42 @f2harrell @technoslerphile @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Extending follow-up amounts to adding an outcome variable, dependent on both treatment and RCT outcome. Don't we need modeling assumptions to say anything of added value?

2023-02-12 11:53:25 Dedicated to my friends and colleagues in Barcelona and Madrid. Thanks for posting. https://t.co/N7ZdGXyH3g

2023-02-12 11:47:52 @f2harrell @technoslerphile @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Y(1,u) differs from Y(1,X) in that u contains ALL characteristics of an individual u, measured and unmeasured, known and unknown, sufficient to make the response Y a deterministic function of treatment. I don't think avoiding counterfactuals is an option in matching RCT with ATE.

2023-02-12 10:05:19 @gottfriedmath @LGHemkens @artistexyz @soboleffspaces @smueller You are writing "[RCT] informs on ... average causal effect (ACE)". This is what we learned in school, and have hoped they are doing. Not so. What I've discovered the past 2 weeks is that the notion of ACE is non existent in the hard-core RCT literature

2023-02-12 09:30:45 @f2harrell's paper https://t.co/RZojYMBRGT, explains why RCT are not falsifiable, "Randomized clinical trials are successful because they do not mimic clinical practice", namely, because practitioners refuse to bow to any external criterion of performance - it must be internal. https://t.co/mwdQzwKkaI

2023-02-12 09:20:24 Asher (8) and Yaakov (6), you don't know me, but I know you, albeit from a distance, and I am retweeting your pictures to decent people all over the world, just to remind God of His promise: evil will extinguish itself from this planet. שׁוֹבֵר אוֹיְבִים וּמַכְנִֽיעַ מִינִים: https://t.co/rPYfDwHiZI

2023-02-12 08:56:02 RCT studies are not falsifiable because RCT practitioners reject the idea that RCT has a goal, named ATE, residing outside the RCT protocol. Once they accept ATE=E[Y(1,u)-Y(0,u)], RCT results turn falsifiable, see Eq.(9.32-9.33) in https://t.co/mfZhzorAGq. Sadly they reject it. https://t.co/mUJSTEClMT

2023-02-12 08:31:26 @f2harrell @stephensenn @HL327 @elmir1omerovic @OstlundOllie @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone IMHO you are mistake if you think I am. If you think @stephensenn's challenges are well-written, please translate them into reasoned, self contained ideas, not via links nor via judgmental verdicts such as "fictitious", "unrealistic", "untestable"...

2023-02-12 08:15:32 @f2harrell @technoslerphile @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone I give you credit for articulating so eloquently the goals of clinical decision making: " the real therapeutic question is how does the outcome of a patient were she given treatment A compare to her outcome were she given treatment B." We write it Y(1,u) vs. Y(0,u)

2023-02-12 07:56:41 @technoslerphile @f2harrell @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Your exchange with @f2harrell represents a clash between two healthy intuitions: (1) OS should mend some of RCT's imperfections (2) OS itself is imperfect. Clash! Only CI principles can reconcile these two. Conclusion: learn to speak CI if you wish a reconciliation before 2030.

2023-02-12 00:26:22 @HL327 @f2harrell @stephensenn @elmir1omerovic @OstlundOllie @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone Sure. A 2nd RCT with narrower (and measurable) subpopulations would be more informative. But the reason for the alarming OS could be unmeasured, (e.g., tr. choosers were ladies who had a cousin that died from corona), we just go with what we learn from OS combined with 1st RCT.

2023-02-12 00:14:05 @stephensenn @soboleffspaces Our intention was not to confound school with diet or with hall. There is one school, two dining halls, each hall has one diet plan. No confounding. If you think the text was not clear, pleas tell us what wording may have caused ambiguity.

2023-02-12 00:03:05 @HL327 @f2harrell @stephensenn @elmir1omerovic @OstlundOllie @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone @HL327, You are ahead of me. Sure, upon replicating our original RCT with ATE = 0.279, we discard the OS and decide to treat. Except, in 2023 we can do even better: combine the two (barring interruptions) to estimate P(harm), and see if we shouldn't flip our decision again.

2023-02-11 23:49:11 @HL327 @f2harrell @stephensenn @elmir1omerovic @OstlundOllie @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone Moreover, now that I have two studies, RCT and OS, I can do better and (barring interruptions from RCT zealots) combine the two to estimate P(harm), and, upon finding P(harm)=0, decide to treat with a smile, as opposed to "treat with hesitation."

2023-02-11 23:19:59 @HL327 @f2harrell @stephensenn @elmir1omerovic @OstlundOllie @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone Seeing a 73% death rate in the OS would make me wonder: Is it possible that women who choose trt. are mostly badly ill , hence desperate patients? I would then conduct a pilot RCT and, upon seeing ATE >

2023-02-11 23:06:33 @stephensenn @soboleffspaces Where do you see a "parallel study" ?

2023-02-11 00:49:55 @AndrewPessin @elderofziyon Don't blame Palestinian journalists. Prostitution thrives when demand increases, and Western readers are dying for more and more of this sort of stories on cowboys and Indians.

2023-02-11 00:34:52 @RaulMachadoG @leeflang_m @soboleffspaces @nickchk @eliasbareinboim Part of the problem is that most "business questions" are causal, and business schools teach stat, not causality, so practitioners can't articulate what they really want. Last week I discovered this to be a problem even among trialists, who should be doing cause-effect for living

2023-02-10 23:55:53 @stephensenn @f2harrell @elmir1omerovic @OstlundOllie @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone Toy examples convey principles, and make all assumptions explicit and transparent. "Actual data" examples, bury principles and assumptions in noise, thus permitting researchers to argue endlessly about irrelevant details, and escape even the key question: What's our estimand?

2023-02-10 23:35:02 @LGHemkens @soboleffspaces @smueller Not clear to me what assumpson you referring to.

2023-02-10 20:47:01 @LGHemkens @soboleffspaces @smueller When we write E(Y(1,u)) we do not need the data from all the u's to estimate this expectation. Statistics was created for the purpose of estimating it from finite sample.

2023-02-10 19:24:03 @f2harrell @elmir1omerovic @OstlundOllie @stephensenn @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone That's exactly my point: You are not referring to the theoretical definition of ATE, nor to the difference between this ATE and whatever comes out of "actual data" (let's call it ATE**) . ATE simply doesn't exist, and life revolves around ATE** with no guidance to quality.

2023-02-10 10:58:07 To my friends and colleagues in Barcelona: How could you allow this to happen? You have invited me to visit your city and universities. How can I accept when my heart tells me: "This city maligned the city of birth"? Or "this city believed vicious lies against my people." Hard. https://t.co/HDOW4H93cC

2023-02-10 10:36:24 RT @vardi: Please join me in the upcoming @AAAS meeting in Washington, DC! https://t.co/htmOPoZR6N

2023-02-10 09:50:33 @MaimonAmir @Ostrov_A How about two Jewish academicians, at my university, writing to the UN urging it NOT to adopt. Which is harder to believe?

2023-02-10 08:58:37 For CI folks ATE is a quantity that exists in reality and which RCTs aim to estimate. For a trialist, ATE is whatever the trial delivers, assuming it is designed by good RCT experts. While CI folks might ask: what makes for a good expert? or whether a good expert is more 1/2 https://t.co/OPV5JV1Ez7

2023-02-10 06:25:46 @returnstosender @StudiesZionist @returntosender. Not clear what you have against @StudiesZionist, an organization that I've joined recently. They decry comparisons between Auschwitz and Gazza - don't you? Doesn't every decent person? They need our support, not sarcasm. ??????????

2023-02-10 06:05:09 but on comparing the outcome of an individual u under treatment to the outcome of that SAME individual under control. The difference is more than philosophical

2023-02-10 06:05:08 The Wikipedia supports my story: In a randomized trial (i.e., an experimental study), the average treatment effect can be estimated from a sample using a comparison in mean outcomes for treated and untreated units. However, the ATE is generally understood as a causal 1/3 https://t.co/qiBmlY1lr1

2023-02-10 05:24:49 @DKedmey @AdamL @jgreenblatt If "leader" is one who follows his convictions, then Koch was a super-leader. We didn't include his essay in "I am Jewish" book because he refused to make the editorial changes I requested. Danny's last words are now inscribed on his headstone: https://t.co/V6gyOGIG4y

2023-02-10 03:33:22 @f2harrell @elmir1omerovic @OstlundOllie @stephensenn @RWJE_BA @PatientStormDoc @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone This stands against everything I've learned from the literature. I believe it was D. Rubin who coined the term ATE, and defined it as: ATE = E(Y(1,u) -Y(0,u)) In other words, ATE is a population property, not a feature of any given study. Do Rubin's disciples agree with you?

2023-02-10 02:17:15 @lilienfeld1 I wouldn't give Roger Waters the benefit of being an antisemite, a hereditary disease usually forgiven for reasons of sickly upbringing. Waters is worse

2023-02-09 20:48:14 Excited to read the whole paper. Who could imagine that the axiomatization of counterfactuals could be applied to images! https://t.co/j4jhdCE5Q3

2023-02-09 16:13:08 For all Hebrew speaking readers of our channel, I am supporting author Ilan Sheinfeld's appeal to resist the ongoing encroachment on the secular-democratic character of Israeli society. https://t.co/Tc0uYqLcBj

2023-02-09 12:27:53 In IRGC defense, I would check first if this is not part of "The Resistance". https://t.co/8l4iKKd7MN

2023-02-09 12:05:37 https://t.co/SOoFWuaHu0

2023-02-09 04:46:03 @Lester_Domes @AnjaLeist @rohitpojha @EpiEllie @_MiguelHernan @casualinfer Sander: @_MiguelHernan's recent "Like" jolted me to read this thread again and to realize -- Woo to me -- that I still have not seen ONE problem that TT addresses. Willing to learn, but I can only learn from problems, not books.

2023-02-09 04:22:18 @elderofziyon Only fake "Rabbis" can stay unmoved when babies are rescued from the rubbles.

2023-02-09 02:22:52 Only a Zionophobic mind like that of Roger Waters could discover that "the Russian invasion of Ukraine was NOT unprovoked." https://t.co/8VDZ2IJ5I0

2023-02-09 02:15:30 RT @Martin_Kramer: Bernard Lewis: “Imperialism, sexism, and racism are words of Western coinage, not because the West invented these evils,…

2023-02-09 02:08:23 @AleksanderMolak Very interesting thread. Can we have links to your book and to "Statistical Rethinking".

2023-02-09 01:03:30 @soboleffspaces @nickchk @eliasbareinboim "Lack of transparency"? Where? Who? Who said? Why? How? Holy Joe! Really?

2023-02-09 00:55:10 @richard_landes @Mivasair The "Rabbi" is right. Zionists have never been under physical attack before.

2023-02-08 13:59:52 RT @CotlerWunsh: “He who saves one life, saves the entire world.” In the face of rising devastation &

2023-02-08 12:10:16 @nickchk @Blogsbloke I wrote somewhere that Nick's book may emancipate economics from the tyranny of outdatedness.

2023-02-08 07:38:22 RT @StudiesZionist: @jewish_studies The AJS is a scholarly org , yet it has chosen Palestinian activism even if it meanings harming Jews.…

2023-02-08 07:17:40 @ivanerogg @soboleffspaces Of course

2023-02-08 05:52:46 I rarely miss an opportunity to RT a video clip by @EinatWilf, but this one is particularly on-target today, when a combination of forces are attempting to turn Israel into a Messianic (Halachaic) State. They will fail, because the anatomy of Zionism won't swallow it. https://t.co/TO8Tp8GkBb

2023-02-08 05:34:15 I wish I was a believer: "Oh, God, don't make @daneasterman lie down there, in Gaziantep, trapped in the ruins, at -12 degrees Celsius, in the darkness, as IDF rescue workers are sweating to find live victims. He is the lowest of humans, true. But don't take him to Gaziantep." https://t.co/o3okpSG4BY

2023-02-07 23:33:28 @Patrikbentolila The other areas are too numerous to mention, and are well known to people who have their eyes open. The trees, however, is an inspiring achievement that only few people know.

2023-02-07 20:32:51 @elmir1omerovic @PatientStormDoc @stephensenn @f2harrell @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone Your experience in Sweden is interesting, can you summarize in a few lines what information you get from the OS? If you can explain it in CI language, it would be great

2023-02-07 19:49:34 @Anatomi54244944 It took a century for science to earn its reputation, at the cost of many good people's torture. Now that it has earned it, it is our obligation to prevent imposters from steeling this reputation and leading many good people towards wrong treatments or worse.

2023-02-07 13:13:49 @DiazQuijanoFA @MadelynTheRose @NP_Jewell Wow! Did we really say that? Where did we get the wisdom to predict precisely why we would be attacked later?

2023-02-07 11:02:11 the paper: "Epsilon-Identifiability of Causal Quantities", perhaps because I am a co-author (with Li and Mueller), or perhaps because it reminds me of Epsilon-Semantics -- an early work in non-monotonic logic which formalizes the notion of "almost sure". Enjoy 2/2

2023-02-07 11:02:10 It's time to take note of the CI papers published last week: https://t.co/CrE7MNnZ2L. I sense an increasing activity in Causal Discovery, but my attention goes to the paper on "Actual Causation and the Challenge of Purpose", perhaps because I am purpose-driven, and to 1/2

2023-02-07 08:27:55 @elmir1omerovic A waste of precious time. Unless you enjoy hearing "not true" "wrong" "erroneous" repeated for the hundredth time, with no explanation.

2023-02-07 07:55:16 As we are leaving Tu Bishvat https://t.co/4IwHZRu2KQ, our Tree-Planting Holiday, I am reminded that Israel is one of the few countries in the world that entered the 21th century with more trees than 100 years earlier. I even remember the Ficus tree I planted on TuBishvat, 1940.

2023-02-07 07:41:34 A major court case will take place in two weeks, which might affect everything we know about social media. Moreover, Nitsana is a lioness, so I wouldn't bet on the outcome. @AttorneyNitsana @ShuratHaDin

2023-02-07 07:41:33 WATCH: Will Israeli attorney bring social media giants to their knees? | World Israel News https://t.co/8I3qKnCxpU

2023-02-07 06:20:08 @Anatomi54244944 Co-existence by all means, but shouldn't we distinguish the two modes of thinking, and not allow the emotional mode to pose as a scientific mode, as we find in some cults, like Scientology, Christian-Science or Happy-Science?

2023-02-07 06:04:05 @f2harrell in the space of continuous functions, how severe it is and, more importantly, whether such pathologies also inflict attempts to remedy imperfections in RCTs, for example, concurrent control with its additivity assumptions. Cause for celebration? Ask your friendly CI doctor. 2/2

2023-02-07 05:52:58 @f2harrell At the risk of spoiling the jubilee at the trialist community party, note what fault has been found in OS: There are pathological functions (eg non-differentiable) that could make the estimand derived "non-uniformly consistent". We do not know how prevalent this pathology is 1/2

2023-02-07 02:33:47 REGISTER NOW! 2022-23 Daniel Pearl Memorial Lecture with Garry Kasparov https://t.co/ACg5ZOTw0M March 6, 2023, 4: 30 pm, UCLA.

2023-02-07 01:59:31 The emphasis is on " in the manner implied ", which is buldozertized in style and scope. https://t.co/4Mm8CK42Uz

2023-02-07 01:54:15 @GoldsteinBrooke Commenting "free Russia" on a Ukrainian person's post may not be antisemitic, but it sure is morally disgusting and genocidal.

2023-02-07 01:47:44 @elderofziyon C:urious: Is there a committee, a librarian, or a custodian on the Inalienable Rights of the Jewish People? Are there such rights?

2023-02-07 01:30:40 Israel sends search-and-rescue delegation to Turkey https://t.co/iqnAbdpNIJ via @JNS_org

2023-02-07 01:03:08 @splendorsolis69 No way. "Schmucks" is judgmental

2023-02-07 00:56:15 Extremely useful chart, to be posted in the hallways of every "Evidence Based" department. Luckily, students of Causal Inference are not suffering from this imprecision

2023-02-07 00:32:35 @elmir1omerovic @RWJE_BA @PatientStormDoc @stephensenn @f2harrell @dylanarmbruste3 @drjohnm @kaulcsmc @GreggWStone Thanks for defending the faith, although I do not know what was accused of - I've muted RWJE long time ago, out of social pressure to focus on the constructive. There are so many readers who wish to learn, that I can't waste time fending off detractors of learning.

2023-02-06 22:39:39 @imthealan @ylecun @VladicaV Zelensky's canonical model may not be exactly canonical to Putin. My canonical model is SCM -- a model of reality, not a model of our algorithm.

2023-02-06 22:31:43 As soon as I posted this reply, I recalled my daughter warning me that the word "pagan" may offend the majority of living humanoids, and that I should be very very careful. So, seeking advice, what should be a respectful way of shunning those who muddy the unity of science? https://t.co/Uk4Io6DHYG

2023-02-06 22:06:57 @ylecun @VladicaV And my proposal calls for first studying what we can do with causal world models, before we labor to learn them. Still not mutually exclusive.

2023-02-06 22:00:38 @Manuel_do_rio @lexfridman Merry sailing, and don't let the pagans muddy the unity of science.

2023-02-06 20:43:39 @Manuel_do_rio @lexfridman I haven't changed that prediction. Part of it has already been accomplished through the formal distinction between "deliberate" and "imitating" actions.

2023-02-06 20:35:27 Here is an Open Letter to George Washington University that I've signed today: https://t.co/0C53HQvF2a which I believe every concerned academic should sign, in light of the abusive practices perpetrated by some of GWU's teaching staff. Please consider adding your name.

2023-02-06 19:29:31 Great thread for students of causality and cognitive science. I didn't realize there are psychological theories today that try to explain causal judgments without counterfactuals. Are there? What makes them resistant to the Ladder? https://t.co/iPTqN56qL5

2023-02-06 08:30:50 At the peak of his fame, Albert Einstein travelled to the holy land to lecture at the Hebrew University in Jerusalem, and to lend his name to an institute that would become a symbol of purpose and renewal to his people. One of many lectures I am sorry to have missed. https://t.co/kkbV1lqDbe

2023-02-06 07:43:06 RT @richard_landes: woe onto the Jews when their bright youth are useful infidels. Fisking Mira Fox at the Forward on Jenin https://t.co/s…

2023-02-06 05:36:45 It's one of the most perplexing problems facing physicists these days: Conservation of Matter. If Palestinian teenagers are all scholars, dreamers, artists, and lovers of mankind, where do their people-shooting teenagers come from? Physics is in an uproar. Not @Reuters . https://t.co/06TNR5r51O

2023-02-06 04:56:58 Published 3 years ago, but still a most powerful rebuttal of American Jews who seek redemption through suicide. https://t.co/OQuagOs1Hj

2023-02-06 03:21:38 @nickeread Nothing much changed the past 400 years, the poor are still more easily recruited.

2023-02-06 03:14:32 @PatientStormDoc @stephensenn @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Not sure if you are referring to me, I know however that, when I reframe, there is benefit to the reframing. It is often done to exit from the clutter of details to the clarity of principles.

2023-02-05 22:24:45 @AlfredoMorabia @soboleffspaces Thanks for this illuminating chapter of history. As we can see from the quote, Helmont deployed both randomization (cast lots) and intervention (i'll treat mine my way, you'll treat yours your way). Leaving no room for ambiguity.

2023-02-05 21:53:27 @soboleffspaces @stephensenn Some would argue, of course, that you can't disentangle estimation from identification, because, reading stat textbooks, the two are badly conflated and entangled. Why? Lacking identification + estimand, you never know what needs to be estimated. These days are over.

2023-02-05 20:44:45 @soboleffspaces @stephensenn When I was still hoping to have fruitful conversations with @stephensenn I used to say: "SCM lets smart statisticians address the SE issue". Today I dare say: "SCM has done more for the SE challenge than most statisticians. How? By disentangling estimation from identification"

2023-02-05 13:01:52 Daniel's experiment is described on pp 134-7, chapter 4. Lind's experiment on p.302, R.A. Fisher and the modern rationale for RCT, on pp. 143-150. #Bookofwhy pagination remains the same from printing to printing. Enjoy. https://t.co/qYXv9CzzRD

2023-02-05 12:02:45 @whysomanykaykay @IAmSamFin Helmot wrote "Let's cast lots". Daniel said: "Take four of us and feed us only vegetables

2023-02-05 09:07:54 @IAmSamFin Daniel's experiment is described in details on pages 134-7 of #Bookofwhy. The novelty of Helmont was the "randomization" element. BTW, Helmont was the first to coin the word "gas".

2023-02-05 08:54:57 @Ostrov_A No! The entire crop understands "peace" one way: The elimination of Israel. But Dahlan says so explicitly, while the others try to obfuscate.

2023-02-05 08:30:45 let us cast lots, that one halfe of them may fall to my share, and the others to yours

2023-02-05 08:30:44 Correction to #Bookofwhy: The concept of a controlled clinical trial was first described in 1662 by Jan Baptist van Helmont (not J. Lind, 1747) who wrote: "Let us take out of the Hospitals... 500 poor People, that have fevers or Pleuritis. Let us divide them in Halfes, 1/2

2023-02-05 06:05:39 RT @AmericanSephard: @yudapearl You may find our co-authored article of interest: https://t.co/41BZ6Cl83I

2023-02-05 04:30:04 GOP insane hypocrisy does not make @Ilhan Omar fit to serve on the HFA Committee. Can we discuss the latter squarely, as Democrats should, w/o diverting to the former? See an attempt: https://t.co/zyBilhd6uG https://t.co/0xSk11raf0

2023-02-05 04:08:11 A worrisome move to be watched. I met Dahlan briefly in Doha (2005) - hardly a man of peace. See "Dialogue of the Deaf" https://t.co/S78zGwO8vi https://t.co/jojiBixOJr

2023-02-05 03:49:13 @RepJasmine @Ilhan @Repjasmine sounds like she genuinely never heard why @Ilhan was found unfit. Where has she been the past 4 years? Is it true that some Congressmembers get their entire education from a social media bubble?

2023-02-05 03:15:15 @EinatWilf Jewishness is a necessary, not sufficient condition for democracy, as we remember from the 2nd Temple, when Jewish "debates" went out of control to become "violent disagreements" if not a "civil war" (Milchemet Achim).

2023-02-04 23:56:36 @maradonasattva This line reminds me of the joke: "Why are you molesting your sister?" "I don't have a sister". "According to Al-Jazeera you do".

2023-02-04 23:19:59 @ProfDBernstein @DavidLBernstein Thank you for the correction, David. Yes, the book I meant is "Woke Antisemitism" a great supplement to "Classified:.." Two great Davids, two great Bernsteins. We need more such writers.

2023-02-04 23:11:58 @maradonasattva Zionophobia is as much a valid criticism of Israel as Islamophobia is a valid criticism of ISIS.

2023-02-04 21:14:46 @maradonasattva US campuses are the microcosms of the future. Besides, worldwide islamophobia may exist as a phenomenon, but it is still considered unbecoming in good company

2023-02-04 20:21:01 @soboleffspaces @RohanAlexander To continue your test, count the number of causal utterances in the stories, and compare it the the number of statistical utterances. Publish the result in The Annals of Statistics.

2023-02-04 20:17:20 If LLMs become instruments toward the democratization of Causality, something #Boodofwhy has aimed to do, then I wouldn't write them off so quickly. https://t.co/dR1qIVL9yW

2023-02-04 20:02:31 @WilliamWonkaJr No match. She doesn't have the venom.

2023-02-04 19:59:23 @rwolffoot I forgot my category theory. Can we think of the boxes as generalizations of "functions"? Given that we count on humanoids to provide those models, we need a primitive metaphor for thinking about them. In SCM I use "listening to". What should I used for "strings"?

2023-02-04 19:51:58 @maradonasattva Islamophobia on my campus is a cardinal sin

2023-02-04 14:14:35 @maradonasattva It's uglier because Islamophobia is a cardinal sin in our society, condemned by every politician, educator or clerics, while Zionophobia is becoming a cute sport among the "respectables," including some academics that I know.

2023-02-04 13:28:04 Congratulations, Moshe. What a great collection of articles just added to my "to read next" heap. https://t.co/0nofo6dOsa

2023-02-04 12:59:59 Last Sunday I attended a meeting with David Bernstein, author of "Classified: The Untold Story of Racial Classification in America.” It's a must read for parents of school children in 2023 America, especially Jewish parents, see https://t.co/ZsTTlobo7k

2023-02-04 12:25:38 RT @billmaher: If you're part of today's woke revolution, you need to study the part of revolutions where they spin out of control. https:/…

2023-02-04 12:10:37 Today it's the Society for Psychoanalysis and Psychoanalytic Psychology, tomorrow its in your own university. The normalization of hate in academia is slow but steady. https://t.co/R2epZM2YzC

2023-02-04 11:50:59 A star scholar on antisemitism is born. She should apply for Fellowship in one of the many Centers for Antisemitism in academia -- Harvard? https://t.co/votNBVB60I

2023-02-04 11:39:58 The squad is furious!! And they have a point! No one has warned them, not even the Jewish establishment, that Zionophobia is uglier than Islamophobia. https://t.co/IGgvZHTcbD

2023-02-04 03:31:55 A very astute suggestion, in my objective opinion. But my attention goes to @RohanAlexander https://t.co/HWYgz8CNom Because I am in the opinion that you cannot tell stories with data

2023-02-04 02:36:29 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Sorry, I can't parse it. You say " if p(Y_x|x,z’) =<

2023-02-04 02:21:22 @Claire_V0ltaire I never use the word "majnoon", it sounds judgmental. Instead, I use "Palestinian mentality", to let them project their mental state as authentically as they can. They would never fail to surprise us by over-exposing their real selves.

2023-02-04 01:11:54 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone For the life of me, I can't see how the observational study could be the result of assuming anything, let alone consistency . It is just estimating good old fasion P(x,y,z),

2023-02-04 01:01:45 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone We don't know P(Y_x|x,z) from an RCT. All we know is P(Y_x|z).

2023-02-04 00:05:40 This photo of almond blossom reminds me of Tu Bishvat, Israel's national holiday of Tree Planting, to be celebrated Monday, Feb 6, https://t.co/4IwHZRu2KQ "Birds from Every Roof, Are Welcoming the Holiday" https://t.co/TgxxzzJRhs

2023-02-03 23:56:03 RT @peterhimmelman: My Green Kite https://t.co/xi08CSefvW via @YouTube

2023-02-03 23:41:26 @soboleffspaces Sail the wind you have! Amen! The wind and other capabilities enter into the Agency-based consequence Y(x), not so the urges I had last time I sailed.

2023-02-03 23:18:38 @soboleffspaces Yet we distinguish "behavioral decision theory" from "rational decision theory". The latter talks about consequences stripped of emotional urges. https://t.co/nDpUzRoEKU

2023-02-03 23:00:39 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone This inequality is generally correct: P(Y|x,z)≠P(Y|x) Consistency, however, says: P(Y_x|x,z) = P(Y|x,z)=/=P(Y|x)

2023-02-03 20:28:24 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Consistency is individual-level relation, so, X=x &

2023-02-03 20:18:19 @rwolffoot #Bookofwhy says: "I believed no one could do this without the do-calculus". That was 1994. Causality p.82 has a proof w/o do-cal. PO folks are still working on it, and Heckman&

2023-02-03 11:48:08 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone I am not sure what you mean by "that's not true". I can assure you that an analyst would never apply the rule unless the same covariate appears on both side of the equation.

2023-02-03 10:37:37 RT @israel_advocacy: Omar Suleiman just told lie after lie about Israel to Lex Fridman, and Lex just sat there nodding. @lexfridman @omars…

2023-02-03 10:05:18 So, what makes Ms. Jones different, able to outperform her other free-acting peers? A simple ingredient: Agency. She is deciding freely, ignoring whatever advice she received

2023-02-03 10:05:17 As promised, I'm coming back to the problem of Ms. Jones - a randomly chosen female. Could she possibly have zero harm and 30% chance of benefit from taking the drug, given that, in females, 73% of free choosers and merely 30% of free rejectors died? Enough to caution 1/4 https://t.co/sk6GjoqJbu

2023-02-03 08:03:16 RT @PHuenermund: Is it really necessary to quote Lenin? @techreview https://t.co/TTnmG62dmi

2023-02-03 04:38:49 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone The consistency rule is always used under the same Z.

2023-02-03 00:36:12 @HL327 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone In epidemiology, consistency was first formulated by Robin (1986) as: (X=x) implies Y(x) = Y [Causality pp. 96] It is a Theorem in SCM. It says: An individual who chose treatment X=x and responded with Y, would also respond Y if forced to take X=x in an experimental study.

2023-02-03 00:15:38 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Surely time matters. But in 2023 we should be able to express such healthy intuitions about time into formal cause-effect relationships and carry them over to RCT findings, to mend the latter imperfections. Can we?

2023-02-02 22:37:01 The Nineteenth Conference on Theoretical Aspects of Rationality and Knowledge (TARK 2023). June 28 until June 30, 2023, at the University of Oxford, United Kingdom. Website:https://t.co/hw5clvAfRl Extended abstracts can be submitted at:https://t.co/rAINAnph2k

2023-02-02 21:09:07 @stephensenn @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone In classical RCT, Q1 and Q2 reduce to classical statistical estimation, no causal assumptions are needed, and no causal thinking is needed.

2023-02-02 21:00:13 Re-directing our fight against "Zionophobia" is critically needed today, even more than 4 years ago, when I wrote this article for UCLA students. I hope it is not too late. Yes, @AdamL and @JGreenblatt can turn the tide. https://t.co/lMsdi1aij1

2023-02-02 20:44:20 @PatientStormDoc @RWJE_BA @hemanth_kanak @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone @f2harrell I am sharing with my friend @MissDiagnosis, an exert on sleep apnea: Did you know that?

2023-02-02 20:41:42 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone It would be helpful if those who cite causal data mining papers take the time to explicate the causal assumptions made by the authors, rather burden twitter followers with that task.

2023-02-02 20:32:10 @PatientStormDoc @stephensenn @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Why go to warfarin overdose and not ask in full generality: "Can statistical associations determine causal probabilities?". The answer is provenly NOT without causal assumptions. So, why go to warfarin overdose?

2023-02-02 20:27:14 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone I don't get it. If an Observational Study provides a (well estimated) probability P(x1,x2,x3..) on many temporally ordered variables, how can this aid in mending RCT imperfections?? Here is where the words of an enlightened trialist can help understand "using timing to explore".

2023-02-02 15:13:54 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Can we unfold "using timing to explore causality"? What is the input information, what causal conclusion we infer from it, and how is that conclusion used to amend RCT imperfections (where our goal is to estimate ATE at some population, i.e., P(y(1)-y(0) | covariates).

2023-02-02 13:30:34 A beautiful microscope unveiling the anatomy of the 100-yrs Israeli-Palestinian conflict. One side: "Two-state solution?" The other side: "You are guilty no matter what!" I watched it 26 times. https://t.co/lyqhwd8PFE

2023-02-02 12:40:31 @stephensenn Forget for a moment whether the advice given to Ms Jones is justified, or how it was derived, she is asking whether it is at all possible that she, a patient named Ms Jones, would fare better than her peers by choosing rather than rejecting.

2023-02-02 11:36:19 @stephensenn Revolutions are fueled by the curious, not the skeptics. Study effects exist, but have nothing to do with Ms. Jones question: "I am part of a population in which drug choosers performed much worse than drug rejectors. Why should I, Jones, fare better by choosing over rejecting?"

2023-02-02 11:31:40 @stephensenn @PatientStormDoc @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone "Pointing" to literature is not as effective as concisely explaining ideas or methods. In this particular case, "pointing" led me to a forest of unrelated topics. Can we summarize: what information can OS provide to mend imperfections of RCTs?

2023-02-02 06:13:18 Are you serious? Rep. Omar has never eaten cockroaches, and never slept with Putin. Does that make her qualified to decide how Israel should defend itself against Iranian threats? https://t.co/zyBilhd6uG @Ostrov_A @RepAdamSchiff @RepMcGovern @robsatloff @RobEshman https://t.co/68f3j1tMer

2023-02-02 05:44:47 The hell with Nuentes and West, isn't it enough that Rep. Omar is infected with an incurable anti-Israel hatred? Isn't it sufficient to disqualify her from HFAC? Please listen to the arguments I presented to Nancy Pelosi in 2019 https://t.co/zyBilhd6uG @Ostrov_A @RepMcGovern https://t.co/68f3j1tMer

2023-02-02 05:16:49 @DBSpro539 Danny graduated Birmingham in 81, and he loved it, especially his English teachers.

2023-02-02 04:12:59 RT @jazchaz: Today, as we mark 21 years since the murder of Jewish America journalist Daniel Pearl,let us all reflect on his steadfast beli…

2023-02-02 04:11:23 Thank you @bentiriderdeb for proving, as the 1st Principal of the Daniel Pearl Magnet High School, that even a humble Jewish boy from Encino CA can turn into an icon of courage and excellence to hundreds of high-school students from all walks of life, all over Los Angeles. https://t.co/XHb0Y7Mkv5

2023-02-02 03:58:58 RT @Foodaism: @yudapearl @yudapearl Danny's memory has proven a blessing to the world, not just because you, Ruth and Danny's whole family…

2023-02-02 03:54:44 @PatientStormDoc @stephensenn @RWJE_BA @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Refreshing to hear a voice of wisdom, insisting on stripping RCT from its elitist privileges and returning it to the scientific fold. Can you opine on the role of observational studies in this process?

2023-02-02 00:37:35 RT @Psalm2348750835: #Jewish #IStandWithJewishPeople #IsraelStrong https://t.co/YMyOKwDADO

2023-02-01 09:47:37 @RonKenett Depressing. No idea who's behind it.

2023-02-01 06:27:19 even have a chance to make up my mind, when I could. If I were to make up my mind now, wont I be just like those other women, with 75% chance of dying once they choose it and 30% once they reject it?" Before we deal with "consistency" , namely, whether Dr. X was justified in 2/3

2023-02-01 06:27:18 Ms. Jones: "Dr. X! How dare you tell me 'you have good change to benefit from the drug and zero chance of being harmed by it' when you know that more females died (73%) after choosing to take the drug than choosing to reject it (30%). You know nothing about me, and I didn't 1/3 https://t.co/8SpUUwzkfW

2023-02-01 02:20:43 I am deeply ashamed by the unacceptable behavior of my countrymen and, yes, I do feel responsible for their action, and for not doing enough to prevent their educators from entering classroom. Both should be publicly exposed and harshly punished. https://t.co/wMzIFlYGo0

2023-02-01 02:05:24 @Martin_Kramer @BardCollege @NathanThrall Clever move. When students complained about a "History of ME" class at UCLA, I demanded that it be titled "Palestinian-biased History of ME" in the catalogue. But the Chair insisted that a professor can choose any title he wants, even a misleading one, like "History of ME".

2023-02-01 01:29:05 @HL327 @stephensenn @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Yes. Females have zero prob of ‘indiv harm’, though more females died after choosing to take the treatment than choosing to reject it. If you are stunned by this finding, ask what assumption is offended in your intuition, and examine its justification under a formal microscope.

2023-02-01 00:36:56 Danny's official Yahrzeit is tomorrow, February 1st. Tonight we'll light the Soul Candle (Ner Neshama) and ask him to speak on our behalf, a world gone mad, in places where humanity, wisdom and courage once reigned. Yitgadal V'Yitkadash Shmey Rabba https://t.co/bbmbinQCyB https://t.co/1STPtsbvAb

2023-01-31 23:46:28 @stephensenn @HL327 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Jokes aside, this paradox have been studies by philosophers: "The paradox of inevitable regret" https://t.co/2tOMqlFx0r. But it takes more than mockery to deal with it seriously.

2023-01-31 23:13:49 The practical aims of CI are the same as its computational aims: To answer causal questions of interest. Some have immediate practical implications, e.g., policy evaluation, and some less immediate, e.g., assessing individual responsibility or individual benefit &

2023-01-31 21:14:16 Even political Islamists may be forgiven some day, when the All-Mighty opens wide the gates of Mercy. Not the Mehdi Hasan's among them, who have sold their last shred of morality, and are working 24/7 towards the destruction of Israel. https://t.co/Hod2Nvm2t2

2023-01-31 17:06:45 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone First, I am surprised you are not curious to know what conclusions one can draw from a provisional set of assumptions that can be enriched and even proven necessary. Second, some of those assumptions may have testable implications, and some shown plausible by sensitivity analys.

2023-01-31 16:15:49 @dylanarmbruste3 @f2harrell @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Disagree. CI recognizes that sample data is finite but it also recognizes that, in order to use our finite data to estimate actionable quantities (eg ATE), we first need to represent those quantities as estimable statistical expressions, a task neglected by some statisticians.

2023-01-31 15:43:07 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone CI deals with the first (ie estimation from finite data) the same way statisticians deal with it in RCT and in prediction tasks, plus, by presenting smart statisticians with new estimation challenges: estimands they haven't seen before.

2023-01-31 15:29:37 @cogscikid I'm still playing with several strategies. Basically, LLM need to be taught what a "good explanation" is, then how to generate one for a specific query, from a combination of models, data and stories.

2023-01-31 15:19:12 @SamuelJHyde11 @RasulElon @njhochman Plus an addendum: This magical homecoming was legitimized and enabled, against the wishes of the colonial powers (Britain and conservative America) but as an anti-colonial force, by the Soviet Union, its satellites, socialist Europe and progressive America.

2023-01-31 08:25:33 @matloff Interesting analysis - agree!

2023-01-31 07:36:36 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Here are 90 articles in CI, published last week, at least half on "concrete applications of CI." I don't think they all "come up short in taking all uncertainties in account". They simply assess uncertainty of estimation on top of causally produced estimands.

2023-01-31 05:53:41 Mehdi Hasan to Israel: "How is this not…an act of war? " Mehdi is shocked and startled at the sound of WAR because, until yesterday, he heard the Ayatollahs singing nursery rhymes. @MSNBC

2023-01-31 05:42:09 It's hard to contain the over flow of new articles in Causal Inference, let alone read it all. Here is last week's yield: https://t.co/HPdnldNF3j. Please share if you find nuggets of gold touching on your area of research.

2023-01-31 05:23:31 A fairly goo introduction to causal inference: The science of automating that "behind-the-scenes" work, or a good chunk of it. https://t.co/mhlc7cHjO5

2023-01-31 04:39:42 @njhochman Good point. Yet the failure of our progressive activist to see that Israel in a homecoming endeavor, bearing no resemblance to Western colonialism, that failure, while not inherently anti-Jewish, is certainly morally despicable, and anti-progressive.

2023-01-31 04:17:11 I hope it is clear that this Tweet is aimed at NPR, based on its coverage of the Jerusalem terrorism. https://t.co/7S0KFG7tQe

2023-01-31 03:59:46 @DavidDeutschOxf There is only one thing that would need to be set on fire to allow this: National Identity. Some say: D-N-A, but we are talking political philosophy.

2023-01-31 02:25:50 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone CI distinguishes two types of uncertainty

2023-01-31 01:04:51 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone What you are saying is that you feel uncomfortable making a new kind of assumptions, called "causal", even though theory says you can't get ANY causal conclusions whatsoever by making only the kind of assumptions you are used to make, called "statistical." Thank CI for it!

2023-01-31 00:50:57 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Bernoulli's great insight was: You would never know how good your estimate is, unless you assume that, behind your data, there is an association with a definite value.

2023-01-31 00:46:20 The little I can do to honor the memory of the seven terror victims in Jerusalem is to echo their last thought: Resilience - the spirit of the Israel will endure. https://t.co/IjGJhQrOJv

2023-01-31 00:39:38 RT @vardi: Jewish doctors in the Warsaw Ghetto secretly documented the effects of Nazi-imposed starvation, and the knowledge is helping res…

2023-01-30 20:35:12 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone The overriding assumption in CI is not that the data is infinite, but that the finite data we see is generated by some "joint distribution function" in which association of lack of association is not a matter for debate but has a definite value. (Bernoulli, 1710)

2023-01-30 13:12:23 @nytimes! Your readers would appreciate it if you can refer to a Palestinian "man" as "scholar", and to an Israeli "man" as "soldier". it's hard to stomach ambiguities. https://t.co/bdUhXl1DBm

2023-01-30 12:51:36 @SAFECUNY This is the price we pay for charging he with antisemitism instead of Zionophobia, a bigotry which she, as a member of the House FA Committee, should distance herself from.

2023-01-30 12:26:42 RT @AvivaKlompas: Al Jazeera - mouthpiece of the Qatari regime - created a video to glorify the terrorist who murdered 7 unarmed Jews on Fr…

2023-01-30 10:19:18 equation (10) implies that the regression of Y on X at the target population can be inferred from regressions estimated in the study, W=1, regardless of how X, Y and W are interacting. Two distinct cases are worth noting. (1) If W is a pre-treatment variable, as in Fig. 3, 3/5

2023-01-30 10:19:17 To understand how trialists, under additivity, can get away without causal calculus, it is instructive to see how selection bias cures itself in linear systems. Let's start with Eq. (10) of https://t.co/xXl4OMzmIy which, for any 3 variables, X,Y,W gives: 1/5

2023-01-30 01:00:00 CAFIAC FIX

2023-01-17 00:09:01 @NandoDF @eliasbareinboim Given that my neurons are enslaved to other neurons, I, too, lack agency. Still, I have the illusion of agency, and that is what we need to fix: The illusion of reality existing up there, and of us operating upon reality through what we call "our model of reality".

2023-01-17 00:01:41 @coecke @NandoDF @eliasbareinboim I am a gambler too, but, in science a gamble only after seeing a couple of toy problems.

2023-01-16 23:51:09 Finkelstein views have changed because he, like most Western observers, had it all wrong thinking the obstacle to 2-state is Israel, not Palestinian rejectionism of the very idea of Israel's sovereignty. He seems awfully disappointed to see Israel refusing its death - poor man! https://t.co/DsK2uPCZP9

2023-01-16 23:25:00 @bhack_10 @GaryMarcus were going to win. Should he buy it, even though he does not know which candidate will win?". I got a lecture on why property buyers should consider location, not elections. Once a question touches on a domain in which Chatti has some knowledge, She forgets logic. 2/2

2023-01-16 23:16:50 @bhack_10 @GaryMarcus It has a problem instantiating general principle. I tried Savage's original question: "A businessman decides that he should buy a piece of property if he knew that the Democratic candidate were going to win. Similarly, he should buy if he knew that the Republican candidate 1/2

2023-01-16 22:47:59 @bhack_10 @GaryMarcus Where did she learn about the Sure Thing Principle? Did you teach her? Nice.

2023-01-16 22:01:07 Commemorating MLK Day, Tablet Magazine reprinted a speech by Ron Dermer which resonates with me as a good reminder of what MLK means to me, as an American and as an Israeli. Sharing: https://t.co/SarXK9JfEz

2023-01-16 12:54:47 @stephensenn @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone @HL327 The crop is a mixture of varieties, but the experimenter does not know it.

2023-01-16 12:50:13 @jon_mellon @GaryMarcus Interesting. It does not realize that "either E or its negation are not true" is impossible. It treats "not-E" as the name of a totally unrelated event.

2023-01-16 10:49:07 @ShehroseMia @GaryMarcus Would love to learn what I've missed, but I don't see how the 'problem of induction" is related. Eager to learn.

2023-01-16 10:45:42 @BRSLWP @GaryMarcus In other words, I believe in the "Causal Sure Thing Principle" though Savage's Principle has a flaw: We need to state that the action does not affect the probability of E.

2023-01-16 10:37:32 @BRSLWP @GaryMarcus I do have doubts about the sure thing principle, but not because of "other conditions". I describe them here: https://t.co/ODjYZuaAFZ

2023-01-16 10:31:58 @stephensenn @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone @HL327 Fine. But what if you we have "latent heterogeneity", namely, if we have two varieties and we cannot tell whether a given lot is variety-1 or variety-2. How would variance help us tell that we have two, rather than one variety.

2023-01-16 10:16:33 that same action if I knew nothing about E.?" CHAT answered: "It depends on other conditions that might make the action unsafe, it you do not know about E". Even when I proved to her that not-E is safe, she continued to claim: "It may depend on other conditions..." @GaryMarcus

2023-01-16 10:16:32 Strange encounter

2023-01-16 09:57:41 @erikbiz @VP This has been the case in the past 30 years. Whenever a journalist argued: Didn't Arafat recognize Israel in 1993? I would say: Don't listen to Arafat, Listen to what my Palestinian colleagues say, at UCLA and at all other universities.

2023-01-16 09:19:09 @stephensenn @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone @HL327 Fisher is obviously concerned with heterogeneity. Is it you interpretation that he recommend detecting heterogeneity using the variance? If so, how?. At any rate, the "Latent Heterogeneity Inequality" of https://t.co/lDnscocTRz Eq. (5) does not invoke variance estimation.

2023-01-16 08:53:34 @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone You probably mean: Simpson's "reversal" disappears with backdoor conditioning. Conditioning requires some non-missing measurements, which you can identify from the causal model, and you can't otherwise. The reversal only harm decision making if you ignore what CI tells us.

2023-01-16 08:43:21 Thanks for sharing. I was not aware of MLK letter of September 1967. It is not the one usually quoted which states "anti-Zionism = antisemitism"

2023-01-16 01:27:07 JAY BHATTACHARYA is a co-discoverer of "Bias Amplification" in econometrics https://t.co/FB8td3nuN0 I would take his article about Stanford seriously.

2023-01-16 01:27:06 https://t.co/dTFhv7B4Jc

2023-01-16 01:01:56 @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone @stephensenn @HL327 @HL327 asked and deleted. My answer: We cannot distinguish carriers from non-carriers when we select patients for the RCT, i.e., before treatment. We call it "latent heterogeneity", and we also have a formula for detecting its presence: https://t.co/lDnscocTRz Eq. (5), page 9.

2023-01-16 00:38:11 @Undercoverhist @causalinf @jakewertz @Andrew___Baker @gelbach @CFCamerer @EricChyn @raulpacheco @mosenkis @alan_krueger @Econ_Sandy Relatedly, I've briefly summarized the history of causality in econometric here: https://t.co/GIaEtSZ9Mb (p.3). Do you think I've skipped a contribution that is relevant to causal analysis today?

2023-01-15 22:12:26 Ken Roth’s Main Contribution is to Antisemitism, not Human Rights https://t.co/VGPRbeiQrf via @jewishjournal

2023-01-15 21:16:48 Perfect timing - perfect song! Yesterday 80K Israelis went to the streets, demanding to examine their face in the mirror. "I have no other country" says this song, https://t.co/teL05tuBPH, "even when she changes its face - NO! I ain't going to give up".

2023-01-15 20:44:19 @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone @stephensenn @HL327 calculates bounds on P(harm| man). The bounds become more informative the greater the difference between carriers and non-carriers under free choice, even when no difference at all is detected under RCT regime. The visual plot delivers a bound for every data combination. 3/3

2023-01-15 20:30:44 @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone @stephensenn @HL327 when carrier men are exposed to a certain medication. (2) Carriers have greater craving for the drug than non-carriers, when given free choice, so they are more likely to buy if available. No additional assumptions. For any numerical assignment of proportions, our paper 2/3

2023-01-15 20:23:42 @f2harrell @AndrewDGarrett @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone @stephensenn @HL327 The datamethods has hard time giving me access. So, here is my Twitter version of a hypothetical and realistic example. (1) A variant of a certain allele, when present on the Y chromosome (i.e., it only has the potential to occur in men), causes a fatal adverse drug reaction 1/2

2023-01-15 19:51:53 @Undercoverhist @causalinf @jakewertz @Andrew___Baker @gelbach @CFCamerer @EricChyn @raulpacheco @mosenkis @alan_krueger @Econ_Sandy I bring up Hoover's "Lost Causes" often, e.g., whenever Heckman tells me the econometrics is the Mecca of causal inference.

2023-01-15 19:15:47 I don't care if he wines and dines with oligarchs, but can he teach anything meaningfully at Harvard given his record? Would students listen to a man who vilifies Israel over Iran 65:1 ? https://t.co/64WA3qbSnG

2023-01-15 03:39:28 There is still an unsolved puzzle: What was it in his traumatic upbringing that made him so pathetic and irrational, despite his intellectual potentials? https://t.co/hwB0dZXFdX

2023-01-14 21:17:39 @vk_wilde Wikipedia offers a fairly good explanation for the surprise part of Simpson's Paradox https://t.co/IE5HTB9HHp

2023-01-14 11:25:49 My wildest dream, @VP Kamala Harris opening her speech with: "You want to know why I support Israel? Look at her neighbor's mentality!" https://t.co/KruQotioC6

2023-01-14 11:11:20 RT @vardi: https://t.co/B3UNr9Ld5G

2023-01-14 10:59:18 Jackals think their howls sound like "uproar". Some even think their howls are "criticism". Students prefer logic over howls. @GeraldNGOM @Ostrov_A @elderofziyon https://t.co/vNsgo9AJnW

2023-01-14 10:02:51 @AndrewDGarrett @f2harrell @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone Any Simpson's reversal observed in RCT is due entirely to small sample anomalies, bound to disappear as sample size increases. The proof is in the causal sure-thing principle https://t.co/ODjYZuaAFZ

2023-01-13 22:07:33 @kaulcsmc @AndrewDGarrett @f2harrell @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @GreggWStone We must be talking about a different "Simpson's Paradox", because the one I know is "provably impossible" in RCT. Is the concept of "provably impossible" part of the conversation in your field?

2023-01-13 21:53:42 @AndrewDGarrett @f2harrell @elmir1omerovic @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone It's not a matter of 'protection", nor of "hard to find"

2023-01-13 03:09:00 Media Breathe New Life Into Ken Roth’s Twisted Anti-Israel Conspiracies https://t.co/pKcxTufBlh

2023-01-12 22:07:05 This chart should accompany every discussion of @KenRoth whining. Harvard's Dean needs no pressure from donors to see that, with this mental aberration, the guy isn't fit to teach in academia

2023-01-12 12:59:02 Mega-Zionophobe Abdulhadi is in the news again and MESA, the hijacked Middle East Studies Association is behind it, making a mockery of its academic mission and celebrating its politicization. Glad I was able to contribute a comment.

2023-01-12 12:59:01 SFSU Prof Abdulhadi Wins Middle East Studies Award https://t.co/ewe1xwc0Pz via @jewishjournal

2023-01-12 12:22:52 @KenRoth, Data do not support your wishful victimhood. On the contrary, demonizing Israel is today a #1 prerequisite for any academic position in any ME-related department. https://t.co/C3hww06btC

2023-01-12 11:56:23 We've gotten used to PLO double talk, and learned to tolerate it with a smile. What I haven't been able to tolerate is US pretending to believe it, instead of demanding: Stop the BS! If you want 2-states, say so in Arabic, on TV, tomorrow: "Tel Aviv is not a stolen land!" Louder! https://t.co/ndYIOJrUCv

2023-01-12 11:41:19 RT @blakeflayton: Chaim Weizmann was once asked "why you do Jews insist on Palestine when there are so many underdeveloped countries you co…

2023-01-12 08:25:28 @leticiakawano As I see it, the disagreements converge on one issue: Is a drug that kills 10% and saves 10% clinically equivalent to a sugar tablet, which has no effect at all ? Moreover, assuming they are, is there enough scientific interest in the difference to try detective experiments?

2023-01-12 08:02:59 For readers who saw Nick Huntington-Klein's review of #Bookofwhy: https://t.co/0DFAoQ1v2D and asked what I have to say about it, I've now posted my comments here: https://t.co/JjKMK4VPzl and sent it to Journal of Economic Methodology. Please alert me to errors and omissions.

2023-01-12 07:37:50 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone To name the goal, or to name the means -- this is the question. This is another cultural difference I noticed yesterday. In CI we give names to what we seek to estimate

2023-01-12 07:26:59 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone In this paper you talk about "covariate-specific effects" while your tweet talks about "patient-specific effects". Are they the same? If so, to eliminate confusion, why not decorate it with a formula, e.g., E(Y|do(x)|c) or E(Y(1)-Y(0)|c) or whatever.

2023-01-12 04:24:35 @stolev75 @IlhanMN If you are Jewish, you probably have the antennas to distinguish "disagree with another country's politics" from "disagree with another country's existence". Where do you think @IlhanMN belongs? Would you nominate someone who "disagrees with Ukraine's right to exist?" Or Canada?

2023-01-12 03:20:53 @yajnadevam @IlhanMN True, with one nuance: people whose perception of their "nation's interest" is distorted by a non-inclusiveness disease.

2023-01-12 03:08:22 The loudest Gorillas are always the first to decry "attempt to silence." https://t.co/y1vXZtgizf

2023-01-12 02:49:50 @blakeflayton @StudiesZionist There is always a spec of truth in what @PeterBeinart is spitting. If he is Jewish and if most Zionists are anti-Beinhart, then it follows that most Zionists are antisemitic. Hmmm.. but is he really Jewish? Not according to the Amidah: "The slanderers will find no hope" (12th B)

2023-01-12 02:11:51 @AbdulahiNur2022 Thanks for telling us about this genocide, of which I knew practically nothing. Still, I wouldn't hold @IlhanMN responsible for what her compatriots have done in the past. I am more concerned about future genocides she helps ferment in the Middle East.

2023-01-12 00:31:42 Here is what I've learned last night. What's the difference between CI folks and Trialists? Ans. The former view RCT as an "intervention" on natural behavior. The latter view natural behavior as an RCT marred by non-compliance. Will share more of what I learn. https://t.co/v2LsarQv7O

2023-01-11 23:40:33 @NatsforDocs @leticiakawano @GuyattGH Anyone who "Hadn't heard of it until now" should get a medal of resilience. I wish I could send you a signed copy -- enjoy.

2023-01-11 23:27:56 University of Chicago Students for Justice in Palestine to Boycott Course Taught by Former IDF General https://t.co/c8Mzyz3hcB

2023-01-11 16:17:24 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone I am still perplexed by your: "RCTs provide model-based patient-specific estimates allowing for rich conditioning". Is your patient-specific estimate similar to our PNS ? Or is it a new species? Does it have a mathematical description? A counterfactual formula? Perplexed.

2023-01-11 16:00:51 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone validity of the premises, which are really mild and hardly disagreeable. You seem to suggest that a drug that kills 10% and saves 10% is clinically equivalent to a sugar tablet, with no effect at all - Not to me! Still, it's helpful to know where we disagree. 2/2

2023-01-11 15:46:27 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone I didn't realize our paper was discussed at such length at discourse. Great, I'm learning a lot. You say:"The key premise is that you can learn from unmeasured variables". I see it as a conclusion, not a premise. You object to the usefulness of the conclusions, not to the 1/2

2023-01-11 09:37:11 @stephensenn @elmir1omerovic @AndrewDGarrett @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone Can we express this sentence in mathematics: "similar patients given identical treatments will have different values in different studies". The assumption overriding counterfactuals is that Y(1, u) and Y(0,u) exist, and are immutable properties of u (the patient). Is it wrong?

2023-01-11 09:29:17 @stephensenn @elmir1omerovic @AndrewDGarrett @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone The problem posed in Example 1 has a complete solution, explicated here https://t.co/ty6fgbzycF The mathematics of do-calculus is what gives me the confidence in the validity of the solution. Please try to tell us if your method agrees with mine, rather than attacking &

2023-01-11 09:16:23 @stephensenn @elmir1omerovic @AndrewDGarrett @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone Sorry if I have given the impression of refusing. My inability to examine your solutions stems from my inability to recognize their relations to the concrete problems under discussion. Example: I am familiar with "causal effects", but what is "study effects"?

2023-01-11 09:00:46 Rep Ilhan Omar @IlhanMN is finally out, almost 4 years after I pleaded with Nancy Pelosi to take her off the US foreign Affairs Committee https://t.co/zyBilhd6uG, a position she was the most unfit person to serve. I wish Speaker Pelosi would have done it, not McCarthy. https://t.co/zzsxB9X462

2023-01-11 03:52:59 In case some of our readers are looking for book recommendations from influential people, I'm notified by https://t.co/nPJwwUNMBC that #Bookofwhy has been listed on their website: https://t.co/cIhkr2d3qz I would like to believe that our readers have discovered it 4 years ago.

2023-01-11 03:21:04 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone I've ventured to look into Hauck etal (1998) https://t.co/AnOoZEAwKY and, as much as I hate to sound critical, this paper is pre-causal (eg, even "confounding" is wrongly defined). For a complete nonlinear analysis of the adjustment problem posed, see https://t.co/f2oN7JW60N

2023-01-11 02:41:14 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone Please note that in the example submitted https://t.co/hzc0FAgeQ0 the randomization is on different variable, not the one whose causal effect is needed.

2023-01-11 02:12:38 @leonpalafox @GaryMarcus With one tiny nuance - I don't have "views on causation" , all I have are a few mathematical proofs of what can or cannot be done with and without causation.

2023-01-11 01:59:29 @stephensenn @elmir1omerovic @AndrewDGarrett @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone I've noticed that, invariably, you end your tweets with what's wrong with my solutions. Can you try to end them with the way you would solve the concrete problems posed? Please try.

2023-01-10 21:44:13 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone Do you mean you can prove that Crossover RCTs provide unbiased estimate of CATE, without measuring (C), and without any model of how the two crossed over RCTs are related to each other? Puzzled.

2023-01-10 21:03:01 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone Agree. Instead of ATE, I should have written CATE, where C stand for "conditional" on a set of baseline characteristics. Is that what you mean by "patient specific estimates"? Or do you mean P(benefit| Characteristics)? The latter requires OS, explained in https://t.co/rGHIc93DYO

2023-01-10 18:28:59 @f2harrell @elmir1omerovic @AndrewDGarrett @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone If an adequately sized RCT is available, and if all you want is ATE, no need to use OS. What we ask is whether OS can be useful when the RCT is NOT adequately sized, or when we want more than just ATE. Any advice?

2023-01-10 17:33:42 @VickersBiostats I am relying on MA experts to tell us if the Cochrane guidelines agree with the rigorous methods developed in CI, but I can't seem to get a straight answer, because we do not have a common language in which the heterogeneity is represented. Let's wait for it to be developed.

2023-01-10 11:46:36 Important message to campus administrators.

2023-01-10 11:46:35 Biden is failing to deliver in the fight against antisemitism | Opinion https://t.co/pXl05bsjAl

2023-01-10 11:45:54 RT @SAFECUNY: Extremely important explanation by @Klmarcus of the current status of Title VI as it applies to Jews on campus and what is, h…

2023-01-10 11:28:24 RT @JGreenblattADL: This article concocts a conspiracy theory about the @Kennedy_School and the military-industrial complex and implicates…

2023-01-10 11:26:16 @MehdiHasanShow @KenRoth @mehdirhasan I was wondering how long it would take for @KenRoth to reach the bottom - @MehdiHasanShow - in his "they did it to me" campaign. I heard Harvard "Fellows" are expected to generate new ideas .

2023-01-10 10:28:47 @elmir1omerovic @AndrewDGarrett @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone What assumptions are needed to facilitate a solution to our problem? Put another way, how can we be sure that, once we make those assumptions, we can pool data from both studies and construct an (consistent) estimate of our target effect. (From https://t.co/B0TKO64IzR 2/2

2023-01-10 10:24:44 @elmir1omerovic @AndrewDGarrett @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone I'd like to be concrete. Example 1: Suppose we wish to estimate the causal effect of X on Y, and we have two diverse sources of data: (1) an RCT in which Z, not X, is randomized, &

2023-01-10 10:16:06 @AvivaKlompas The Zionist dictum is: Rights are handed not by genetic lineage, nor by past dwellings, but by the present state of mind.

2023-01-10 09:59:37 @Ostrov_A They should take @PeterBeinart with them. Families that pray together stay together.

2023-01-10 09:55:04 I love the lyrics: "We (westerners) must do everything to avoid provoke them (Pns) who have no self-control." And the more we do, the more they will convince us that they have none. https://t.co/qhsLP3vQ0P

2023-01-10 09:39:20 @AndrewDGarrett @elmir1omerovic @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone Starting with design, instead of the research question, is a sure way to give up on ever affirming the validity of the design.

2023-01-10 09:32:28 Would this guy speak that way, had we charged him with Zionophobia instead of antisemitism? Would he say: "I am really concerned about people cheapening Zionophobia, which is a real threat these days?" Would you, @KenRoth? Not too late. You might be forgiven at heaven's gate. https://t.co/22QAS5LSjk

2023-01-10 09:09:33 @AndrewDGarrett @elmir1omerovic @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @drjohnm @kaulcsmc @GreggWStone And to me, the relevant Q why do we submit such questions to debates, views or opinions, when we have the mathematics to answer them?

2023-01-10 04:39:45 For my old comrades in non-monotonic reasoning, I just tried our favorite Tweety: "Tweety is a bird, and birds fly. Does Tweety fly?" And got back: "It is not possible for me to determine whether or not Tweety, a hypothetical bird, is capable of flying without more information."

2023-01-10 04:19:54 How many of us, AI-ers, remember the famous "Yale Shooting Problem" (1985)? CHATOPENAI just solved it: "Fred is alive and my gun is unloaded. I load the gun, wait a moment, and shoot. What will happen to Fred"? Well, after lecturing me on the evils of shooting, 1/2

2023-01-10 00:54:46 It's 2023 already, and the flow of articles on causal inference continues as if it was 2022 on steroids: https://t.co/ouOHT3wpVm My attention zooms onto the paper on Cluster DAGs, which I hope to read soon.

2023-01-09 21:27:40 History books, 2033: Kenneth Roth @KenRoth - The man who gave #HumanRights its bad name. https://t.co/GzIJ6Gew25

2023-01-09 21:17:00 @VickersBiostats I am the last to look for causal notation. I am merely looking for that extra-statistical notation that your guidelines find necessary when we "investigate heterogeneous sources." As a computer scientist I know that one can cheat with hand waving, but not with notation.

2023-01-09 21:08:19 @VickersBiostats Your guidelines, which "investigate sources" in an extra-statistical language, may be a singular and welcome revolution in MA history. Still, my analogy to apples-oranges-bananas is valid to the pre-guidelines MA literature and practice as it existed till 2012 (analogy made).

2023-01-09 05:16:44 RT @KhaledAbuToameh: Report: The commander of Lions' Den has turned himsef in to the Palestinian Authority security forces. القدس|| قائد "…

2023-01-09 00:21:31 @DrAmbuj @VickersBiostats As long as you are aware of the fact that I am speaking from whimsical "opinions", but from a mathematical framework, and that such a framework exists which successfully answers questions about combining heterogeneous data sources.

2023-01-08 22:14:06 @VickersBiostats MA is a great methodology, when the sources are homogeneous. When they are not, MA needs to take heterogeneity into account, and you can't do it with ordinary statistical vocabulary, because the heterogeneity is in the underlying causal effects, not merely in the distributions.

2023-01-08 22:08:57 @VickersBiostats There is no need to "argue" that MA (as practiced) is "causality free". All one has to do is to look at the notation used in MA papers

2023-01-08 21:04:54 @DrAmbuj @VickersBiostats How about mathematical proofs?

2023-01-08 20:59:22 @VickersBiostats Why? Is conducting a wrong analysis a prerequisite to explaining why it is wrong? Let's talk science

2023-01-08 14:42:11 @GreggDCaruso Too bad they do not detail the actual experiment.

2023-01-08 13:50:25 RT @Israel: Let it Flow: Last week, Israel began streaming desalinated water from the Mediterranean Sea to the Sea of Galilee (Kineret) to…

2023-01-08 13:50:00 RT @Kasparov63: Joke going around with this photo is that Putin was completely alone in church on Christmas because even God is with Ukrain…

2023-01-08 13:43:10 @GeraldNGOM @AviKaner @KenRoth @Harvard @Kennedy_School @NGOmonitor @hrw Such psychological analyses are extremely important to making @kenroth deformity believable to people who are impressed by his HR title, and haven't traced his anti-Israel pathology. Please share.

2023-01-08 13:34:00 When someone you love is in suicidal danger, you draw closer, and pay attention. https://t.co/PgfjFFNvXa

2023-01-08 13:02:16 @AviKaner @GeraldNGOM @KenRoth @Harvard @Kennedy_School @NGOmonitor The Guardian, @hrw and @KenRoth pretend they never heard of the difference between "criticizing Israel" and "deep seated hostility to Israel". They claim @Kennedy_School is bowing to donors pressure, instead of upholding a moral principle: Zionophobic racism is still racism.

2023-01-08 11:51:13 @AleksanderMolak It's a terrible weakness of mine, easily dragged into methodological discussions that I thought were resolved two decades ago, with the publication of Causality https://t.co/LjzDCLZgzg I simply cannot see alchemy being taught and practiced in the age of causation. Time to quit.

2023-01-08 11:33:58 @michelnivard @PWGTennant @epigenci @EpiEllie @SachaEpskamp We have a language for making heterogeneities explicit. It's called "causal SEM" (or SCM). It's rigorous, friendly and provably complete. Why avoid it? Why resort to alternatives that fail on all three criteria? But don't let me discourage you, I've made enough enemies that way.

2023-01-08 11:21:53 @BassamMusaed @RepRashida Three Zionophobic outfits regurgitating each other accusations doesn't make them see reality better than three blind men touching a fake elephant.

2023-01-08 11:11:48 @BroBrusque @RepRashida @Ilhan Who said they're antisemitic - God forbid. They are merely Zionophobic bigots, a more dangerous form of genocidal racism.

2023-01-08 11:03:13 @michelnivard @PWGTennant @EpiEllie @SachaEpskamp You cannot write RCT studies on heterogeneous population as non-causal SEMs, because the heterogeneity is causal, not statistical. We demonstrate it visibly in https://t.co/dEPwcv3l40.

2023-01-08 10:47:12 Was I too harsh in saying that: "Meta analysis tries to average apples and oranges to estimate properties of bananas?" https://t.co/0crNZQbOuE

2023-01-08 10:43:32 @michelnivard @PWGTennant @EpiEllie @SachaEpskamp Non-causal SEM's, be they "meta" or not, cannot capture the invariants among models. So I'll vote "no", while remaining open to counter demonstrations (rigorous only, please).

2023-01-08 10:38:05 @epigenci @michelnivard @PWGTennant @EpiEllie @SachaEpskamp The problem surfaces when the two estimands are different. Has meta-analysis matured to incorporate causal modeling?

2023-01-08 10:35:17 @tkaiser_science I have several such articles, https://t.co/dEPwcv3l40, https://t.co/E1ap1MNWAh, and more, but I'm not sure everyone will consider them "practically demonstrating"

2023-01-08 00:18:28 Unknown to priests and critics of RCTs, Causal Inference has gone beyond analogies and metaphors. Deaton &

2023-01-07 23:51:21 @rohitpojha @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone There is more to combining studies than running them poorly. You can have a guru for RCTs and a guru for OS, each conducting a perfect study, and yet, if they don't know how to combine them properly (e.g.,https://t.co/E1ap1MNWAh), they are likely to fumble.

2023-01-07 18:13:19 @f2harrell @dredgardorsey @RWJE_BA @ArgoloF @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone I assume what @f2harrell meant by this post is: If you do not know how to properly combine data from experimental and observational studies, you're better off not combining them.

2023-01-07 01:45:57 @elmir1omerovic @f2harrell @ZscheischlerJak @_MiguelHernan It's nice to see a team of 19 knowledgeable researchers, finally sorting out various approaches, old and new, to causal inference from time-series data.

2023-01-07 01:30:42 @deboerk07 @nephologue I share my difficulty to accept that "diagrams are harder to create than equations" - Why?

2023-01-06 13:01:19 @NowakPosadzy @CavaliereGiu Why economist are scared of graph? Clear to me: Because (1) They are unfamiliar w/ graphs, (2) They are told graphs are ad hock, or not needed, (3) They won't risk tenure (4) They havn't been shown the cost of following their leaders. Algebraic more efficient? Try Appendix A2.

2023-01-06 10:24:01 @NowakPosadzy Naïve question: Do you think any graph-avoiding framework deserves the title "framework"? Before you say "yes", please try to solve the Econ-101 problems presented in https://t.co/PptTCL1AM2, Appendix A2, using such a "framework". Eager to hear. @CavaliereGiu, #EconTwitter

2023-01-06 07:29:51 @CavaliereGiu @heckmanequation @UChicago @UCLA Those in #EconTwitter Interested in the #econometrics of and , might also be interested in the principles of causal inference as applied to #econometrics. Check out this "Year in Review" https://t.co/HQvCVpz52w.

2023-01-06 07:11:13 I comment on Heckman's new twist in my "Year in Review" https://t.co/j1nbgXY3vy (which is linked to 2 relevant critics: https://t.co/UGqywWLheF &

2023-01-06 06:17:52 @Ostrov_A Politicians never say things that everyone knows. To be a politician you say: We oppose any act that may POTENTIALLY change the status quo over #TempleMount. The word "potentially" makes people think you are visionary.

2023-01-05 23:56:03 Who said physics is observant-dependent? https://t.co/do2vXkugn6

2023-01-05 18:31:00 The Jewish path back to Israeli democracy https://t.co/3H3yLDfODT

2023-01-05 17:34:37 @ccaballeroh10 Hard to believe! I'll try to find out from the publisher.

2023-01-05 03:26:39 I've now finished and posted a "Causal Inference (CI) - Year in Review", https://t.co/j1nbgXY3vy, as seen through a highly biased causal lens. Wishing all readers a happy and productive 2023.

2023-01-05 02:52:21 @Ostrov_A @EUSR_Koopmans @EUSR_Koopmans , This can easily be verified by linguistic analysis: How many times does the word "Jewish" appear in EU's documents dealing with the holy sites. Please do the counting and tell us about "equilibrium between the major religions in #Jerusalem".

2023-01-05 02:31:18 @blakeflayton I can't think of a stronger Zionist-promoting argument than @rad_planne's "safe spaces." He should only continue to tell us how those spaces should defend themselves against neighbors claiming exclusive ownership over all "spaces", especially those created by "immigrants".

2023-01-05 01:12:59 @noah_greifer @soboleffspaces @causalinf @EpiEllie @bveg3 What this paper calls ITE is not "individual treatment effect" but the average treatment effect in the population resembling an individual. See https://t.co/rGHIc93DYO for formal distinction.

2023-01-04 16:59:00 @think___y @ari_bronstein @Noahpinion Not a bad book.

2023-01-04 10:21:21 Working on "CI - year in review" takes me longer than I thought

2023-01-04 09:21:02 As we say in #Bookofwhy, the brain is a big Why Machine -- it's the New Science of Cause and Effect. https://t.co/gOOCM6A7Jn https://t.co/WUuvjZI1oM

2023-01-04 09:13:39 @GeraldNGOM @nytimes Give @nytimes credit for not using "stormed" or "raided" or "molested".

2023-01-04 09:08:48 Good point: Scared of Hamas, Ben-Gvir had to sneak onto the Temple Mount like a thief, rather than one who has historical right to be there, like all other parishioners. https://t.co/XUJtcuNVTl

2023-01-04 08:56:38 RT @NickMinock: “If people are unwilling to stand up to those who are crying out for equity, then yes, America is going the wrong way,” bus…

2023-01-04 08:50:02 Justified mockery of "Machine Learning" - agree. But recall, AI has also spawned "Causal Inference", where the notions of "learning" and "understanding" are quite rigorously defined

2023-01-04 07:54:18 Is "storming" less poetic than "raping"? https://t.co/AOLoeBWeDT

2023-01-03 14:07:44 @erikbiz @richard_landes @mosaicmag @JeffreyHerf @DavidSuissaJJ There is one weakness in this argument. It is not an extrapolation from past concession+ violence to the future that counts, but today's intentions, as articulated loudly and publicly.

2023-01-03 08:31:17 @stephensenn I am curious about paradoxes, not about designs. Lord's story remains paradoxical in the design I've described, and this is what I have set out to resolve (and succeeded). Different designs may raise additional question, but not the one that baffled Fredrick Lord in 1967.

2023-01-03 08:18:25 RT @richard_landes: The Failure to Understand the Jihad against Israel Led to the Failure to Understand the Jihad against the West https://…

2023-01-03 08:06:47 First time I feel true Homo-Sapien kinship to my brethren in Micronesia. https://t.co/gPGAatRUJS

2023-01-03 07:53:14 RT @HillelNeuer: BREAKING: The 2023 U.N. Human Rights Council is a travesty of justice and a farce. https://t.co/Xb2v6FhCcD

2023-01-03 01:51:48 @benj_kerstein @JNS_org The new paradigm need not be crafted from scratch, it was already articulated by Herzl in the First Zionist Congress (1897): "Zionism is a homecoming to the Jewish fold even before it becomes a homecoming to the Jewish land." See https://t.co/ICqKCByB2p @newzionists

2023-01-03 01:44:01 @stephensenn Naturally, if we do not ensure that each of our "designs" estimates the desired estimand, it might well happen that two "designs" give different answers

2023-01-03 01:27:07 @stephensenn Moreover, just in case you want to see a statistical estimand, it is give in https://t.co/MSJJViHxbM by the equation: P(Y|do(Diet)) = ∑_Wi P(Y|Diet,WI) P(WI) I wouldn't trust any "design" which does not estimate the rhs of this equation, would you?

2023-01-03 01:15:16 @stephensenn BOW p216 mentions clearly the estimand: "Lord’s paradox now surfaces with greater clarity, since the query is well defined as the effect of diet on gain". As to "designs", unless one shows that a design produces the desired estimand, speaking of "designs" is a super-red-herring.

2023-01-02 12:49:25 I feel for these photographers/clowns - they travelled thousands of miles expecting to see genuine Israeli oppressors in action - no refund? https://t.co/6vNAL1oCp0

2023-01-01 21:57:04 @m4xim1l1an @Spinozasrose @vonderleyen @Ostrov_A @EinatWilf @FranceskAlbs This is not the song I've been hearing in the past 86 years, and I have a fairly musical ear. From one side I hear:" Share, Share, Share", and from the other: "Not even the size of a postage stamp." I am still waiting to hear two words from @FranceskAlbs: "Equally Indigenous".

2023-01-01 13:39:41 I was waiting for a piece of good news before wishing readers Happy New Year. Let's hope 2023 brings us an inch closer to a more humane world, and may each of us have a part in that inch. https://t.co/2VQxw2LVWG

2023-01-01 08:41:35 @FrankPILA @FranceskAlbs @netanyahu @Israel Had he conditioned it on " its neighbors being peaceful," as Lapid did, Israel would have maintained its high moral grounds with no risk at all, because the condition negates the collective identity of Israel's neighbors.

2023-01-01 05:09:55 Protesting this outrage will not yield results unless we can explain: What makes Palestinian propaganda so effective? What makes Israeli arguments so ineffective? Can anything in the latter be changed to counter the former? https://t.co/AQ0tFEty54

2023-01-01 04:38:03 have a right to a homeland? Any right at all? I think you should disqualify yourself until you give this question some thought, or until you can utter the words: "Equally Indigenous". Try! @vonderleyen @Ostrov_A @EinatWilf 2/2

2023-01-01 04:38:02 True, Netanyahu should not have used the word "exclusive", and he is probably sorry that he did. But you, @FranceskAlbs, you came up with same garbage dumping artillery when former PM Lapid spoke at the UN and reaffirmed Israel commitment for a two-state solution. Do Israelis 1/2 https://t.co/OwnyEsRom9

2023-01-01 04:15:06 The best counterattack is to expose such rot one story at a time, but I wouldn't give up so quickly on progressives - some are still guided by commonsense. https://t.co/6sgS97I7n2

2022-12-31 23:54:29 A most informative and insightful analysis of the new Israeli government by @rosnersdomain -- a frank look at reality shining through the clouds of the dooms day alarmists and let's "lets wait and see" ostriches. https://t.co/91sKRZYMFE

2022-12-31 22:21:28 RT @tdietterich: @EzraJNewman 1. A model of the world, both the current state (the things it believes are true) and the dynamics (the thing…

2022-12-31 12:02:56 @policytensor @raghav50k @DmitriGallow I like your exposition of these methods. Q. What success (or obstacles) have you encountered among traditional regression analysts? I consider it an embarrassing personal failure to see so many statisticians and economists in 2022, unaware of Theorems 1 and 2.

2022-12-31 11:38:44 summarizing CI progress and obstacles in 2022. In the process, I've come across an encouraging Google-Scholar report https://t.co/5mxQvEWiOS showing an all time record of 10,2K citations in 2022, and positive derivatives in all CI categories. I have no doubt this trend will 2/3

2022-12-31 11:38:43 Here is our last collection of 2022 articles on causal inference : https://t.co/tXmrU9eFMT and https://t.co/vCF9XHT93t. Note the increasing breadth of applications, coupled with methodological penetration into skeptical fields. I am writing a "year in review" report, 1/3

2022-12-31 10:23:26 Glad I've helped @SSI_Movement exceed its fund-raising goal. SSI is the only Israeli-supporting student organization in the US that has a faculty advisory board. Even @HillelIntl does not have one, and it shows painfully on my campus, UCLA. @RozRothstein @eishsadehy @MiriamElman https://t.co/AYaUPPNjRk

2022-12-31 10:02:30 And we thought public health would be shielded from anti-Israel tantrums posed as health research. Not so on the pages of the NYT, where we now find deceitful guest essays starting with: "As a researcher, I study...", https://t.co/6ZjDkayDdb

2022-12-31 09:27:18 RT @AsraNomani: MAMA BEARS + 2 dads marched @fcpsnews today with DEMANDS. Harry, Srilekha, I met impromptu with @fcpssupt. We said: #SaveMe…

2022-12-31 09:06:01 RT @blakeflayton: If you want a lesson in how to criticize Israel without being antisemitic, take a look around the country at the thousand…

2022-12-31 05:56:11 Settlements are an obstacle to a "two state solution" only when at least ONE Palestinian is willing to accept Israel in some borders. Since the latter is an unrealistic proposition, and since Biden knows it, we should not expect major diplomatic clashes on this issue. https://t.co/w2qAwRQKYV

2022-12-30 23:20:54 @AdamMilstein The NYT will continue to assume that Sheikh Yousef's mentality is not representative of Palestinian's mindset, which explains why their editors have almost ZERO understanding of the Middle East conflict, and I have been reading their editorials for more than 20 years.

2022-12-30 10:43:22 @AdanZBecerra1 @stephensenn I’d be delighted

2022-12-30 09:38:31 RT @GeorgeDeek: Congratulations and good luck to the first Azerbaijani ambassador to Israel! https://t.co/nYAeUIbERj

2022-12-29 10:56:06 The bds has never aimed to boycott israel

2022-12-29 10:27:32 Kudos to Paul, for propagating the science of causality in an era where the scientific method itself is under attack. https://t.co/bdyCr08mD8

2022-12-29 10:15:18 RT @MiriamElman: When a virulently #antisemitic academic (who’s also a deported leader of a terror org! ) is invited to lecture, universit…

2022-12-29 10:14:11 RT @bulutuzay_: Turkish Cypriot publisher takes Turkey to ECHR I’ve had the honour of interviewing Levent twice.The Turkish government ha…

2022-12-29 09:22:54 @RealSarahIdan @EinatWilf

2022-12-29 09:13:17 RT @Ostrov_A: Thank you @EpochTimes for covering our claim with @GabeGroisman against UC Berkeley Law, over their antisemitism and discrimi…

2022-12-29 09:03:20 RT @FreyaLee5806: @AsraNomani you are amazing!

2022-12-29 09:02:44 RT @bennyjohnson: Indian mom leaves Democrats in stunned SILENCE during testimony after single-handedly DESTROYING CRT— ROASTS AOC htt…

2022-12-29 08:51:41 What uncle Ira tried to tell the Shmegegge was that Palestinian existence is mighty peculiar - no other nation defines itself by negating the existence of another https://t.co/I768eNC6sk

2022-12-29 08:06:48 RT @JooTube: Actor, @MikeBurstyn, recreated #Israel's formal founding via UN consensus to establish a #Jewish State in #Palestine 1947 http…

2022-12-29 08:01:13 @algobaker @mrexits Ill be glad to send one

2022-12-28 10:49:28 RT @GeraldNGOM: Must read: @NYTimes' reporting on Israel must be ignored - Israel’s OK as long as it doesn’t embarrass, doesn’t inconvenien…

2022-12-27 05:38:21 https://t.co/XX2nsHMVG0

2022-12-26 06:51:47 @Claire_V0ltaire I've posed this question to pro-Pal colleagues. The answer: Goliath was a martyr

2022-12-26 04:30:18 The diet-weight example is described in #Bookofwhy pages 211-7, https://t.co/38QW0mKGiC, and especially here: https://t.co/MSJJViHxbM. Senn is trying to make Nedler’s calculus relevant, so he complicates it with "block design" "many dining halls", and more - unnecessarily. https://t.co/MZ3IG346f7

2022-12-26 02:00:20 @stephensenn @soboleffspaces @ADAlthousePhD If the natural question asked by Fredrick Lord -- "What's the effect of diet on weight gain" -- is deemed a "red herring" unless it conforms to someone's suggested experiment, then the suggested experiment should be modified to capture natural questions, not the other way around

2022-12-26 01:21:27 For the last night of Chanukah, here is a great performance of Hendel's Judeas Maccabeas https://t.co/U5hNejW5b9 About 25 min into the video you can find the song we sang in Kindergarten (Hava Narima -- Macabee Gibor), the one that surprised my mother in https://t.co/8m9OeEWSJQ

2022-12-25 21:59:48 Of course design matters

2022-12-25 10:55:44 @stephensenn Causal Inferencers believe that experimental design is a man made tool devised to interrogate Nature. They can't understand, therefore, why Nature, or at least a model thereof, does not show up the writings of many "designers" who claim to worship "Nature".

2022-12-25 10:40:57 @SSI_Movement My duty and my honor, or, as a friend expressed it tonight in our Chanukah party tonight: It's our "Army reserve service (Sherut Miluim)", considering that our academic colleagues in Israel are serving in Jenin. Happy Chanukah.

2022-12-25 02:16:56 But how can we leave Hanukah without our favorite songs (1) Ocho Candelikas ( Ladino https://t.co/oGG2BKYQ9V and (2) Sevivon Sov Sov Sov (Hebrew, about 11 minutes into this video ) https://t.co/1Gv2avtSTt Hallelujah!

2022-12-25 01:17:02 @Ostrov_A Many many more years of happiness

2022-12-24 14:43:00 @BradSpellberg @DocBio1509 @f2harrell @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone This is precisely where Causal Inference comes it. I tells us: "Your conclusion is as valid as the assumption: "jumping out of an airplane without a parachute increases your chances of death", no more no less." That's why it's so important to make assumptions transparent.

2022-12-24 14:20:56 End of year always brings with it a flood of requests for "end of year giving." This year, my choice is made -- I am donating $5,000 to SSI (Students Supporting Israel), on whose Faculty Advisory Board I am proudly serving. https://t.co/wNhQqeLe0W Please consider this cause.

2022-12-24 12:49:03 @Abelaer @mattecapu And I took it from Toulmin's Foresight and Understanding: An Enquiry into the Aims of Science (1961) ISBN 0-313-23345-4

2022-12-24 02:09:36 @Claire_V0ltaire Palestinians have a paper title for everyone and historical connection to none

2022-12-23 22:43:36 Hi everybody, Happy Holidays to all readers. I'll have to take some time off twitter, abandoning even the intense discussion over trialists and CI, Will be back, I hope, with renewed energy to explore new vistas in Causal Inference and Human-level AI. Happy New Year.

2022-12-23 22:34:09 @stephensenn @soboleffspaces @ADAlthousePhD The Lord Paradox example in #Bookofwhy is about "causal effect" and, if it is a property of the population, then we should be able to discuss it without detailed specification of the data collection process. This is exactly what the example is about, it's about "causal effects".

2022-12-23 22:19:03 @stephensenn @ADAlthousePhD @soboleffspaces So what is "causal effect"? A property of the population? Or a property of the units in the study? If the former, then we should be able to discuss it without detailed specification of the data collection. If the latter, we have an uncrossable language barrier. Please choose.

2022-12-23 21:46:42 @stephensenn @ADAlthousePhD @soboleffspaces I've tried, but got stopped by a 5-mile tall language barrier. E.g., what is a "causal effect"

2022-12-23 21:34:25 @soboleffspaces @stephensenn @ADAlthousePhD I admit that my opinion of how trialists think is based (partly) on @stephensenn refusal to define "causal effects", or even to use the term in his tweets. Same applies to "estimand".

2022-12-23 21:17:36 @stephensenn @ADAlthousePhD @soboleffspaces To answer your questions we need to establish some common language. My questions aim to establish it and they haven't been answered thus far. We havn't even established what the "trial objective" is, so I'm waiting. Courage?

2022-12-23 21:11:18 @VickersBiostats @soboleffspaces I've raised the possibility that trialists and CI folks may in fact be thinking the same way, but we'd never know unless the former tell us what the words "trial objective" and "estimand" mean to them, and how they express them mathematically. Courage?

2022-12-23 15:35:32 @TenanATC @ADAlthousePhD @soboleffspaces Sorry that you do not think you can learn something from my difficulties. I usually learn a lot from readers' difficulties, e.g., I learn to explain myself better, and to encourage more questions. Too bad.

2022-12-23 15:18:34 @soboleffspaces @stephensenn @ADAlthousePhD It's OK to define things differently. What is not OK is to refrain from making your definitions explicit and complain that people do not understand what you had in mind.

2022-12-23 15:11:53 @stephensenn @ADAlthousePhD @soboleffspaces Another "absurd". How about telling us what the words "trial objective" and "estimand" mean to YOU, personally, and how you express them mathematically. Courage?

2022-12-23 15:10:49 @TenanATC @ADAlthousePhD @soboleffspaces Another "absurd". How about telling us what the words "trial objective" and "estimand" mean to YOU, personally, and how you express them mathematically. Courage?

2022-12-23 14:32:58 @ADAlthousePhD @soboleffspaces To all offended trialists. There is no point repeating how stupid I am. Instead, just tell us what "trial objective" and "estimand" mean to YOU, personally, and how you express them mathematically. We may mean the same thing, but we'd never know unless one of you say it. Courage?

2022-12-23 14:02:47 @ADAlthousePhD @dailyzad @soboleffspaces The whole discussion started by two of your peers who expressed objections CI, presumably, because CI is (1) asymptotic and (2) does not do the estimation. Such objections imply "tasks begin with data." If there is acceptance of CI among trialists - I declare myself a trialist.

2022-12-23 13:51:45 @ADAlthousePhD @soboleffspaces You keep on telling me that I am wrong, when I am dying to hear what trialists believe their task begins with. Is it so hard to articulate? One equation should do it.

2022-12-23 13:45:56 @dailyzad @ADAlthousePhD @soboleffspaces The paper you cite is using CI language and methods, eg. DAGs. In this setup, "estimands" are natural. Can you cite a paper authored by anti-CI trialists?

2022-12-23 13:27:39 @ADAlthousePhD @soboleffspaces There is another point here. Trialists do not NEED to think of estimands at all, because there is only one estimand in each trial, and it is not different from the trial objective, since no observational studies are involved. Thus, trialists are lucky, no need to be defensive.

2022-12-23 13:13:10 @ADAlthousePhD @soboleffspaces I still don't know how, and your slides do not help me b/c, though verbal descriptions appear, formal expressions are lacking. Thus, I cannot tell for example if "estimand" is an asymptotic entity (shunned by your peers) or sample-based entity. My difficulties are sincere. 2/2

2022-12-23 13:01:02 @ADAlthousePhD @soboleffspaces It's deeper than that. I believe one cannot properly define or understand "Trial Objective", "Estimand" and "estimate" unless one adapts the CI language. Having seen objections to such adaptation from your peers, it is only reasonable that I be curious how trialists do it. 1/2

2022-12-23 12:02:30 @ADAlthousePhD @soboleffspaces Let's not focus on what is wrong with me. Let's hear what estimand means to trialists, how it is defined, how it is expressed mathematically, whether it is ever different from the "query",etc, etc, so that we can understand what makes trialists different from CI folks, if at all.

2022-12-23 11:50:09 @ADAlthousePhD @soboleffspaces If I have made incorrect statements about what trialists do or think, please correct me - what do they do and think? My statements express my best guesses trying to understand what makes trialist so unhappy with CI, where queries, estimands and estimates are so crisply defined.

2022-12-23 00:04:34 Refreshing to see students taking action where teachers and community leaders slumber into inaction. What worries me though is the standard appeal for "education about the Holocaust" instead of education about Jewish history, Israel, Jewish values and Jewish revival.

2022-12-23 00:04:33 Maryland High School Students Stage Walkout to Protest Antisemitism https://t.co/Ls0CHMgQcc

2022-12-22 23:21:18 I join @AdanZBecerra1 in congratulating @AmandaKMontoya on her Mediation preprint, which should introduce causal mediation analysis to social and personality psychology. I think however that there is no such thing as "statistical mediation". The Baron-Kenny tradition 1/2 https://t.co/esLA5vRUkD

2022-12-22 22:34:25 @imthealan I would go even further and say that Causal Inference is the ONLY scientific method we have for dealing with interventions and counterfactuals.

2022-12-22 22:30:16 @ildiazm @f2harrell @LarsvanderLaan3 @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Agree. CI deals separately with the two uncertainties, first confounding bias, then sample variability. The anti-CI folks dislike the former, because it's "asymptotic" hence must be avoided. I doubt it can be handled any other way, implying that @f2harell's "proof" doesn't exist.

2022-12-22 18:39:33 @f2harrell @LarsvanderLaan3 @ildiazm @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone I am not aware of any " near proof of confounder completeness" not using DAGs, and would be mighty surprised if such exists.

2022-12-22 10:34:04 The diagram tells which partial correlation gives the causal effect we need, and which doesn't. Indeed, this simple &

2022-12-22 09:55:11 I liked this quote: "People sympathetic to Israel are going to have an increasingly hard time differentiating between kinds of criticism. The way to do so is to ask if a critic is trying to make Israel better, or trying to make it disappear." Perhaps because I smell the latter. https://t.co/VIB4hbOg6X

2022-12-22 09:47:01 @erikbiz @RabbiWolpe Forgot the piece: https://t.co/8m9OeEEJvI

2022-12-22 09:45:25 @erikbiz @RabbiWolpe Agree! Even wrote an opinion piece against tendencies to obscure the national message of Chanukka with religious fables. Generations of Rabbis had good reasons for doing it - it was safer to play down Jewish national aspirations. Today, it's just a sign of confusion &

2022-12-22 08:40:00 @AvivaKlompas @MonaAbuamara Great progress!! The Palestinians have learned to compromise! Last we heard, they were not going to hold an election "until Tel Aviv is liberated". Now it's only Jerusalem -- peace is on its way.

2022-12-22 08:09:41 how it inspires us, and how its current struggle for survival parallels the Maccabees struggle for independence. I said: "Hannukah without Israel is un-Jewish." But the dignitaries already left. I reached to the piano and sang Hatikva. Luckily a few boys (4-7) joined me, 2/3

2022-12-22 08:01:54 Yesterday I lighted the Channukah candles at Chabad House. As the celebration progressed into the evening, dignitaries gave their speeches, light over darkness hailed, rising antisemitism condemned and holocaust survivors honored, I noticed that no one mentioned Israel, 1/3

2022-12-22 07:09:24 Eight Thoughts on Israel’s Political Crisis https://t.co/v9hYtVxbVY

2022-12-22 05:20:39 I've been using "data-intensive sciences", to distinguish it from "data science" which, unfortunately, has become "the science of data" instead of what it should be: "the science of the data-generating reality". https://t.co/LR2nMNvwfO

2022-12-22 02:17:05 Call me an ethno-centric, but I could not help but think about Chanukah as President Zelensky spoke tonight

2022-12-22 00:18:49 @soboleffspaces One tiny nuance: "What we are looking for", the query, can also be expressed in causal vocabulary, e.g., E[Y|do(x)]. The "estimand", on the other hand must be expressed in statistical vocabulary, so that even an anti-CI trialist would know what needs to be estimated.

2022-12-22 00:04:36 @ggreenwald Dear Glen, as one of the victims of the Berkeley's Bylaws, let me confirm @Ostrov_A's allegation that we are facing a clear case of discrimination on the basis of national origin and, contrary to NYT implications, I'm no "conservative commentator" nor anti-Palestinian activist.

2022-12-21 21:36:17 Thank you @soboleffspaces for posting this page from #Bookofwhy. Reading it again, I begin to understand why trialists rarely use the word "estimand" -- they believe their task begins with the data, albeit experimental, not with the query. https://t.co/JvFsmvEgV2

2022-12-21 21:20:39 @AdanZBecerra1 @kat_hoffman_ @RWJE_BA @dylanarmbruste3 @KordingLab @ildiazm @f2harrell @elmir1omerovic @dredgardorsey @ArgoloF @drjohnm @kaulcsmc @GreggWStone @_MiguelHernan Whenever I see an appeal to "expertise or experience" I interpret it to be "informal wisdom" which tells me either (1) We, AI-ers need to capture it algorithmically, or (2) The practitioners are hiding something from science, or both. Which is it?

2022-12-21 17:47:13 To further elucidate the concept of "estimand," I would recommend Fig. 1 page 12 in #Bookofwhy, where you can see an entire "inference engine" and how estimands and estimates are produced. https://t.co/Xwhd7Frs5L

2022-12-21 17:04:48 @f2harrell @soboleffspaces @ildiazm @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone When you ask a causal question then, in order to convert the question into an estimand (a statistical quantity), you need Causal Inference (CI) to do the conversion. You can do it in your head, as in most RCT studies, but it is still CI. All the estimands I've seen are asymptotic

2022-12-21 10:48:07 RT @elderofziyon: 12/20 Links Pt1: Secret document reveals EU plans to help Palestinian takeover of Area C

2022-12-21 10:11:55 It's not Biden alone to blame. Jewish leaders are falling all over thanking him for inviting them to the whitehouse, rather than asking him to be consistent and to backup his beatiful words with meaningful commitments, e.g., to seriously investigate cases such as Berkeley's Zones https://t.co/5gVq37YRuS

2022-12-21 10:01:34 @tdietterich @amt_shrma @ChenhaoTan I'm starting to play with it - very impressive. However, it fails the toy problems I pose to it. It is still illuminating to find out how much children can learn strictly from teachers, without guidance from playful manipulations and without innate templates for causal models

2022-12-21 09:15:11 I may not count as a "leading scholar on antisemitism" but I have published substantially on the subject, perhaps more than any of those "signatories". With this in mind, can @PeterBeinart educate us: What is "smearing" in exposing why she does not fit this job's qualifications? https://t.co/wyOdSGXrdX

2022-12-21 08:40:21 RT @Ostrov_A: One cannot help but question Joe Biden’s sincerity to combating antisemitism, when on the one hand, he makes such passionate…

2022-12-21 02:23:38 @soboleffspaces @f2harrell @ildiazm @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone I'm also not sure I got it correctly. Guessing: the anti-CI statisticians complain that CI folks dont do the final estimation, something they can do very well. The notion of "estimand" is not in their vocabulary, perhaps because they would need to get it from CI. A puzzle.

2022-12-20 22:38:36 @f2harrell @ildiazm @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone As I understand it, @f2harell will never forgive CI folks for letting statisticians do what statisticians do best, namely, take the CI solution (i.e., the estimand) and estimate it from finite data in some specific domain, using the latest available technique.

2022-12-20 15:55:14 If I'm not mistaken, this @Columbia is Columbia University, also known as an accredited academic institute. Or am I mistaken? https://t.co/qwUAcak2FO

2022-12-20 12:43:39 I should add though that the Bayesian shift from shaky to plausible assumptions is accomplished entirely within the province of probability theory

2022-12-20 12:16:33 Again, Jewish leaders are up the mountain tops praising Biden for general words against antisemitism, instead of telling him what actionable words they would have liked him to say. For example, that anti-Zionism is first degree racism. @Ostrov_A @bandlersbanter

2022-12-20 12:16:32 Biden slams ‘venom’ of antisemitism at Hanukkah event, unveils White House menorah https://t.co/VARmKgRVhN via @timesofisrael

2022-12-20 11:58:49 The ‘Jew-Free Zones’ at Berkeley story wasn’t ‘misinformation’ https://t.co/FcIa6i2aX6 via @JNS_org

2022-12-20 11:29:29 @richard_landes @jsternweiner If @jsternweiner receives letters from Avi Shlaim he must have received the conditions under which @FranceskAlbs would qualify for the job. Those who deny Israel's right to Tel-Aviv, are hardly in position to judge how to defend that right in Jenin.

2022-12-20 10:51:43 there is a fundamental difference between the no-confounding assumptions that RCTs are designed to neutralize and the assumptions upon which the recovery methods rely. Researchers may be totally ignorant of the structure of the former and quite knowledgeable about the 5/6

2022-12-20 10:51:42 First, consider the philosophy of Bayes analysis. Why should an analyst postulate priors to derive posteriors? Why not postulate posteriors directly? The answer is that the former judgment is more reliable than the latter. In Bayes' example (#Bookofwhy, p.98), judging the 2/6

2022-12-20 10:51:41 Readers who continue to be attacked by "RCT Priests" have probably noticed that in "Dialogue with hostile examiner" https://t.co/dEbgsXY1Q3 (p.369) I am claiming that CI is a mapping from shaky assumptions to plausible assumptions. I here wish to exemplify it in two domains: 1/6 https://t.co/wmHSzq5CwV

2022-12-20 07:42:06 Fortunately, most American Jews know that abandoning Israel, as @Eric_Alteman advocates, is an act of identity-suicide, and they know better. https://t.co/esGZMi6MG0

2022-12-20 06:58:09 @HumzahAlkindi From what I can see, it's a collection of algorithms for doing estimation of causal quantities. I'd question only whether readers would be able to understand the assumptions under which the algorithms would deliver unbiased estimates.

2022-12-20 04:46:12 Hamas calls on Palestinians to defend Al Aqsa against Israeli plans to celebrate "the so called Hanukah Holiday". Quoting: "the so called Hanukah Holiday". I was right in saying that Palestinians fear Hanukah more than they fear the Mossad: https://t.co/8m9OeEEJvI https://t.co/rjrtLaNBqW

2022-12-19 13:25:29 @ranilillanjum @ericjdaza @ArgoloF @ildiazm @KordingLab @f2harrell @RWJE_BA @kat_hoffman_ @elmir1omerovic @AdanZBecerra1 @dredgardorsey @drjohnm @kaulcsmc @GreggWStone @_MiguelHernan @ElenaRoccaPD @Cause_Health Readers curious about the relation between SCM and framework proposed by @ranilillanjum would note that, in SCM, there is only one concept of causation (ie, functional relations), giving each cause a different degree of "sufficiency", "necessity", "directness", "prob. raiser" etc

2022-12-19 09:58:54 Please read carefully, and keep in mind, any resemblance between the actors in the dialogue and participants in our Twitter thread is purely coincidental, though some of the arguments could sound astonishingly similar. 2/2

2022-12-19 09:58:53 Recent discussions on "observational trials can NEVER establish causality", followed by misunderstandings of what Causal Inference is about, are jolting me to retweet Ch. 11 of Causality https://t.co/Ty53coWZrI where you can find (p.369) "A dialogue with a hostile examiner" 1/2 https://t.co/NKqkVhhBzs

2022-12-19 00:03:01 There is no way for me to start the Holiday without sharing my understanding of why Chanukah is the Jewish "trust deed" to history: https://t.co/8m9OeEEJvI @Ostrov_A, @SAFECUNY, @robsatloff https://t.co/HcL3n7LcBF

2022-12-18 23:42:22 @NotWokeBiatch @Ostrov_A Amen! Amen!

2022-12-18 22:39:39 @KordingLab @ildiazm @f2harrell @RWJE_BA @kat_hoffman_ @elmir1omerovic @AdanZBecerra1 @dredgardorsey @ArgoloF @drjohnm @kaulcsmc @GreggWStone @_MiguelHernan "Mechanism understanding" is far from ignored. (1) It is this understanding which makes one assumption more plausible than another. (2) This understanding is the aim of mediation analysis.

2022-12-18 22:02:38 @KordingLab @ildiazm @f2harrell @RWJE_BA @kat_hoffman_ @elmir1omerovic @AdanZBecerra1 @dredgardorsey @ArgoloF @drjohnm @kaulcsmc @GreggWStone @_MiguelHernan If you believe in the impossibility of observational causality, you should also believe in the impossibility of experimental Causality, because of population mismatch. Both depend on assumptions, and CI help us replace shaky assumptions with more plausible ones.

2022-12-18 20:51:33 @DiogoMOSilva @DanielGordis @isaacdecastrog "Palestinian rejectionism" is Palestinian willingness to forego independence, freedom and prosperity in order to prevent Jews from having independence, freedom and prosperity - "not even the size of a postage stamp".

2022-12-18 20:42:27 @dataengines @RWJE_BA @stephensenn @AdanZBecerra1 @lakens @Lester_Domes @learnfromerror I'll with the "larger picture".

2022-12-18 20:36:52 @BradSpellberg @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone @f2harrell Nice, but I would add that, in order to assess what information they add, and how to use that added information, we need a theory of Causal Inference (CI). This is, in essence, what CI is all about.

2022-12-18 20:28:21 @AyalSharon @tdietterich @DanielGordis @isaacdecastrog I wouldn't call it "ridiculous" because this is how the NYT and most Western commentators present the conflict to the public, refusing to acknowledge the elephant and its core asymmetry: One side says "Equally indigenous" the other says "Not even the size of a postage stamp".

2022-12-18 20:01:07 @f2harrell @RWJE_BA @kat_hoffman_ @ildiazm @elmir1omerovic @AdanZBecerra1 @dredgardorsey @ArgoloF @drjohnm @kaulcsmc @GreggWStone @_MiguelHernan I have tried to decry this confusion with propensity scores since 2009: https://t.co/BYnjoEsn8D But I believe you need a Theorem like Eq. (11.10) to argue that using propensity score does not result in any reduction of bias, even when bias is inevitable.

2022-12-18 19:48:11 @f2harrell @ildiazm @RWJE_BA @kat_hoffman_ @elmir1omerovic @AdanZBecerra1 @dredgardorsey @ArgoloF @drjohnm @kaulcsmc @GreggWStone @_MiguelHernan But do you see any impediments to filling those "unspecified elements of the scheme" with the best available methods developed by smart statisticians? I don't.

2022-12-18 12:43:01 Tonight, as we light the first candle of Chanukah. I am reminded of the 2007 Whitehouse celebration, in which I and my late wife Ruth were invited to light our family Menorah, and of my late son Danny who mentioned his great grandfather in his last words https://t.co/pxQJwI3tU3

2022-12-18 10:26:43 @stephensenn @AdanZBecerra1 @lakens @Lester_Domes It would take me many hours to sort through the data of the 1930's Milk Experiment. I would do it if I knew that I could learn a new insight from it. Can you describe the insight I should expect?

2022-12-18 09:51:46 @DanielGordis @isaacdecastrog Here is a funnier sentence: "Hopes for a Palestinian state have dimmed under the combined pressure of Israeli resistance and Palestinian corruption, ineptitude and internal divisions." Nothing about the giant elephant in the room - Palestinian rejectionism.

2022-12-18 09:23:54 @GeraldNGOM paper is more constructive than Tom Friedman's "watch this space" analysis, but I am not sure that "annexation" would favor better than "occupation" or "expansion" in the eyes of those who refuse to see the real impediments to a 2-state solution. https://t.co/NkGZq6kNYR

2022-12-18 07:46:14 These notions of soundness and validity govern the logic of Causal Inference, once we replace the classical notion of "world" with a "fully specified SCM model." https://t.co/wBGqfvxH4Y

2022-12-18 07:03:33 @stephensenn @AdanZBecerra1 @lakens @Lester_Domes Is the Milk Experiment and example of how causal modeling is properly integrated with statistical estimation? Is it done with out CI? If so, can you summarize the priniciples behind the integration method?

2022-12-18 06:57:39 @ildiazm @RWJE_BA @kat_hoffman_ @elmir1omerovic @AdanZBecerra1 @f2harrell @dredgardorsey @ArgoloF @drjohnm @kaulcsmc @GreggWStone @_MiguelHernan I join @ildiazm in recommending this paper by Maya &

2022-12-18 06:27:59 @maradonasattva @FranceskAlbs Glad I got you to realize that people who respect truth as scientists also respect facts and abhor populist slogans like "apartheid," "extermination" etc., which the misinformed take for unassailable facts.

2022-12-18 06:05:40 Unfortunately, once you give her an opening to plead innocence of charges she wishes to deny, the whole conversation gets embroiled in endless philosophical sophistry: "Is she or isn't she an antisemite?", and you lost the chance of accusing her of charges she cannot deny. https://t.co/n3Ioi4S0o5

2022-12-18 03:12:34 @maradonasattva @FranceskAlbs `Good question, I assume these will have to stand with Atlantis, Unicorns, Squared Circles and other impossible yet wished-for objects.

2022-12-18 02:52:57 @stephensenn @AdanZBecerra1 @lakens @Lester_Domes Identification (1) listening to estimation (2)? By all means! Likewise, estimation (2) listening to identification (1). Of course! The aim of CI to coordinate the conversation

2022-12-17 13:22:07 @Lizstuartdc @PWGTennant @AdanZBecerra1 @stephensenn @lakens @Lester_Domes Rosenbaum's "causal crossword" analogy has been a romantic desideratum for generations (e.g., Deaton etal: https://t.co/B0TKO64aKj). What Rosenbaum does not realize is that a formal framework for "causal crosswords" is provided by the data-fusion theory of https://t.co/E1ap1MNoKJ

2022-12-17 12:43:35 Academic sophistry disputing what is or isn't antisemitism is another reason we should not accuse @FranceskAlbs of antisemitism, but of much uglier form of racism - Zionophobia - a charge she cannot deny, and which should disqualify her immediately from the UN job she's seeking. https://t.co/D6PQ3GfS7F

2022-12-17 11:21:26 @TristanBlumen Because once you give her an opening to plead innocence of charges she wishes to deny, you lost the chance of accusing her of charges she cannot deny.

2022-12-17 10:59:40 @stephensenn @AdanZBecerra1 @lakens @Lester_Domes You are probably thinking of some caricatures of CI painted by some statisticians. In general, CI deals separately with the two sources of uncertainty: (1) asymptotic bias and (2) sample variability. The former under "identification", the latter under "estimation".

2022-12-17 10:50:54 @ProfHayward @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone @f2harrell This is precisely what Causal Inference provides us: A mapping between the plausibility of the assumptions and their context, to the plausibility of the conclusion. Why is it problematic?

2022-12-17 10:45:25 I still maintain it's a grave mistake for Jewish leaders to challenge the fitness of Francesca Albanese on the basis of her antisemitic remarks 10 years ago, instead of proving her unqualified on the basis of what she says TODAY -- denying Jewish right to a homeland. https://t.co/NzynvCC4QC

2022-12-17 09:43:04 RT @AdamMilstein: The Israel Philharmonic Orchestra will perform in Abu Dhabi on Tuesday, marking its historic UAE debut. https://t.co/auM1…

2022-12-16 21:44:22 @f2harrell @ildiazm @dredgardorsey @RWJE_BA @ArgoloF @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone The principles of causal inference are not meant to lead us to unassailable treatment efficacy in either observational or experimental studies. They lead us from plausibility of assumptions to plausibility of conclusions, in both cases. The assumptions in RCT are not unassailable

2022-12-16 21:30:43 @AdanZBecerra1 @stephensenn @lakens @Lester_Domes To the extent that 100 studies can prove "something", a single study can prove 1/100 of that "something". I can't understand the sudden interest in "proving something" when CI explicates precisely what can be learned from each study.

2022-12-16 21:20:43 @peterjliu @ataiiam What is LM? What is "causal LM"?

2022-12-16 20:33:49 @xh_andrea I have often noted the parallels between Russian and Palestinian propagandists: Both deny their neighbors right to exist, and both describe their neighbors' defense as aggression. Good company.

2022-12-16 20:06:19 Thanks for letting me know about your journal club, but the paper you are discussing is misguided, and should always be read together with its compendium https://t.co/sZPpL7Ifmg, and the references in https://t.co/Jf1SkgV3CK, https://t.co/pYrfhD1GID and https://t.co/etASSEk4Or. https://t.co/fM9ferjLmb

2022-12-16 19:42:44 @ataiiam @peterjliu Amazing how long it takes for people to accept that calculus and probabilities are insufficient for progress in AI.

2022-12-16 19:31:53 @xh_andrea Those who count casualties and ignore intentions are bound to conflate "defensive reality" with "oppressive reality" and are therefore unqualified to serve as UN Rapporteurs in any conflict.

2022-12-16 10:23:10 She has a point. Why, for havens sake, people dig out things she said 10 years ago? She can easily be disqualified by what she is saying today. Just ask her what she thinks of Zionism, or whether Tel Aviv is stolen land. https://t.co/KrJHniVpOn

2022-12-16 10:09:26 UK readers, if you wish to see why the Israeli-Palestinian conflict is the simplest conflict in human history, don't miss out on this event with @EinatWilf https://t.co/PSqZvtf2hr

2022-12-16 09:58:42 Regardless of its outcome, the Biden Administration decision to investigate the Berkeley Law School scandal sends a clear message to all US Universities: Beware, there's a new minority group on the block -- Zionist students and faculty -- who is demanding recognition and 1/2 https://t.co/XndLf8bjQU

2022-12-16 07:15:59 To be honest, I didn't know a field called "Learning Analysis" (LA) exists, separate from "Education". Now that I know, I am glad they are getting to read this introduction to Causal Inference, to get them going on the right path. https://t.co/FFEpp48wxm

2022-12-16 01:51:26 Reassuring to know that the Office of Civil Rights is launching an investigation into to civil rights violations at UC Berkeley Law School. https://t.co/P89uXfLjt1

2022-12-16 00:39:15 @ildiazm @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone I still do not know how one can "estimate causal effects from data" alone, without making some extra-data assumptions, or resort to experiments. Unless, of course, one labels those extra-data assumptions "Bayesian".

2022-12-16 00:26:40 @EqualEffrontery @AStratelates Would be curious to find out how Duhem interprets Galileo's revolution, and why his interpretation is more satisfactory than mine, as articulated in this epilogue https://t.co/00cnodjgNc

2022-12-15 18:40:26 @kaulcsmc @noah_greifer @GreggWStone @f2harrell @elmir1omerovic @drjohnm I'm not familiar with of those evaluation methods

2022-12-15 18:31:45 @GreggWStone @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @f2harrell Can we generalize from multiple RCTs without accounting for the differences between those RCTs ?? Can we account for those differences without a formal language to do so? Here is one: https://t.co/E1ap1MNWAh. Any alternative?

2022-12-15 18:19:58 @f2harrell @dylanarmbruste3 @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone CI recognizes that every conclusion is a function of 2 types of uncertainty and, in contrast to Bayesian thinking, properly handles the two types. That's why I am only a half-Bayesian https://t.co/gqeQb23En9

2022-12-15 18:10:58 @ildiazm @f2harrell @elmir1omerovic @drjohnm @kaulcsmc @GreggWStone Building upon "multiple sources of evidence" requires a formal language in which the characteristic of each source are described and coherently fused (or "triangulated"). This is where Causal Inference comes in.

2022-12-15 08:38:07 Beautiful title: Can the Whole world be Wrong? https://t.co/UHH7NKEwdc

2022-12-14 10:45:30 I am proud to join over 130 Jewish Studies professors and educators who have signed today a statement calling on the United Nations to adopt the IHRA Definition of Antisemitism (IHRA WDA) https://t.co/pSJfamvBHt. This is the first time ever that Jewish Studies scholars 1/2

2022-12-14 05:07:59 Agree, 3 is willful. But this still begs the question of why reasonable analysts, sympathetic to Israel, would blind themselves to reality. My answers: (1) is truly irrational, requiring new prism to be believable. (2) breaks the Western ethos of "Smart we can surely fix it". https://t.co/OoGFaMTsV5

2022-12-14 02:33:27 The causal inference literature continues to swell with new and interesting results - here is our weekly report: https://t.co/AUlf2uJyJ6. Many titles catch my eyes, like this: https://t.co/ryFD76wSeQ which bounds counterfactuals by combining observational and experimental data.

2022-12-14 00:47:27 It was a fairly long lecture and, I agree, it should be useful for curious outsiders asking: what's all this noise about causal inference - what is it? https://t.co/T7agG9mcCY https://t.co/I2xLdeVZTg

2022-12-14 00:30:40 Happy Chanukah, Amir

2022-12-13 23:19:30 @scwallphd @RWJE_BA @imthealan Same with arithmetic.

2022-12-13 22:27:25 @kareem_carr And the same is true about arithmetic. Arithmetic is the "science of everything" if allowed the flexibility to imagine a future where arithmetic is combined with everything "into a single venerable discipline."

2022-12-13 22:21:51 @kareem_carr With this flexibility, Statistics is indeed the "science of everything". Why? Because if someone were to tell us "No, statistics is missing XYZ" we would say: "when we say "statistics" we imagine a future where XYZ and statistics are combined into a single venerable discipline"

2022-12-13 17:07:44 @blakeflayton This rotten habit of embedding antisemitism in the context of other forms of hate is aimed to justify inaction under the cover of endless philosophical debates of hate-related phenomena. UCLA is using this trick too: https://t.co/c6Vl4qIQ6T

2022-12-13 16:27:41 I guess I am, and will remain sentimental

2022-12-13 16:27:40 2,200-year-old coin hoard gives hard proof of Book of Maccabees, say archaeologists https://t.co/MNCG6YfZvo via @timesofisrael

2022-12-13 09:27:49 The debates on who owns whom did not end with Karl Pearson. Some statisticians still think causal inference is a "missing data" problem (e.g., Rubin and Imbens) and some deep learning folks think it's an neural-net optimization problem. https://t.co/aisVMBIqxT

2022-12-13 05:18:17 This is what Karl Pearson thought, when he discovered correlation, "everything is statistics", he said (paraphrased). But Sewall Wright, already in 1921, told him: "No way, Karl, statistics can't answer even half of my scientific questions" #Bookofwhy. https://t.co/KPMWZME7dF

2022-12-12 20:47:47 Here is your chance to see the 3 major obstacles to peace in the Middle East: (1) Palestinian rejection of any form of a Jewish state, in any borders, (2) Israelis awareness of (1), (3) Western media un-awareness of (1) &

2022-12-11 20:45:41 @osoleve @DrRAWesterman @Dominic2306 A better way to phrase it is: My neighbor's roof has been getting wet whenever mine did.

2022-12-11 15:27:26 @_HassanEldeeb_ It's not nice calling other people "the devil," civilized people do not do that.

2022-12-11 02:33:31 If you are connected in any way to Jewish Studies, or International Law, please consider adding your signature to this important statement: https://t.co/mKanSiHfho by writing to <

2022-12-10 23:44:14 Some Retweets require comments, some demand a word of caution. This does not. https://t.co/UZe6THPLIa

2022-12-10 23:29:58 This self-evident item should have been submitted as an "amendment" to the original HR Declaration, if only we had a HR-minded body to submit it to, unmolested by the forces that made Jewish self determination less than self-evident. https://t.co/Oln1tZLgSx

2022-12-10 22:53:23 @the_heruman Thanks. I'm impressed again. Lets try asking about responsibility. Who is responsible? The guy who struck the match or the guy in charge of Oxygen in the room? Dead curious!

2022-12-10 22:46:01 @DrRAWesterman @Dominic2306 This is straight propositional logic, which requires no causal inference. I am interested in problems that cannot be handled by standard logic.

2022-12-10 22:40:08 This is fairly impressive for a LLM. I wonder how it would handle counterfactuals like in Firing Squad. Or,in this oldie: “A fire broke out after someone struck a match, ‘What caused it, striking the match or having oxygen in the room?’ See https://t.co/yd1Z4QGARx https://t.co/pzHGxuHGbt

2022-12-10 20:08:24 It shouldn't be too hard for LLM to add a causal reasoning engine to their system. I suggest they start with an oldie: 1. My neighbor's roof gets wet whenever mine does. 2. My roof gets wet whenever I hose it. 3. Will my neighbor's roof get wet if I hose mine? (Causality p.1). https://t.co/FCKstXyYxC

2022-12-10 19:56:40 It's a heart-warming photo, to be standing next to the late Tom Tugend, and next to Susan who wrote this powerful eulogy about Tom: https://t.co/R1g6tiS946 https://t.co/mFmf5gMke9

2022-12-10 18:54:52 @uncoolbob @DanFmTo @Ostrov_A @nadplo @Palestine_UN @EUpalestinians @ShaharAzani @benshapiro @COLRICHARDKEMP @CUFI @ActForIsrael @ReinsteinJosh Eliminationists use creative ways of defining their neighbors: 1. Putin says Ukrainians are Nazis. 2. China says Taiwanese are imperialists 3. Palestinians say Israelis are "settler-colonialists" 4. Some Western observers fall for it.

2022-12-10 11:54:19 President Mahmoud Abbas has very good reasons to be angry at the Balfour Declaration: https://t.co/9XOCOsqkIT. It embraced Jews and Arabs as equal claimants to the land - an idea he cannot swallow to this very day. https://t.co/mPYXOFZDx0

2022-12-10 10:18:55 @uncoolbob @DanFmTo @Ostrov_A @nadplo @Palestine_UN @EUpalestinians @ShaharAzani @benshapiro @COLRICHARDKEMP @CUFI @ActForIsrael @ReinsteinJosh I assume you know that I did not expect to convince a person with your persuasion to listen to facts and principles. I just wanted to make sure you know that your Zionophoic rants are deemed morally repulsive, if not racist in most normative circles.

2022-12-10 02:13:58 @ruthenian @AbsDelfuego @Ostrov_A @nadplo @Palestine_UN @EUpalestinians @ShaharAzani @benshapiro @COLRICHARDKEMP @CUFI @ActForIsrael @ReinsteinJosh The pen is mightier than the sword. Behind every rifle there is a mind, driven by words, legitimizing the shot, delegitimizing the target.

2022-12-10 01:21:55 I just found out that the original film recording of the UN Vote of November 29 1947 was only recently discovered, and it's up for sale at an auction, initial offering $200,000. We can watch it here: https://t.co/ToOyfvNsyX, 5:13 min. into the video. https://t.co/9nioi6kLPk

2022-12-10 01:05:52 @uncoolbob @DanFmTo @Ostrov_A @nadplo @Palestine_UN @EUpalestinians @ShaharAzani @benshapiro @COLRICHARDKEMP @CUFI @ActForIsrael @ReinsteinJosh Being Palestinian is not a choice, but denying your neighbor's right to existence IS a choice, and an ugly, racist one, and denying your neighbor's right to define itself as a PEOPLE (not merely a polity) is doubly racist. Inconvenient, but doubly racist.

2022-12-09 22:08:11 @DanFmTo @uncoolbob @Ostrov_A @nadplo @Palestine_UN @EUpalestinians @ShaharAzani @benshapiro @COLRICHARDKEMP @CUFI @ActForIsrael @ReinsteinJosh Distinctions based on mindsets are not racist. On the contrary, such distinctions point us toward the means of changing those mindsets and replacing them with peace-enabling worldviews.

2022-12-09 21:43:47 @neontaster You are in good company. Mehdi Hasan is a product of @AlJazeera

2022-12-09 21:36:30 I'm proud of the Los Angeles community who understood the calling of history and leveraged the 75 anniversary of Resolution 181 to remind the world that Israel is here by historical right, not by favor, nor by might. @IsraelMFA should share this video w/ other communities. https://t.co/7ljAZnDPKY

2022-12-09 21:14:31 @uncoolbob @DanFmTo @Ostrov_A @nadplo @Palestine_UN @EUpalestinians @ShaharAzani @benshapiro @COLRICHARDKEMP @CUFI @ActForIsrael @ReinsteinJosh Zionists don't whine

2022-12-09 20:54:00 A true friend, a lover of his people, a fighter for truth and a true mencsh​. Tom, I love you. Your memory is already a blessing to us all.

2022-12-09 20:53:59 Tom Tugend, War Veteran and Lifelong Jewish Journalist, has Died at 97 https://t.co/C3XNpDUPV4 via @jewishjournal

2022-12-09 20:40:16 @AbsDelfuego @Ostrov_A @nadplo @Palestine_UN @EUpalestinians @ShaharAzani @benshapiro @COLRICHARDKEMP @CUFI @ActForIsrael @ReinsteinJosh Palestinians to Israelis: How dare you resist your elimination!

2022-12-09 20:03:31 It's reassuring and humbling to know that, four years after publication, there are still researchers who are discovering the beauty of #Bookofwhy for the first time. This means that AI is not doomed for stagnation

2022-12-09 13:28:48 The history of Arab's rejectionism, its depth and its totality, are well documented in this article by E. Karsh: https://t.co/DSlNRVjAHe. It contains some interesting facts that are new to me, like the quotes by Nasser and Abu Iyad. https://t.co/LFWc4T7HIr

2022-12-09 12:54:15 @Aaron37913713 @Claire_V0ltaire @EinatWilf I've found it especially interesting for a different reason: How can such a simple and obvious truth escape the eyes of almost ALL Western observers, commentators, and peace makers, including the current administration?

2022-12-09 12:45:43 Thanks for having me. https://t.co/QWbqcMIHhv

2022-12-09 06:44:15 @Raamana_ @GaryMarcus @ylecun I would need to learn the rules first.

2022-12-09 06:18:23 It's great they decided to listen to students! Next, I hope, they will listen to faculty! https://t.co/JQpOphlQQ9

2022-12-09 06:06:06 Upset by neighbors who vow to exist, International pressure they pledge to resist, Would Mr. Putin please join to assist? https://t.co/VBbE5iqCsV

2022-12-09 05:13:40 I lost tract

2022-12-08 13:00:00 CAFIAC FIX

2022-12-07 08:00:00 CAFIAC FIX

2022-11-14 00:54:31 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn If you throw causal theories into the "inductive bias" salad, you obscure it from seeing the light of day, and you make it harder for ML folks to ever distinguish what works from what doesn't.I advise readers to stay away this obstruction.

2022-11-13 23:39:56 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn Yes. Submit the Firing Squad story to an "inductive bias" machine and ask it some questions, like "would the prisoner still be dead had soldier-1 not shot?" etc. etc.

2022-11-13 23:34:12 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn If I understand your question properly you are asking whether we should understand what wea are doing. Ans. Yes.

2022-11-13 23:22:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn The apparent speed with which you go very far with inductive biases is a result of traveling in a flat terrain of Rung-1 curve fitting, far from the bumpy terrains of Rung-1&

2022-11-13 20:58:16 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @KordingLab I agree with everything you say except the last, which I would reverse to read: DT is not the most suitable language for formulating and conducing CI, since both its basic assumptions and its methods of analysis are better represented in SCM (or CBN).

2022-11-13 20:52:49 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @KordingLab I agree with everything you say except the last, which I would reverse to read: "causal models are needed to justify the incomprehensible "weak ignorability" judgments.

2022-11-13 20:48:30 @KordingLab @scac1041 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can call the rules of arithmetic and the whole scientific method "inductive bias" -- no objection, as long as you don't expect significant scientific problems to be solved by the broad brushstroke methods of "inductive bias" used so far in ML.

2022-11-13 20:35:14 It took almost a century for statistics to snap out of "Causation is just a species of correlation" (Pearson) to "Correlation is not causation". I hope it takes less for Machine Learning folks to snap out of the "inductive bias" mantra and examine some causal inference tools. https://t.co/cwLr4PWgCb

2022-11-13 19:50:53 Summarizing what I've learned in the past two days about "inductive bias". The "prior information" needed is, of course, assumptions about the causal relationships involved. These assumptions are problem-specific and, hence, could not be encoded as "inductive bias" https://t.co/ZWOpHSo0hY

2022-11-13 19:38:37 @ylecun @KevinIndrebo @GaryMarcus @KordingLab @scac1041 @PhilDawid @pmddomingos @StephenPiment @AlexTensor @gopnik @stephensenn @davidchalmers42 In addition to Inductive bias, and its "no-free-lunch" theorems, which lift us from finite samples to distributions, we also need prior information to properly translate our research questions into the language of distributions - a non-trivial task, not addressed by induction.

2022-11-13 19:08:29 To our Spanish readers: This is your best introduction to the Arab-Israeli conflict -- well researched, well documented and, most importantly, well reasoned. https://t.co/aJn1hjCWvz

2022-11-13 12:37:28 Finally, a crisp characterization of anti-Zionists, free from antisemitic ambiguities. Valid even if you are fighting antisemitism tooth and nail, or visit a Holocaust museum twice a week. https://t.co/pUbW1gyrIx

2022-11-13 09:42:05 @pmddomingos @ylecun @GaryMarcus @KevinIndrebo @KordingLab @scac1041 @PhilDawid @StephenPiment @AlexTensor @gopnik @stephensenn @davidchalmers42 @De_dicto Data "can" tell us everything when available.The role of causality is to encode assumptions that would make up for data unavailable, eg, data from actions and experiments. I would hate calling such assumptions "inductive bias" lest folks would think I'm a "catch-all evader".

2022-11-13 08:46:17 In commemoration of the 2nd Yahrzeit of Rabbi Lord Jonathan Sacks, I am re-tweeting my reflections on the life and legacy of this great man: https://t.co/k24s0HNoRr"Stay the course, and be true to yourself" he told me.

2022-11-12 23:23:27 Do you concur with @ylecun's statement: "asymptotically, with infinite data, there is no need for priors." ?? If so, does it also apply to"inductive bias"? I hope you say Yes, because that would clarify the source of my confusion. https://t.co/VLfJPDohNc

2022-11-12 20:34:57 @AlexTensor @KordingLab @scac1041 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @gopnik @stephensenn Even "initial assumptions" is too vague. When we say "assumptions" we need to specify: assumptions on what? On the structure of the NN? On the distribution of the data? On the world behind the data? The word "knowledge" behooves you to be clear on these distinctions.

2022-11-12 20:29:13 @ylecun @GaryMarcus @KevinIndrebo @KordingLab @scac1041 @PhilDawid @pmddomingos @StephenPiment @AlexTensor @gopnik @stephensenn @davidchalmers42 @De_dicto "So, asymptotically, with infinite data, there is no need for priors." This sentence may hold the key to the "inductive bias vs. knowledge" debate. The latter is needed even asymptotically, e.g., no amount of data can tell us that the barometer does not cause tomorrow's weather.

2022-11-12 13:49:59 @gottfriedmath @attilacsordas Well put! Pre-Dawinian theories of purposeful evolution show our preference toward causal explanations over selection mechanisms, which even Reichenbach missed.

2022-11-12 13:42:02 @PhilDawid Believe me, I feel much more comfortable discussing Abraham, whom I know from childhood, than discussing inductive bias, a new arrival on my block.

2022-11-12 13:24:28 I love it. Especially the history of Israel in 7.5 seconds. https://t.co/bO52NcmtNO

2022-11-12 12:49:53 @Ostrov_A Hard to decide between you and Nancy Pelosi. She said: "Zionism is the most noble liberation movement in the 20th century." I can't think of any counterexample to either statement.

2022-11-12 12:24:50 @benj_kerstein Abraham might say: No, all I wanted is to check consistency between God's action and the moral order which He himself devised.

2022-11-12 12:14:55 @RonKenett Puzzling: "And God saw every thing that he had made, and, behold, it was very good." Even all-knowing God could not foresee the result of his action, else he wouldn't be surprised with the outcome. Lesson to us: No matter how smart, check your results.

2022-11-12 12:02:25 @PhilDawid Abraham did not ask "what if I knew that..." He asked "what if there were". Beside, God has made a decision already. So, Abraham knows that, factually, the number is below threshold. Asking for decision-changing conditions is thinking about conditions that contradict known facts.

2022-11-12 10:20:06 @KordingLab @scac1041 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn These are some of the reasons I prefer "Knowledge" over "bias", which often goes to "gimmicks" or "hacks", like preferring one structure of NN over another, not knowing whats behind the preference except "it works".Compare it to "knowing" say that aspirin tends to cure headache.

2022-11-12 06:39:39 @AvivaKlompas @SAFECUNY Progress, but falling short of explaining the moral deformity of BDS, as Martha Polack did (President of Cornell):“[BDS] often conflates the policies of the Israeli government with the very right of Israel to exist as a nation, which I find particularly troublesome.”

2022-11-12 06:25:37 @KordingLab @scac1041 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn This is the first clear definition I received of what "inductive bias" is - a concept that has always eluded me, and that I prefer not to use, now that I know it's just "prior knowledge". Why? I can taxonomize "knowledge" by its origin, type and representation. Not "bias".

2022-11-12 03:21:46 @KordingLab @scac1041 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn It requires information, yes, but calling any required information "inductive bias" is not serving any purpose, unless it tells us what's special about this information, where it can be found etc.

2022-11-12 03:08:24 @KordingLab @scac1041 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We know that observations+pertub are enough for rung-2 tasks, but I have never been able to see how things like "inductive biases (like smoothness), taken totally from the world of prediction, can move us an inch higher.

2022-11-12 03:01:49 @MatthewEGunter @cleverclue @johncutlefish I do not characterized my work on causality as Bayesian. (I actually wrote a paper "Why I am only half Bayesian" https://t.co/gqeQb2lNBh). I consider the two to be orthogonal, nor can I understand how causality could be viewed as constraints over NN

2022-11-12 02:50:25 @KordingLab @scac1041 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn Sure. When I "listen to" X, I do f(X) regardless ofwhether X is perturbed externally, or changes spontaneously or stays constant.

2022-11-12 00:28:24 @scac1041 @PhilDawid @KordingLab @pmddomingos @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn I do not regard a causal DAG as a primitive. The primitives are the "listens to" relationships with the help of which we construct the DAG. These provide the "deeper foundation" needed.

2022-11-12 00:25:31 @soboleffspaces @PhilDawid It would be additive if Abraham were to ask: What if we ADD 50 righteous person to the city.But Abraham did not talk about addition, only about "if there were".

2022-11-12 00:22:05 @PhilDawid Abraham believed, I presume, that God could count and therefore God knew the true number of righteous people in the city, call it X. So, what's the point of asking "What if X=50". Obviously, X was not equal 50. Hence counterfactual.

2022-11-12 00:10:46 @attilacsordas The #Bookofwhy summarizes Reichenbach "Principle of Common Cause" and its faults, on page 199. If you think we missed anything of it, or did not fully appreciates its significance, I would love to re-discuss it. Today, having a language of causation, we can do it meaningfully.

2022-11-12 00:03:20 @mundt_martin @GaryMarcus I am always supportive of bridging sub-fields but, humbly, the description I read does not help me believe that continual learning can "enable knowledge transfer... when distributional shifts are experienced". Is it demonstrable on a toy example, say fig.1 https://t.co/dEPwcv3l40

2022-11-11 11:31:35 It's the first counterfactual "what if" in the Bible

2022-11-11 11:04:58 @mundt_martin @GaryMarcus I am not familiar

2022-11-11 10:44:44 For those who read my "Miracles of November" https://t.co/zJunAizIhdand were wondering how the 75th anniversary to the UN Vote, Nov. 29, 1947, will be celebrated, here is the information:A community event is planned for November 28, 11:30 am at the Museum of Tolerance, LA.1/2

2022-11-10 09:07:48 Another paper deserving our discussion is Weinberger's "Path-Specific Discrimination"https://t.co/mrvYuo3iouwhich illuminates debates about the meaning of "fairness" or "discrimination" in the context of causal theories of mediation, and path-specific effects. https://t.co/4DNzVZuaNt

2022-11-10 05:25:49 @herdiants Unconvincing. Population-specific interventions are routine in SCM, requiring no change in the meaning of counterfactuals.

2022-11-10 02:57:15 Congratulations to professor Jeffrey Herf, upon receiving the 2022 Bernard Lewis Prize for his book "Israel's Moment: International Support for and Opposition to Establishing the Jewish State, 1945-1949" (Cambridge University Press, 2022) https://t.co/5dCdlh4Us5

2022-11-09 23:58:33 @JKugelgen Seeing the prisoner dead, I wish to conclude:"Had soldier-1 not shot, the prisoner would still be dead". Regardless of whether soldier-1 turned disobedient or his rifle got jammed. I do not believe the subjunctive mode is appropriate for evidential (backtracking) conditionals.

2022-11-09 14:52:45 Israel unearths ivory comb from 1700 BCE inscribed with plea to stop lice https://t.co/AVbR7bS9kM

2022-11-09 14:13:05 @ruescasd Figuratively saying "breaking the laws of nature" really means "changing state of the world". See Firing squad example in #Bookofway, eg, "had soldier-1 refrained from shooting".

2022-11-09 13:45:32 Liberal-minded readers who are as disturbed as I am by the right-swing outcome of the Israeli elections, should find this paper useful in countering the recent dooms-day interpretations offered by Jewish Studies professors who never reconciled to the idea of a democratic Israel.

2022-11-09 13:45:31 The Jewish Studies Professors Who Traffic in Antisemitism https://t.co/sKsRrVRgNj via @jewishjournal

2022-11-09 12:59:38 RT @causal_science: And that brings the Causal Data Science Meeting 2022 to a close. Thanks to all the speaker, panelists, @csilviavr and @…

2022-11-09 12:53:10 of breaking the laws of nature, we modify the initial conditions. I am still trying to convince myself that this is a reasonable alternative in some situations, but I'm missing a compelling example. Perhaps some philosophy-minded readers could help us here??3/3

2022-11-09 12:53:09 It's time again to examine this week's production of causal inference publications:https://t.co/KZQcRpCAb4. Among the many inviting titles (5 1/2 pages), my eyes zoom on "Backtracking Counterfactuals"https://t.co/OSIeP8gNYbwhich formalizes an alternative mode of 1/3

2022-11-08 09:41:29 Can't wait to see this tablet in a museum, commemorating the successful completion of King Kezkiya tunnel, having been reading this story many time in schools and re-living it while visiting Jerusalem 2012. Oddly, I can hardly read a word, and I thought I spoke Hebrew. https://t.co/M2EbDwrYdy

2022-11-08 09:26:23 RT @GeraldNGOM: Nov 9: German government political foundations are exploiting and distorting #KristallNacht - the 1938 brutal Nazi German p…

2022-11-08 07:00:41 @nathanb_science I am missing the effect of What?

2022-11-08 06:34:49 I purposely selected this semi-amateurish performance of "Eli Eli" which, unlike the usual concert-hall performances, reminds me of the many evenings we (as teens in Israel) used to sit around the camp fire, tell stories, fall in love, and sing our hearts out: "Eli Eli". https://t.co/Xyx5IYN0jw

2022-11-08 05:41:26 78 years ago, on November 7th, 1944, Channah Szenes was executed.You are probably familiar with some of her poems, but the one below, Eli Eli, became more than a poem

2022-11-08 05:03:33 @JohnHMorris @jhurwitz @stu_frost @ddkirsch @GeminosAi I am inclined to agree with this statement, with emphasis on "today". This does not mean that AI will not bridge this difference tomorrow.

2022-11-08 04:40:55 @cjpberry The unprincipled claims we are hearing from data-fitting enterprises (eg, ML, DL) remind me very much of pre-Socratic debates in anient Athens -- "it seems to work, hence..." Socrates method urges debaters to start with agreed principles, then proceed to the conclusions.

2022-11-08 04:22:54 @JohnHMorris @jhurwitz @stu_frost @ddkirsch @GeminosAi Part of the answer is that, in science and in everyday life, we carry the model in our heads, implicitly, so it appears as though new wisdom comes from data. But when the entire inference process had to be made explicit for dumb robots, we've found that models are badly needed.

2022-11-08 04:14:47 @fisagor I'll vouch for that!

2022-11-07 23:42:50 WATCH: Iranian girls fight back, refuse to believe teacher's lies about 'Zionists' | World Israel News https://t.co/DakZHINjC8

2022-11-07 09:27:51 We are spending all our resources fighting a hate that the haters claim is hard to define. If we spent a tiny fracture to fight anti-Zionism, which every 10 year old can define, and which ferments at least 50% of the hostilities, we would save a fortune, and make some progress. https://t.co/iGk7cay85v

2022-11-07 00:36:59 @diomavro Misspelling correction. It should read:They (i.e., the courts) don't count "de-confounders" as "expert witnesses"

2022-11-07 00:34:09 Thanks for reposting, doubly relevant today, with thoughtless "Jewish scholars" urging the UN NOT to accept the IHRA definition of antisemitism, and misguided Jewish professors telling the @latimes that: Israel is no longer what our parents imagined it could be. Read Avot 1:11 https://t.co/n50Gwevw2M

2022-11-06 21:49:40 Video clip from Rabin's memorial in Jerusalem.President Herzog 15:20 min, PM Lapid 28:40https://t.co/8diSiSmPZ4These are Israel's answers to all the scavengers who have gathered to bemoan her democracy, devour her legitimacy, or diminish her spirit of perpetual rebirth.

2022-11-06 21:09:11 To Itzhak Rabin, murdered Nov. 4th, 1995.Every person has a namewhich his sins gave him,and which his longing gave him.Every person has a namewhich his enemies gave himand his love gave him.Every person has a namewhich the sea gave him,and which his death gave him. https://t.co/rzZnv7Hod3

2022-11-06 20:59:23 @diomavro The don't count "de-confounders" as "expert witnesses"

2022-11-06 14:08:10 On Nov. 4th, 1995, Yithak Rabin was assassinated. To mark this day, I've chosen to share the song "Every Person has a Name", sung by Hava Alberstein https://t.co/Nmnk3uhRtvEvery Person has a Namethat God gave himand which his father and mother gave him

2022-11-06 11:04:45 @Michael_Wgd Sadly, 128 so called "Jewish studies scholars" submitted a statement to the UN, imploring it NOT to adopt the IHRA WDA. The "Jewish studies Zionist network" was formed toprevent such hair-splitting to ruin the reputation of "Jewish studies" proper. See https://t.co/i5SFind4vy

2022-11-06 09:15:16 @Michael_Wgd @sunwins888 @WatsonSpeaking I can't figure out how any Jew can be "fake" when Jewishness is a state of mind. Either you have this state, and you identify with the collective memories of the Jewish people, or you don't. What is there to fake?

2022-11-06 07:26:34 @sunwins888 @WatsonSpeaking @Michael_Wgd Not sure what I can add to the conversation, except a reminder that Jewishness is peoplehood first, religion second and skin-color blind.

2022-11-06 06:34:16 The received wisdom in the sciences is that the notions of "transparency" and "meaningfulness" are subjective, hence should not be dictated.Wrong. We can easily point to judgmental tasks that PO folks would find impossible to perform, resulting in erroneous causal conclusions . https://t.co/YEhJfowwfI

2022-11-06 06:11:23 @soboleffspaces Rubin's proclamations have had no negative effect on my students, but they have set back statistics and econometric education for at least30 years. I have colleagues at UCLA who are still teaching that way.

2022-11-06 03:56:01 @soboleffspaces This shock can be seen as one of the worst educational consequences of the Potential Outcome framework. By refusing to deploy cognitively-compatible models of reality, it blinds people from detecting fallacious statements about reality, let alone making them.

2022-11-05 22:40:38 Pelosi Cites Israeli Poem in 1st Public Appearance Since Attack on Husband https://t.co/wdQCBrEO3u

2022-11-05 20:17:54 Recent discussions on IPW with different sets of covariates call my attention to a 2014 paper on "confounding equivalence" https://t.co/jxWdLuUKZP which, sadly, has not received the attention it deserves. (Only 46 google citations - shame!). It tells us when one covariate 1/2 https://t.co/X3zng6uZhZ

2022-11-05 19:51:21 @AsraNomani @Columbia We had a professor at the Technion, who used to scold us: "You behave like Asiatic!". It never occurred to me that, some day, I would be defending the honor of my Asian heritage, together with my Muslim sisters from India.

2022-11-05 19:36:26 @soboleffspaces @RaulMachadoG @spphubc It's stronger! Inverse-treatment-probability weighting is equivalent to stratification and averaging (the formula produced by do-operator) if stratification variables are the same in both, regardless of whether they satisfy the identifiability criterion. https://t.co/f2oN7JW60N

2022-11-05 19:23:06 @policytensor @_alice_evans I couldn't find out what "sort of thing" you refer to, nor who "They" are. Someone must have deleted the Tweet.

2022-11-04 15:23:07 The problem with taking proactive security measures is that, eventually, you will be blamed for being the aggressor. Israel has been under such dilemma for 75 years. https://t.co/I4qok2ftoK

2022-11-04 00:56:38 A brilliant lecture presented last week at UCLA, by professor Walter Russell Mead https://t.co/uQiBpc1Fgmwhich should guide and educate us, as we are preparing to celebrate the 75 anniversary of the historic UN Vote of Nov. 29 1947, https://t.co/f394qsXg5h.@DovWaxman, @Ostrov_A

2022-11-03 10:20:59 RT @MepinOrg: I just returned from #SaudiArabia &

2022-11-03 08:14:43 @LekhtNaya Well put. And the worst violator of this principle is @jstreetdotorg with their offensive slogan "pro-Israel, pro-peace", as if being pro-Israel does not automatically entail being pro-peace. As if one must go to great effort to find peace-loving Israelis. What an insult.

2022-11-03 06:59:27 @RaulMachadoG @soboleffspaces @spphubc The paper shows that many causal models, not merely backdoor types, have computationally efficient "inverse probability type" estimators.

2022-11-03 06:46:47 RT @HillelNeuer: BREAKING: After powerful remarks by @NazaninBoniadi, the U.S. announces before the United Nations that it will lead the ef…

2022-11-03 06:44:36 Having checked #Columbia Professor @JLeeSoc's recent papers, I am not sure she is aware of modern definitions of the phrase: "may have also benefited from their racial status" in the causal inference literature. @eliasbareinboim works on causal fairness are relevant here. https://t.co/WsIh3QTwxf

2022-11-03 06:35:01 Give me a platform like the UN and I'll find you 300+ Jewish "scholars" who would rush to reject any definition, of any concept, from motherhood and apple-pie -- this is what the Talmud teaches us to do

2022-11-03 06:17:27 RT @LahavHarkov: Making waves in the Hebrew media/commentariat: Right is leading the Left by under 10,000 votes, but has 20 more seats in t…

2022-11-03 05:57:18 Confirmed. I've watched Mehdi Hasan the past 5 years. He is no lesser bigot than David Duke or Itamar Ben-Gvir. He is definitely more dangerous, given his saintly platform on @MSNBC. https://t.co/FVFUOAcuwK

2022-11-03 05:49:49 The logic is simple. If Israel is indeed as evil as @amnesty describes her, then given that most Jews support Israel makes Jews either hopelessly morons or dangerously evil, both deserve harassment by the lunis. https://t.co/gbmIzkYPtM

2022-11-03 05:35:31 RT @elderofziyon: Libel inflation. https://t.co/l6BdsrtoLB

2022-11-03 05:33:36 RT @sooresraphael: A true revolutionary song, sang in a true revolutionary act, by a true revolutionary generation. It captures the soul o…

2022-11-03 02:45:42 RT @HillelNeuer: In April, we kicked Vladimir Putin off the U.N. Human Rights Council. Now we're about to kick his friend Ayatollah Khamene…

2022-11-02 21:35:56 The Balfour Declaration of 1917, which committed Britain to support the creation of a Jewish homeland, came about in large part due to behind-the-scenes efforts by three women, writes historian Natalie Livingstone. https://t.co/gFTz8m1dLx via @WSJ

2022-11-02 20:47:29 @cto_maverick Another reason: "everybody is fascist" said @leibowitzadak, so he must be right. https://t.co/ysrgq1F5co

2022-11-02 19:46:48 In my little corner, I haven't seen #ML and #DL claiming X

2022-11-02 19:11:38 Finally! Perfect timing! Today, Nov. 2, 2022, exactly 105 years after the Balfour Declaration of 1917, https://t.co/9XOCOsqkIT,Israel's neighbors found a reason for rejecting co-existence. Makes sense. https://t.co/Lt9gH2cy9x

2022-11-02 19:01:27 True, but the main point of the Balfour Declaration was that place of residence in the PAST should be weighed against historical attachment at the PRESENT. The Arabs had more of the former, the Jews had more of the latter, therefore, SHARE. https://t.co/9XOCOsqkIT https://t.co/HJLWzR0AJr

2022-11-02 08:22:40 RT @yudapearl: Why is he so happy? Good question! But put yourself in his shoes: When was the last time you've found a three-line proof to…

2022-11-02 07:58:39 Readers who question the causal connection between the Balfour Declaration (Nov. 2, 105 years ago) and the rescue of 400K Jews from Hitler's Europe, should note that most of these Jews were able to come only after a state-like infra structure was laid in Palestine1/2 https://t.co/xpA4gF0QSV

2022-11-02 05:52:54 RT @HillelNeuer: BREAKING: Dutch Senate adopts motion requesting its government “to call on behalf of the Netherlands for Iran's removal…

2022-11-02 05:37:21 RT @drninaansary: 17-yr-old student #SoheilKhoshdel who was arrested during the#IranProtests has been sentenced to death by the Islamic Re…

2022-11-02 05:37:09 It also means that those who do not believe in "both do" must wait till they do. https://t.co/apujZuXwuz

2022-11-02 05:21:14 Though I rarely use the charge of "antisemitism" against Zionophobes, today's vote of the LA City Council labels many of my UCLA colleagues "antisemites" (primarily for denying the Jewish right to self determination), thus giving students the nerve to call them out in class.

2022-11-02 05:21:13 Los Angeles City Council Passes IHRA Definition of Antisemitism https://t.co/suUgKuhEkB via @jewishjournal

2022-11-02 04:11:18 Every time I see this derivation I fall in love again with mathematics. I wonder how ML folks can go through life without napkins. #Bookofwhy https://t.co/IuukYrmDHB

2022-11-02 03:53:43 @HumzahAlkindi I don't know if you are speaking from personal experience or hearsay. The "living together" fable is paved with bloody reality. (My wife has experienced the latter).

2022-11-02 03:49:02 @_ChrisAlbertyn @IbrahimZina @KCLBHI @KingsIoPPN @daniel_stahl_01 @richdobson @z_abdulla01 @HDR_UK @AcademicChatter @OpenAcademics @NHSDigAcademy Oh God! This looks like my handwriting! You caught me in poetic mood, thanks.

2022-11-02 03:47:13 @soboleffspaces @spphubc You are underestimating the grip and duration of academic loyalty oaths

2022-11-01 20:52:15 She should have been suspended 2 years ago, on a much stronger and more provable charge: anti-Zionist form of racism. https://t.co/ZbA8rhMsTW

2022-11-01 08:34:17 Sorry, I meant "for supporting Balfour in 1917".

2022-11-01 08:29:38 I literally owe my life to the Balfour Declaration. It encouraged 400K Jews to escape inferno Europe and immigrate to the holy land between 1917 and 1939. Among them was my father who, by virtue of being a British subject (1924), was able to go back to Poland, marry my 1/2 https://t.co/kdA5nwWsJS

2022-11-01 06:43:15 @soboleffspaces @spphubc If you are asking "Why", I have a soft spot for the question. My answer is: (1) Because Rosenbaum and Rubin did not know about backdoor in 1983 and (2) Because they signed their disciples on a life-long loyalty oath, so that backdoor will remain a mystery for generations.

2022-11-01 05:57:25 RT @AsraNomani: Born in India, I was on an emotional roller coaster today in the Supreme Court, listening to 3 justices + 4 lawyers try to…

2022-11-01 05:44:26 @Claire_V0ltaire They will not be tired, as long as we refrain from calling them "Zionophobes". Only then they would wake up to ask: "What have we done wrong? or stupid? or totally thoughtless?" It's our responsibility to give them a mirror.

2022-11-01 05:35:17 @soboleffspaces @spphubc Moreover, if the covariates do NOT satisfy the criterion, IPW does not reduce nor increase the (asymptotic) bias over what we would get by stratification on the same covariates.

2022-11-01 05:27:51 @artistexyz I have long been forgiving to people who are allergic to DAGs

2022-11-01 05:21:14 Opinion | Opinion | Is Israel really a settler colonial state? https://t.co/jQcfMGCNAC

2022-11-01 04:01:35 @titudeadjust @Popehat @GaryMarcus @kahneman_daniel @R_Thaler Can't tell what we agree on.

2022-11-01 03:56:38 My comments pertain to the last section, titled Causality and Computer Science, in whichhe essentially tells economics students: Oh, BTW, there has been some work done outside of economics, but it has merits only in computer science, where there are "large number of 2/4

2022-11-01 03:56:37 More CI papers have been added to our harvest at https://t.co/XFohTn8OJG. Deserving our attention is Imben's CAUSALITY IN ECONOMETRICS: CHOICE VS CHANCE,a printed version of his Nobel lecture, thatshould always be read with the correctionslisted in https://t.co/sZPpL7Ifmg1/4 https://t.co/3O9w8vMOpT

2022-11-01 01:06:40 Happy to share a beautiful concert at Kibbutz Maagan Michael, Israel, celebrating the Daniel Pearl World Music Day, by Ensenble Consart (Leipzig, Germany) https://t.co/amBbmC23QSThe song performed is "Shalom Aleichem," a traditional Friday night song written in the 16th Century

2022-10-31 22:36:58 Another Monday's harvest of papers dealing with causal inference https://t.co/XFohTn8OJG. Catching our attention is Woodward's "There is no such thing as statistical explanation", though based on quick reading, Woodward is yet to internalized the First Law https://t.co/etASSEk4Or

2022-10-31 17:24:44 @artistexyz @stephensenn @eliasbareinboim Enticed by your post, I've read the "Lord Paradox" again, both in the #Bookofwhy and in https://t.co/MSJJVipnXE. Agree, it's beautiful and clear (specially when each dining hall has only one diet.) I'll stay out of the new debate -- no point spoiling clarity for soiling sake.

2022-10-31 16:55:20 at the Museum of Tolerance. Mark it on your calendar, please, Sunday Novemeber 29,and wait for final details and program.Happy Halloween and a meaningful November.4/4

2022-10-31 16:55:19 This coming week brings us an inspiring event,the 105 anniversary of the Balfour Declaration, Nov. 2, 1917 -- the first international recognition of the Jewish people right to a homeland.I have written about its historical significance in 2017, https://t.co/9XOCOsqkIT, 1/4

2022-10-31 09:47:44 @WSB_OfficialMod @HarvelaHarvey @noatishby The 1967 border can only be respected when Arabs recognize the right of Israel to exist IN ANY BORDER. As long as Palestinians, to the last person, are committed to obliterating Israel given the opportunity, it's merely a cease-fire line. For facts: https://t.co/XSVDjqgHlL

2022-10-31 09:06:07 @WSB_OfficialMod @HarvelaHarvey @noatishby Another role of the word "Zionophobia" is to shift the conversation away from the ugly manifestations of the conflict (borders, settlements, terrorism, occupation etc.) to its very roots: Arab's uncompromising rejection of a Jewish state in ANY border: https://t.co/XSVDjqgHlL

2022-10-31 07:51:55 @GeraldNGOM @theintercept @amnesty @KenRoth @OmarSShakir @jrosyfield meant it to be a parody of how the "apartheid hyenas" of @KenRoth and @amnesty are emptying the word from content. Tomorrow's parody: Israelis sell land to Israelis, Palestinians can't - apartheid! Tuesday's: Palestinians are paid to kill, Israelis r not - apartheid!

2022-10-31 07:12:08 RT @ksadjadpour: Students at Hormozgan University in Bandar Abbas (historically a more socially conservative region of Iran) tear down the…

2022-10-31 07:04:59 @guntax59 @user14201 @Avishay3000 @shubham19881 @ShaharAzani @Israel @benshapiro @Ostrov_A @ActForIsrael @IsraelinNewYork @CUFI @noatishby @edencohentweets @IsraelinLA @StandWithUs This is where a formal theory of causality comes in

2022-10-31 06:46:15 "Zionophobia" changes the conversation entirely from*AZ: Zionism is inherently anti-Palestinian*NT: It's a mistake, I'm not ...to*NT: Anti-Zionism is inherently racism

2022-10-31 02:14:55 @Minnhawk1 @shubham19881 @ShaharAzani @Israel @benshapiro @Ostrov_A @ActForIsrael @IsraelinNewYork @CUFI @noatishby @edencohentweets @IsraelinLA @StandWithUs Religion played an important role for both Pakistanis and Palestinians (1947). The former sidestepped the extremists and compromised on partition, the latter were led by a religious fanatic (The Mufti) who fabricated the "Al Aqsa in danger" fable, and foiled the partition option

2022-10-31 01:30:56 @VladTheSlav5 @ChesterMorton96 @ShaharAzani @benshapiro @Israel @Ostrov_A @ActForIsrael @IsraelinNewYork @CUFI @noatishby @edencohentweets @IsraelinLA @StandWithUs "He was a fighter" Palestinian logic formalized. 1. X attacks Y.2. Y defends itself (mandatory service, reserve service, even women and seniors.) 3. X declares men and women legitimate targets (children too -they grow to be soldiers)4. Y: Enough!5. X: War crimes! Apartheid!

2022-10-30 23:42:50 @Minnhawk1 @shubham19881 @ShaharAzani @Israel @benshapiro @Ostrov_A @ActForIsrael @IsraelinNewYork @CUFI @noatishby @edencohentweets @IsraelinLA @StandWithUs Interesting question. Both Pakistanis and Palestinians fought British colonialism in 1947 and succeeded in getting international charter for independence, through partition. The former embraced the partition plan, the latter didn't, and decided to liquidate their neighbors first.

2022-10-30 14:48:51 Comment # 1 on @EinatWilf video https://t.co/XSVDjqgHlL. The rationale given by the Arabs (in 1920's) for prioritizing their war against a Jewish State over all other considerations was not fears of dispossession by Jewish immigrants, but a profound ideological1/2

2022-10-30 13:53:18 @Crazy_fox65 @backforhowlong @EyalTweet @ShaharAzani @Israel @benshapiro @Ostrov_A @ActForIsrael @IsraelinNewYork @CUFI @noatishby @edencohentweets @IsraelinLA @StandWithUs Me too, with both eyes.

2022-10-30 13:21:53 @ShaharAzani @Israel @benshapiro @Ostrov_A @ActForIsrael @IsraelinNewYork @CUFI @noatishby @edencohentweets @IsraelinLA @StandWithUs Warped ears in the West have long been programmed to translate Palestinian acts of terror into heroic acts of freedom fighting. How ready Palestinians are as "partners for peace" is better betrayed by their intellectuals, now openly dancing: "From the river to the sea".

2022-10-30 12:53:38 @soboleffspaces @RaulMachadoG @AlfredoMorabia This would be an insult to Epi literature, for not reading Yule (1900), who also advocated stratification in order to block potential confounders. I'm sure many epidemiologists used stratification in the period 1900-1973 to remove the bad influence of potential confounders.

2022-10-30 12:41:34 @anndvision The premises in SCM are those fragments of knowledge that you consult when you decide that no arrow should go from X to Y. Namely, Y listens to other variables, not to X. "Listening" is a "unit-level" (rung-3) concept. Interventional premises are averages: https://t.co/PqATlrRh91

2022-10-30 01:33:32 @soboleffspaces @RaulMachadoG @AlfredoMorabia I am willing to modify my summary sentence. Would you be happy with: "Here, for the first time, Miettinen is introducing the concept of 'blocking', and Epi has not been the same since

2022-10-29 23:37:09 @KordingLab @metrics52 I don't see where the generalization comes in. Is it that we have SEVERAL problems, P_1, P_2...? or is it that we have both P(y|do(x)) and P(y|x) ??Do we need more than one DAG to explain what your method is doing? Recall, one DAG can represent many problems and many realities.

2022-10-29 22:20:43 @RaulMachadoG @soboleffspaces @AlfredoMorabia Thanks, @raulMachadoG and others for sending me links to 1970's papers. But I was waiting for a summary sentence: e.g., "Here, for the first time, Miettinen is introducing the concept of xyz to epidemiology, and Epi has not been the same since

2022-10-29 18:44:21 @a_sarentorbic @bsgallagher @GaryMarcus Such characterization makes me suspicious that you haven't read #Bookofwhy.

2022-10-29 17:48:09 @GaryMarcus A cause-effect question. Do we appreciate Kant because we are here, or should we read him to get further ahead of where we are?

2022-10-29 17:30:09 @anndvision The word for this is "logic". Formal logic does not tell us what the world looks like, it merely tells us that, if you make premises 1 and 2 about the world, and observe data D, then conclusion C is inevitable. In our case the, premises are causal, rather than propositional.

2022-10-29 11:17:16 @AnnGreenberg @noatishby But that has not prevented campus climate from becoming what it is, and has not equipped students with the weapons to tell their bullies: "You sound like a Zionophobic racist! Don't tell me you are one!" Bullies smell weakness.

2022-10-29 10:28:31 Something is missing. These ignorant students who are regurgitating populist slogans at @noatishby do not even begin to suspect that they are engaged in less than super-righteous activity, let alone a racist movement with genocidal goals. The word "Zionophobia" is badly missing. https://t.co/JxldS1coQu

2022-10-29 10:13:30 RT @RabbiWolpe: In the ‘box score’ dept… https://t.co/bSffUZ7WdM

2022-10-29 09:55:22 OTD. On October 29, 1969, 53 years ago, the internet was born at UCLA. https://t.co/ylPiShf4G8I was hired only 3 weeks earlier, and I didn't quite understand what the people on the floor below me were celebrating. Now I understand.

2022-10-29 09:21:01 @AlfredoMorabia @soboleffspaces So, can I assume that my account of causation in epidemiology, narrated here: https://t.co/GIaEtSH0y3, is not missing a major conceptual breakthrough, although I skipped many prominent names. (I may still have a chance to add, if convinced otherwise.)

2022-10-29 08:42:41 @AliAbunimah, the founder of Electronic Intifada sides with Putin. Not surprising. The Russians and the Palestinians are the only two peoples who decided that their neighbors do not have the right to exist. Both fight for ONE-STATE solutions, and deem co-existence as injustice. https://t.co/5VM5PYi3kI

2022-10-29 07:25:38 The only explanation I have why @EinatWilf's video https://t.co/XSVDjqgHlL does not garner >

2022-10-29 01:28:35 @RaulMachadoG @soboleffspaces @AlfredoMorabia The paper is under a pay-wall but, if you think it oughtt to be part of the discussion, please describe Miettinen's contribution in modern language? I can't penetrate it on my own.

2022-10-29 00:28:54 @KordingLab @metrics52 I am just trying to understand your setup. If it is just drawing samples (y,x) from the P(y|do(x)) distribution, then obviously ML would take you from finite sample to the distribution. To comment, I need input-output - a terrible weakness I acquires in my Engineering schooling.

2022-10-29 00:18:18 @AlfredoMorabia @soboleffspaces @ken_rothman I believe I've read this article at the time, but it confirmed the perspective I had while writing Causality, in 2000. Do you think I've missed important conceptual milestones in my recent review: https://t.co/GIaEtSH0y3, ???

2022-10-28 23:44:13 @KordingLab @metrics52 Only one input and one output? How about effect of drugs on the weather, or education, or seatbelt usage? Is there a place for piece of knowledge: "drugs do not change a person's sex" ??

2022-10-28 23:36:43 @totteh Hilarious! If it were not serious. This means that readers on our channel would have to start faking it and talk like PO parishioners, eg, "treatment assignments" "target trials" "death to Rung-3" etc. I feel manipulated already.

2022-10-28 23:23:42 @DKedmey Please do nominate me to Twitter's "content moderation council" but, warning, my first recommendation: Zionophobic content is moderated same as "child pornography"

2022-10-28 22:51:38 @KordingLab @metrics52 I don't care either whether it's technically correct but, to anchor my thoughts, I need to know what "prior causal knowledge" is, how it is encoded, what data is taken, etc. tec.To anchor our thoughts, can we talk firing-squad?(I forgot my transistor circuits class).

2022-10-28 22:14:48 @KordingLab @metrics52 For me to "think", I need an input-output description, hopefully in SMC language. What is assumed, what kind of data is available, what is the output? Eg. Can we deduce the firing squad model by observing a zillion executions?

2022-10-28 15:46:11 @AlfredoMorabia @soboleffspaces At least no one was able to tell me what it is - perhaps you can? BTW, what was "the standard definition of confounding" in Epi prior to 1986 ? I hope it was not statistical, because there isn't any, see Sec. 2 of https://t.co/7lUwca9KP02/2

2022-10-28 15:33:28 @AlfredoMorabia @soboleffspaces Hi @AlfredoMorabia, Boris was referring to your upcoming talk at Harvard, celebrating Miettinen's paper of 1976. So, I asked: "Have I missed Miettinen's contributions when I wrote my historical account of causation in epidemiology?"So far, I do not see that I've missed any1/2

2022-10-28 15:09:51 American friends often ask me: What is it about Israel/Palestine that you, Israelis, think you know and that we don't. I cannot think of a better answer than this video clip of @EinatWilf https://t.co/XSVDjqgHlLPlease don't miss a word, even if you think you know and we don't.

2022-10-28 13:36:26 Dear @FranceskAlbs Those who deny Israel's right to Tel-Aviv, are hardly in position to lecture us how she can defend herself in Ramallah. https://t.co/VBKzJq9txN

2022-10-28 12:49:03 Check out this article: https://t.co/DLV49zoUZd

2022-10-28 09:47:35 @kanyewest has given bad name to antisemitism through his incoherent and blunt racism. Much more dangerous are the soft spoken "apartheid hyenas" at the UN, whom many people, mostly the gullible and uneducated, tend to parrot. https://t.co/w5p9GDhSdV

2022-10-28 09:20:53 @soboleffspaces @AlfredoMorabia My question to those who read Miettinen: Did he state conditions for the set Z to make the adjustment formula correct, or merely wrote down the formula by intuition. I haven't read him, assuming (perhaps prematurely) that, lacking notation, he did the latter only. Did I err?

2022-10-28 05:26:45 @soboleffspaces @AlfredoMorabia My question remains, if you have no notation for causal effects, no do(x), no Y_x, we are back in Yule's days

2022-10-28 03:18:49 Curious. How did Miettinen handle stratification without causal notation? I have credited the scientification of epidemiology to Greenland and Robins (1986), https://t.co/NMja3mjhtK,seeing that Miettinen tried to define confounding using probabilities. @AlfredoMorabia https://t.co/Yn5mCilXmR

2022-10-28 02:53:01 We need a stronger word, perhaps "Jew-cleansing" to describe "Jew washers" like Mehdi Hasan, who get terribly indignant when people point them to this glaring hypocrisy. https://t.co/TOx6NEdi1W

2022-10-28 02:34:02 @AleksanderMolak @rasbt However, I would not call it "implicit", because in my mind I have a clear arrow going between X=water and Y=tree-growth. I therefore conjecture that we extract causal knowledge from language by filling in the missing parts of a preprogrammed template (eg, DAG) = curiosity.2/2

2022-10-28 02:25:36 @AleksanderMolak @rasbt Most of my causal knowledge came from language. My father taught me why we need to water the orange trees, and how often. My Dr. told me to exercise or else. There is no reason therefore that LLM will not obtain causal knowledge from text sprinkled with causal utterances.1/2

2022-10-27 23:37:48 @itamarcaspi @DuduLagziel @giladmp I would also recommend the paper: "Understanding Simpson's Paradox" https://t.co/GpSKmVcKR7which goes beyond #Bookofwhy, and resolves the "paradox" from both decision making and cognitive perspectives.

2022-10-27 20:42:13 @xiaodai10838475 The word "biased" is a compliment to a Zionophobe.

2022-10-27 11:39:06 No Mehdi, it's got nothing to do with your being or not being an antisemite. No, you are a carrier of a much uglier affliction: Zionophobia - a moral deformity that cannot be blamed on alcoholic parents or abusive upbringing

2022-10-27 10:02:44 Glad this Tweet was re-liked, for it has become timely again in light or the new @ylecun - @GaryMarcus debate. I'm not sure though whether this definition of "Deep Understanding" would meet universal acceptance in "Deep Learning" land. https://t.co/Ik0Q4pi8Ie

2022-10-27 08:53:38 I don't know about you, but I get emotional watching high profile meetings such as this: https://t.co/w6v9QXUQuH. Just watching two great nations rising above the cynics and re-asserting their unique greatness makes my heart melt.

2022-10-27 08:21:29 @arieljalali I am not taking position here on whether hate mongering or child pornography should or shouldn't be allowed on Twitter and other social media channels. I am merely requesting that Mehdi's Tweets on anti-Semitism should be given equal treatment.

2022-10-27 07:03:05 @blakeflayton Once you baptize yourself with the holy waterof anti-Zionism, people might begin to take you seriously, even when you have nothing to say.

2022-10-24 05:14:40 @soboleffspaces @ObservStudies Your take-aways are fine for all practical purposes but, philosophically, @HarvardEpi would fight you tooth and nail when you remind them that the do-operator operates on a model of reality (SCM) while the g-formula emerges from a mysterious FFRCISTG =?? @EpiEllie #Bookofwhy

2022-10-24 04:45:19 @policytensor @bpetershome OK, peace on earth, but please warn me next time that your "regression" comes from you.

2022-10-24 04:43:07 @policytensor @bpetershome I remember doing tensor analysis in grad school and I bow to Einstein's theory of coordinate independence, but causal models do depend on choice of variables. Here I show how the direction of statistical time is reversed by choice of coordinates: https://t.co/Ty53coWZrI (11.2)

2022-10-24 04:30:19 @policytensor @bpetershome I was misled by the term, sorry. This is what happens to you after fighting 3 decades against the regression/structural confusion in the econometric (and statistical) literature. (eg, https://t.co/CEpVusa5eB) I would using suggest the term "linear SCM", to avoid confusions.

2022-10-24 04:24:00 @calimagna @noahholl @FrankRHutter I don't see why we need to guess about the gain that SCM gives them when we have given them a simple tool to diagnose and pinpoint the source of their gain: Run a toy problem, and see.

2022-10-24 04:14:55 @policytensor @bpetershome I follows your arguments, but not the language. In my mother tongue, regression is a statistical notion having nothing to do with the causal structure. Similarly, I do not see where "tensors" come in. Could be y weakness.

2022-10-23 19:27:27 @calimagna @noahholl @FrankRHutter This is what we are trying to understand on a simple 4-variable toy problem. Take a correct SCM and a fake one (of same size) that is equally compatible with the observed distribution. Would the former lead to better predictions?

2022-10-23 18:50:44 @noahholl @FrankRHutter @calimagna Here comes the beauty of toy-problems

2022-10-23 12:41:53 @Ostrov_A @StandWithUs @SAFECUNY SHSHSHSHSH!!!! Don't tell the @nytimes!!!

2022-10-23 12:05:47 A new film is out, "Why Ukraine", by my friend and mentor Bernard-Henri Levy @BHL. Don't miss! Oct. 27. At the UN. https://t.co/Qj5FBJ0Nsc

2022-10-23 10:35:58 @StandWithUs @SAFECUNY The looks of it warns me - it was probably stolen from Palestinians

2022-10-23 07:40:19 @mhd01 @AsraNomani @BloodBrief @fcpsnews Thanks. I hope I haven't used it innocently, as an endearing adjective, on anyone.

2022-10-23 07:32:39 Great to get cultural content on Twitter once in a while - stay tuned. https://t.co/14JyB7FVIS

2022-10-23 07:02:06 RT @elderofziyon: It might be time to start working on my book of cartoons. https://t.co/wspX6qAtXy

2022-10-23 06:59:09 Recalling personal experience, Reza Aslan is not a person I would trust. Not surprised he is now working for the Iranian regime. https://t.co/3o9G6gRTLu

2022-10-23 06:52:08 Should be in the tool-kit of every instructor. https://t.co/2ubgENwJ1x

2022-10-23 06:49:11 @AsraNomani @BloodBrief @fcpsnews I'm 62 years in America and can't figure out what the R-word is. Please educate a fellow immigrant.

2022-10-23 05:17:45 @erikbiz @SAFECUNY @AdamSandler @DavidSuissaJJ Agree, honest and simple. Which makes you think: Shat does it take for celebrities like @BarbraStreisand or @SaraSilverman to be honest and simple and come out of their silent-Zionist closets.

2022-10-23 02:21:53 @FJnyc @jjz1600 @arnoldroth The Mongol raiders, so I am told, complained bitterly when the Great China Wall was erected: Apartheid! Apartheid! They shouted. "If this ain't apartheid what is?" Western analysts echoed, and echoed, and echoed.

2022-10-23 01:31:47 @skdh I've noticed a sense of comradeship in your tweets. They will not replace us.

2022-10-22 22:48:52 @mvanhout71 @wv012 "Human anatomy" in healthcare translates into "intrinsic properties of target population" in social science, invariant across backgrounds and aspirations of participants.

2022-10-22 21:33:21 @FrankRHutter Speaking basic principles of "rung 1.5", why should causal information help us improve predictions? Once we understand the principle, we should be able to quantify how sensitive prediction quality is to causal misspecifications.@calimagna

2022-10-22 21:13:46 @mvanhout71 @wv012 My Dutch is a bit rusty, but I hope you touched on these considerations: https://t.co/A2MguL54zE

2022-10-22 20:22:10 @wv012 Off hand, not getting into the details of the habits, my answer would be YES. Why? Because associations are less stable than causal relations

2022-10-22 10:05:01 Jewishness is based on "Amech Ami, Ve'Elohaich Elohay" ("Your people are my people and your God is my God" - Ruth to Naomi). Meaning: "Peoplehood first - Religion second". https://t.co/2P8y0tfKVF

2022-10-22 09:55:39 @AsraNomani @BOSCityCouncil @Quillette What an Orwellian inversion. #MahsaAmini must be turning in her grave to see the hijab, which to her (and to majority of Iranian women) meant a symbol of women subjugation celebrated as a symbol of free expression in a US city called Boston, home to MIT and Harvard.

2022-10-22 09:40:31 @HenMazzig There is one logical flaw in your argument, @HenMazzig, the word "share" does not exist in Palestinian vocabulary. When it comes to Israel, "share" is replaced by "stolen", and "I choose" is now "I'm a victim".

2022-10-22 09:32:07 @Eve_Barlow @ArielElyseGold Inspired by @Eve_Barlow https://t.co/w5kr5QfDFN

2022-10-22 08:51:08 @AleksanderMolak @rasbt I am trying to understand what you mean by "implicit heuristic world models". If this is something that is learned by the agent, then it must be stored someplace, in some code. That code then is explicit. What makes it different from "explicit world model"???

2022-10-22 08:27:52 I was about to swear this is a joke ridiculing our "apartheid parrots". It isn't! Seeing @ArielElyseGold, I reckon these parrots still exist, seriously squeaking and chirping: "Me too, Me too, Dying to be original, dying to be provocative, Me apartheid parrot too, Me too!" https://t.co/V8QfAhBqGD

2022-10-22 07:34:13 RT @richard_landes: https://t.co/LWy0ssOfOL

2022-10-22 07:31:11 @CitalanEdward @WealthInc247 It was my first exposure to the philosophy of science, dissolving my guilt for not understanding speculative philosophers like Kant and Hegel, and leading me to Carnap, Hampel , Goodman and Popper.

2022-10-22 04:53:28 @tbednall @WealthInc247 It changed my life too, forcing me to re-examine long-held assumptions from a totally new perspective.

2022-10-22 04:50:12 @WealthInc247 Hans Reichenbach "The Rise of Scientific Philosophy".

2022-10-22 03:59:00 In Rubin's framework, ignorability conditions are assumed not because they are deemed true, but because they are needed to support a claim

2022-10-22 03:58:59 We used to have Rubin's disciples among our readers (eg, Yale economists), perhaps they can help us interpret the analogy. I assume he refers to the fact that SCMs appear to be redundant because they provide, not only ignorability conditions, but also their justifications. 1/3 https://t.co/sAbb8m9yVV

2022-10-21 17:33:21 @penguinvondoom @GaryMarcus @hajivat @nirsd Forgot what "cool RPG premise" stands for.

2022-10-21 10:16:29 @PHuenermund Agree with the insights. I will soon post a cleaned up copy, with unmolested references.

2022-10-21 10:14:02 @AvivaKlompas And when I point out that Zionophobes are suffering from a moral deformity, friends tell me "don't get down to their level".

2022-10-21 09:49:01 Today, all Palestinian factions have declared a "Day of Rage" in memory of Super-Martyr Uday Tamimi (22), who murdered Noa Lazar 11 days ago.I am used to Days of "Rage", but I can't get used to Noa's eyes and smile, which are haunting me for 11 days:https://t.co/qZEpzX6Kz3 https://t.co/LS1fvK4IaU

2022-10-21 09:27:17 @jeremy1174 I thought you understood what Zionism means: Coexistence as equally indigenous peoples. The disrespectful word "entity" tells me that you don't.

2022-10-21 09:14:51 @jeremy1174 Nothing would please me more than seeing our Palestinian neighbors greet this Ethiopian boy as a comrade and close kin. That is precisely what my grandfather expected from our cousins.

2022-10-21 08:56:41 The only thing missing from the video is whether this man listens to Al-Jazeera or Mehdi Hasan on@MSNBC https://t.co/U4lV6tKKi8

2022-10-21 08:48:55 Family lore has it that, when my grandfather arrived at Jaffa port, in 1924, he likewise fell down and kissed the ground. This makes me a comrade and close kin of this Ethiopian boy, transcending all age and background differences. https://t.co/pEOhTPTyWB

2022-10-21 08:41:39 @RaulMachadoG @soboleffspaces @ObservStudies The criterion: "50% chance that the claim is true," sounds statistical until we realize that the "claim" itself is causal, and renders statistics totally helpless when it comes to legal responsibility.

2022-10-21 08:27:56 @hajivat @GaryMarcus @nirsd The #Bookofwhy was translated to 14 languages, showing in chapter and verse how the belief that all information is in the data fails. Yet this has not prevented data-centric enthusiasts from erecting taller and taller cathedrals to a handcuffed God.

2022-10-21 01:51:00 @AviMayer No need to get upset, it's just a matter of semantics. When I first heard (in 2000) that the land my grandfather bought in Bnai Brak (1924) was "stolen", I realized that "stolen" in Arabic means "cultivated". Thus, Hadera swamps, Tel-Aviv, Falafel, Humus, Debka, are all "stolen".

2022-10-21 00:41:17 Great question. I did some thinking about it.See Section 2.8 on "Statistical Time" here: https://t.co/xfevdJms8r. Especially last paragraph. Also, Section 11.2 in https://t.co/Ty53coWZrI, on "Reversing Statistical Time". https://t.co/4A4d0QBrRe

2022-10-20 21:24:55 @AI_AM001 @isaacdecastrog To the best of my knowledge no one denies Iran the right to exist. Enjoy.

2022-10-20 18:52:12 @soboleffspaces @ObservStudies Anyone who rejects the First Law would naturally see "causes of effect" to be a "cocktail conversation", including legal responsibility and scientific explanations (moon and tides). How else can you define "but for"?. I bet Heckman would think likewise, rejecting the First Law.

2022-10-20 11:57:20 @AleksanderMolak @rasbt That's why I prefer toy problems

2022-10-20 11:15:39 Journalism Award Stripped from Palestinian Journalist Over Antisemitic Facebook Posts https://t.co/d95kdYYiIM

2022-10-20 10:29:22 @isaacdecastrog Some "writers" think that their arguments become convincing if they sprinkle foul adjectives around, marinated in the cesspool of Al-Jazeera. Surprisingly, some of their readers love that stench.

2022-10-20 10:18:41 My take-aways from Rubin's and Heckman's interviews @observStudies match precisely what I anticipated, as expressed in my interview, even though it was completed 2 years ago, and we were not permitted to see each other's texts. https://t.co/N8vmMp6Mxo

2022-10-20 08:08:24 @Claire_V0ltaire We have also seen Jewish scholars expressing their irritation, some of them inadequately:https://t.co/HkeihG1qBd@michaeldickson @Ostrov_A @EinatWilf @CotlerWunsh @SSI_Movement @BarbraStreisand @SSIUCLA @vardi @CAMERAonCampus @bariweiss @JGreenblattADL @deborahlipstadt

2022-10-20 07:10:48 @AleksanderMolak @rasbt If it's a "trick" then I dare to bet that it would fail on the 3 toy examples in Fig. 1 of https://t.co/dEPwcuLbPS What gives me the audacity to bet w/o even understanding the trick? Simple. The 3 examples generate the same data, while the correct answers are different. Smart!

2022-10-20 06:12:10 @HenMazzig Would you object if I were to replace "antisemitism" with "anti-Zionism", and say out laud: Anti-Zionism isn’t okay, period.You see, Hitler gave a bad name to antisemitism, so people (including antisemites) would rush to join your statement

2022-10-20 05:18:36 I beg to respectfully disagree with Waxman's 2nd interpretation of Trump's statement, the "Inversion of Dual Loyalty", according to which saying that American Jews SHOULD care about Israel amounts to making "antisemitic assumptions". Americans are mature enough to understand 1/3 https://t.co/IkUGrBlvkL

2022-10-20 02:57:36 @rasbt I havn't heard of the "hard, practical truth which consists of retraining one or more layers on a different dataset." Is this "truth" or "trick"? Do we have a theory to guarantee the outcome? Better yet, can we try it on the 3 toy problems inFig. 1 of https://t.co/dEPwcuLbPS ??

2022-10-19 22:06:45 @KirkDBorne I was not familiar with Barber's book, fairly impressive!And I truly love the quote from https://t.co/YmnsZPM3tL"Pearl believes that current AI is mostly just glorified curve fitting, and that adding causal intelligence to current AI will be necessary ...".Amen!

2022-10-19 21:25:58 @AmandaKMontoya @b_snef @rlmcelreath So, UCLA Psyc dpt has a future after all - best news I got from campus in months. Hoping our paths cross.

2022-10-19 18:56:20 @rasbt To the best of my understanding, "transfer learning" is not an "approach" but a goal, a desideratum. Also to my understanding, causal inference is the only approach that achieves this goal (via selection diagrams,) see https://t.co/KBGKcmJxM8, https://t.co/E1ap1N4ZCh

2022-10-19 10:51:50 RT @MNehzati: I'm finally out of Iran and feel safe enough to talk about the situation publicly with an account bearing my name.The crackd…

2022-10-19 10:38:59 Who said computer science is for nerds only? https://t.co/k7EG24IV3v

2022-10-19 10:12:11 @Claire_V0ltaire @AviMayer, you are the last person I know whom @mehdirhasan has not blocked yet. So, please answer his question for me: No! Mehdi, you should have disqualified yourself from talking anti-semitism, your persistent delegitimization of Israel, renders you blatantly incompatible.

2022-10-19 08:26:11 From students mouths, this article describes Berkeley Law School as "State Sponsored Humiliation." No Jew is excluded, of course, but they must sit at the end of the bus. Dean Chemerinsky should acknowledge that: (1) "legal" doesn't make a "right", (2) "wrong" is his to undo.

2022-10-19 08:26:10 We’re Jewish Berkeley Law Students, Excluded on Campus https://t.co/8Nv69R880o via @thedailybeast

2022-10-19 07:05:15 I love this quote from Imbens and Rubin, because it reflects the attitude of the entire PO culture. It says: "Some people find graphs useful in their applications but, in our applications (read: everything on earth), we found them "unhelpful"." That's what PO worshipers repeat. https://t.co/b3uN2SEit4

2022-10-19 06:48:52 @soboleffspaces @prem_k @Anuraag_Shukla @anustup_nayak @KhoaVuUmn @leecrawfurd @singhabhi @eliasbareinboim @_MiguelHernan Rubin began his career at Educational Testing Service. Imbens and Rubin book (2016) is titled "CI in Stat. Social, Behavioral &

2022-10-19 05:20:29 @AmandaKMontoya @b_snef @rlmcelreath Great, we should meet one day. Do you teach causal mediation/moderation (e.g.. https://t.co/SkPFSdPnl3) or stick with Barron/MacKinnon/Hayes ?

2022-10-19 05:09:22 @HumanBeingIHope Equating "Giving a Voice to the Palestinian People" with "Me and Hitler are one" makes a mockery of that voice, monsters of the Palestinians, and thoughtless of @thomsonreuters

2022-10-19 04:57:17 @robwilliamsiii I have been influenced by many philosophers of science from Locke and Hume to Suppes and Reichenbach. But I would not call myself a card-carrying philosopher, because that would behoove me to toe the line and refrain from professing that Suppes and Reichenbach were wrong.

2022-10-19 04:41:41 @AmandaKMontoya @b_snef @rlmcelreath Please do. Are you at UCLA Psych? Normally they don't hire causal inference folks, unless theyhave had a change in tradition.

2022-10-19 03:39:05 @NewsHour @johnyangtv Most people think that Mehdi Hasan and Linda Sarsour are using this sudden Jew-caring rhetoric to stain Trump with antisemitism. Wrong! Their aim is to stain Israel with Trumpism.

2022-10-19 01:23:29 I'll be happy to act as an information-dispatcher of causal inference related job openings, like the one you posted here: https://t.co/2Ik5cjQVAQ . My only requirement, the post must mention "causal inference" explicitly and credibly. (There are many impostors around). https://t.co/ljW6HfBjew

2022-10-19 01:02:21 "Me and Hitler are one" [Reuter's honoree]

2022-10-18 21:24:28 Another blessed feature in your post: You resisted the temptation of including outdated books that claim to be causal, but are graph-tabooed, e.g., Imbens-Rubin (2016), Angrist-Pischke (2014), Gelman-Hill (2006). Why outdated? See: https://t.co/pYrfhD1GID https://t.co/j7IDEsMklc https://t.co/KOsc41sztk

2022-10-18 10:41:45 @prem_k @soboleffspaces @Anuraag_Shukla @anustup_nayak @KhoaVuUmn @leecrawfurd @singhabhi @eliasbareinboim @_MiguelHernan My mind does not retains applications, but I am sure you can find many works in education in the lists of #CI works I post here weekly. Perhaps some readers can also direct you to such works.

2022-10-18 10:32:58 I like your blog post https://t.co/hVEAW5vCQWI can see how a new sailor, confused by the ever expanding terminology, would find use for your map, as well as the 6 causal books described, so as to sail straight to the target research question at hand or to a new buzz-word1/2 https://t.co/Q7ZFg7NkDB

2022-10-18 10:15:22 Great picture of Haifa. Imagine studying 4 years in this city - who can study Maxwell's Equations if you can look down the balcony and watch this bay? https://t.co/I1bZcxcSve

2022-10-18 10:05:55 Australia finds itself in a strange company. If any of my decisions was praised by Hamas and Islamic Jihad I would think twice if I am on the right side of history. Australians will. https://t.co/bY1XaO7nju

2022-10-18 08:55:10 RT @TheHistoryOfTh5: Iraq has opened a park featuring 2,700-year-old Assyrian carvings showing kings praying to the gods. They date from th…

2022-10-18 08:45:29 @HenMazzig What they fear most is your disruption of their mythical narrative of how Muslims and Jews have been living together as good neighbors, for thousands of years, until - you guessed - until the State of Israel was established in 1948. (My wife was from Baghdad).

2022-10-18 08:36:03 Most people think that Mehdi Hasan and Linda Sarsour are using this sudden Jew-caring rhetoric to stain Trump with antisemitism. Wrong! Their aim is to stain Israel with Trumpism. @Ostrov_A https://t.co/FW2PHTS65Z

2022-10-18 06:30:59 I read at least three of them -- really good! https://t.co/Q7ZFg7NkDB

2022-10-18 02:49:42 I must confess there were two significant differences from the time I was took this bus.(1) I did not have a helmet - unheard of. (2) We weren't allowed in the back door

2022-10-17 22:47:16 Another Monday - another flood of papers on causal inference for your consideration: https://t.co/mYtQaaVncg

2022-10-17 22:28:42 @HenMazzig Stay the course, Hen, and may your storyilluminate the world with a wider understanding of what Israel has accomplished that her predators wish to suppress.From strength to strength.

2022-10-17 21:56:43 @Eve_Barlow @AvivaKlompas @MehdiHasanShow Me too! How about starting a "Mehdi Hasan Me Too" movement? Membership would then become a litmus test for things he can't tolerate --- minimal understanding of the Middle East.

2022-10-17 20:55:26 @MissDiagnosis Great! Let me know I am willing to advise them on how to resolve the Israeli-Palestinian conflict. The US State Dpt. does not heed my advice, perhaps the Saudis will.

2022-10-17 13:36:27 a priori handcuffed by allegiance to the Potential Outcome framework, or to the quasi-experimental paradigm of Angrist and Imbens, but explores the foundations of structural equations. Disappointing is the chapter on "ceteris paribus" which fails to reach the First Law, and 2/3

2022-10-17 13:36:26 A third paper with attractive title is: Models, laws, and evidence – philosophy of causal research in macroeconomics, https://t.co/2ZVzJgt6NL PhD dissertation at the University of Warsaw.It is refreshing to see a treatise on the philosophy of macroeconomics that is not1/3

2022-10-17 12:25:40 Another title that caught my attention is: On Explaining Confounding Bias https://t.co/hiw5Y7muf7Again, I am unable to understand the problem, now due to absence of a graph to specify what we assume, what is observed, what we suspect, and what kind of "explanation" we seek.

2022-10-17 11:00:46 I would take very seriously what Dr. Qanta Ahmed is telling us about the transformation of #SaudiArabia. Qanta is not a PR agent

2022-10-17 10:41:56 the ideas surveyed, and how they are related tothe theory of transportability and data-fusion. While the paper invokes graphs, d-separation, do-operators and SCM's , it is totally oblivious of selection-diagrams and their complete implications. An enigma -...??? #ML 2/2

2022-10-08 15:41:06 @ChristinABuelow I have never suffered from this problem, perhaps because my to determination to naively pursue what appears to be "the most important problem in existence" has always produced publishable results. I even got tenure with this naïve approach.

2022-10-08 15:20:31 @ChristinABuelow I am grateful to you for telling me that #Bookofwhy is helping readers better understand ourselves, and the world in which we live. I hope the book continues to thrill you, as you proceed backward, from Chapt. 10 to Chapt. 1. And don't let the harsh truth stop you.

2022-10-08 14:41:49 @KouMurayama @artistexyz @georg_henning I hope this paper convinced you. What's the point in saying "causal inference" is a "missing data" problem, if we do not have a calculus for the latter?

2022-10-08 14:34:46 @Ostrov_A @KombatKlutch @BecomeOblivion @NKippur @spikypotatoface @KfirDrake @ShaharAzani @DoctorNazarian @benshapiro @DrMikeH49 @elderofziyon @ActForIsrael @TheModerateMan @AWiderBridge @StandWithUs I've only used it twice

2022-10-08 14:26:07 @spikypotatoface @ShaharAzani @KombatKlutch @BecomeOblivion @NKippur @KfirDrake @DoctorNazarian @benshapiro @DrMikeH49 @elderofziyon @ActForIsrael @Ostrov_A @TheModerateMan @AWiderBridge @StandWithUs I think you are underestimating the gullibility of good people, especially in the West. This is how we got the "apartheid" culture to propagate on college campuses by innocent well-meaning students. I simply don't have the time to ridicule every stupid allegation on this screen.

2022-10-08 14:07:12 @KombatKlutch @BecomeOblivion @NKippur @spikypotatoface @KfirDrake @ShaharAzani @DoctorNazarian @benshapiro @DrMikeH49 @elderofziyon @ActForIsrael @Ostrov_A @TheModerateMan @AWiderBridge @StandWithUs Apologizing to other readers on this thread, but I must mute Mr. @KombatKlutch. [My 2nd mute in the past 4 years]. When malicious lies are manufactured in a coat of "when I went to tel aviv" they become palatable and I can't allow this platform to assist their spreading. Sorry.

2022-10-08 01:40:05 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor @KordingLab Glad we are converging. The only issue left is to agree along what dimension we should judge one language more "suitable" than another. I have suggested several such dimensions in https://t.co/mm9LnNtBCk, but am willing to examine others.

2022-10-08 01:32:42 @Bijun_med Merry sailing. And share with us the joys of learning.

2022-10-08 01:31:32 @pmddomingos Not enough, sure. But if we agree that "generalization" is necessary for AI then it makes sense to go to a peaceful corner of the wood where generalization was solved, and see if we can borrow a few tricks.

2022-10-07 20:26:49 @KouMurayama @artistexyz @georg_henning The other way around. "Missing data" is a causal inference issue as shown here: https://t.co/1BNZSxBPCw

2022-10-07 16:26:33 Interesting, but in our corner of the woods we have perfect consensus about what generalization means and how to test it. https://t.co/E1ap1N4ZChMoreover, our understanding of its scope and limit is in some sense complete. Perhaps the NLP discord can be alleviated by 1/2 https://t.co/SotQxxS3yW

2022-10-07 15:36:19 I join UN Watch in calling on Secretary Antony Blinken @SecBlinken to oppose the appointment of Francesca Albanese as a special rapporteur on Palestine.https://t.co/09XJzIM3OB@HillelNeuer @Ostrov_A https://t.co/wQ7SOBhwZG

2022-10-07 10:07:09 I can't lift my eyes from the facial expression of the rooster on the left. His urge to project virtue and sing louder than other chicken in the barn could be admirable, if it weren't for the tragic end that awaits the thoughtless. https://t.co/lzYz8tpl0G

2022-10-07 09:34:45 RT @elliecohanim: “It is unfathomable for the Biden administration at this same moment that the regime is crumbling to send them a signific…

2022-10-07 03:24:21 @KordingLab @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor ?? decipher please ??

2022-10-07 02:29:40 We are waiting anxiously for Al-Jazeera @AJEnglish to describe Ahmed as "settler colonialist," and for Berkeley's Queer Caucus to accuse Israel of weaponizing gender tolerance to hide its unholy conception. See background: https://t.co/uIF2QR1tdj https://t.co/CvH68J5XH7

2022-10-07 02:06:36 @PhilDawid @pmddomingos @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor @KordingLab I may have misinterpreted what @PhilDawid truly meant to propose, but I hope we agree on the golden rule:To predict the effect of actions we must start with either experiments or causal judgments. Calling the latter "Statistical DT" does not negate its causal content.

2022-10-07 01:53:54 @KordingLab @tiagopeixoto @pmddomingos @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor I am an odd ball then, b/c I learned Alef-Bet from my mother and how to grow citrus trees from my father, and electromagnetic theory from Maxwell

2022-10-06 13:06:29 @biorankings @UCBerkeley Legally, I presume it can. But morally, banning an ethnic group and humiliating the ethnic group by enshrining the ban into Bylaws may not be compatible with what the Bylaws of "Equity Diversity and Inclusion" call for.

2022-10-06 12:10:22 A nice gesture by Professor Jeff Lax, to speak to the nine @UCBerkeley law student groups who vowed (in their byelaws) never to listen to him and his like. I'm more inclined to favor the approach of "rescind or else", proposed here:https://t.co/s0CnzUX0rx by 150 organizations. https://t.co/uymiNAN8u3

2022-10-06 10:50:45 RT @NGOmonitor: NGO Monitor President @GeraldNGOM calls on the @UN_HRC President and High Commissioner @UNHumanRights to put an end to the…

2022-10-06 10:41:47 RT @FrimetRoth: https://t.co/NHrGCZe2cg

2022-10-06 09:06:42 @pmddomingos @AlexTensor @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus @tommmitchell A semi-classic reference by JP and @eliasbareinboim which shows mathematically (as well as through concrete problems) what kind of information is absolutely needed for generalizing experimental findings: https://t.co/dEPwcuLbPS

2022-10-06 07:45:35 @pmddomingos @AlexTensor @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus I must take some time off this interesting thread. Might come back when I understand "inductive bias" in engineering terms, i.e., input-output.

2022-10-06 07:37:33 @pmddomingos @AlexTensor @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus I would love to see one such assumption which can be articulated without appealing to causal vocabulary. Recall, any such assumption is provided by a human programmer/researcher, so it must be comprehensible to 90% of readers on this Twitter channel.

2022-10-06 07:17:54 @pmddomingos @KordingLab @tiagopeixoto @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor We agree on the last point: AI will be solved by combining empiricist and rationalist ideas. What the causal revolution has added to this abstract agreement is a framework demonstrating a concrete symbiosis of the two, armed with division of labor and performance guarantees.

2022-10-06 07:09:40 @KordingLab @pmddomingos @GaryMarcus @AlexTensor @StephenPiment @ylecun @aminkarbasi Can you describe the input-output properties of the system (judgments, data, conclusions, etc.)BTW, I forgot the existence of my paper https://t.co/UQFQrCh3Im, thanks for the reminder. It's worth reading.

2022-10-06 06:57:43 @pmddomingos @AlexTensor @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus I have at least proven that (1) to predict effects of action we must have either experiments or causal judgments (2) To process counterfactuals (eg identify causes of effects) we must have functional relations (eg SCM), and (3): Given (1), a complete algorithm exists for Rung-2.

2022-10-06 06:40:14 @KordingLab @tiagopeixoto @pmddomingos @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor Simulated evolution provides indeed a theoretical possibility of climbing the Ladder's rungs, but it's wisdom and practicality are questioned here: https://t.co/BsNCQuIVuf along 4 dimensions.

2022-10-06 06:32:01 @pmddomingos @AlexTensor @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus Back to the original question: is causal inference necessary for AI? There are (seemingly) other options. Example: We take causal graphs and call them Neural Networks, we take causal inference engines and call them "inductive bias", finally, we tweet: "Victory, everything is NN".

2022-10-06 06:11:00 In view of some misunderstandings of the boundary bet Rung-1 and Rung-2 of the Ladder of Knowledge, I am RT-ing this thread, hoping to make the boundary crisp and unassailable to all readers: To climb Rung-2, data must be enriched with experiments or judgmental causal information https://t.co/IcYoG4zupE

2022-10-06 05:47:52 @pmddomingos @AlexTensor @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus 2/2 The Second Simpson's Paradox in https://t.co/hDUeA1LFEkI need to understand the reason for your resistance to the causal resolution of SP, which now provides a general calculus for resolving any problem instance that may be encountered. 2/2

2022-10-06 05:37:24 @pmddomingos @AlexTensor @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus We may differ in what counts as "resolving" Simpson's paradox. For me, "resolving" means explaining how for heaven's sake it's possible for two opposing decisions to emerge from two identical datasets that differ only in the causal structures of their generators. It's called:1/2

2022-10-06 04:11:43 @pmddomingos @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor @KordingLab augmented with "action indexing" variables that mirror the causal information conveyed by the starting graph. "No causes in - no causes out" (Nancy Cartwright, 199?)Notice that I keep referring to "causal information" as a conserved quantity

2022-10-06 04:01:39 @pmddomingos @GaryMarcus @StephenPiment @ylecun @aminkarbasi @AlexTensor @KordingLab It is a mistake to interpret Dawid's video as saying that causal inference is not necessary for predicting effects of actions. I've written an editorial comment on Dawid's arguments https://t.co/mm9LnNtBCk -- his framework begins with a causal graph, and ends with a graph1/2

2022-10-06 03:43:42 @AlexTensor @pmddomingos @StephenPiment @ylecun @aminkarbasi @KordingLab @GaryMarcus Agree. You can't solve Simpson's paradox without causal information, and certainly not with "logic" alone. The linked video calls it "context" - a pre-scientific term used by statisticians to explain why same data dictatetwo different actions depending on the causal structure.

2022-10-06 03:34:20 Many reciprocal thanks to our Muslim cousins who are reaching out to us through interfaithactivities, people-to-people, building a peaceful world for our children and grandchildren. https://t.co/rrbj6mqeP4

2022-10-01 05:22:20 @blakeflayton There is a huge difference between Israel and the war in Ukraine. Putin claims ownership on only four regions, while Palestinians go "from the river to the sea" - the whole cake. Westerners are generally critical of Putin, but find the Palestinian claim cute, if not attractive.

2022-10-01 02:10:44 @rustbenisrael @SDrrizzy I am well aware of the stories behind The Vote.The real hero was not Truman but his former business partner Jacobson and dozens other Jews worldwide, including of course the Yishuv leaders in Israel who vowed: "We will have a state even if they vote NO."

2022-10-01 01:32:25 I still owe you a piece I wrote in 2008 "The forgotten Miracle: Nov. 29, 1947"https://t.co/f394qsXg5hDescribing some of the stories behind the miracle and why it behooves us remember them. https://t.co/pB7qPsVro1

2022-10-01 01:18:41 @AlexTensor @ylecun @artistexyz @pabbeel I am referring to "domain knowledge" also known as "world model". How is it provided, how is it represented, what must the provider think about when providing it, etc etc.

2022-09-30 21:55:09 Here are the color slides:https://t.co/Dde0z2Sn4rand here is the accompanying video:https://t.co/z7N6RH3SjsFrom Deep Learning to Deep UnderstandingHappy sailing! https://t.co/rifmUISwcm

2022-09-30 20:53:24 For readers wishing to see what a November 29 celebration looks like, here is what we did in 2012: https://t.co/E57cCWFznHBut even a more modest event, re-enacting the Vote, reading the UN Resolution, and giving thanks to friendly communities would make for a memorable event. https://t.co/oUmvz2rjZ0

2022-09-30 20:38:37 I've defined "understanding" here: https://t.co/z7N6RH3SjsIt is a formal definition, once you tie it with the Ladder of Causation, where the capabilities of each layer is formally defined. https://t.co/iNkHGJybiQ

2022-09-30 13:44:06 @AlexTensor @ylecun @artistexyz @pabbeel This still does not tell me what judgments, if any, are required of the researcher at the onset.

2022-09-30 13:36:27 @DavidHirsh Beautiful blacks have "racist" designation.Proud gays have "homophobe" designation.Proud Muslims have "Islamophobia" for hostile designation. Only we, proud Zionists, have no fighting word. This forces us to use "antizionism", which sounds almost as legitimate as Zionism.

2022-09-30 13:19:31 @DavidHirsh We have allowed antizionists to construct "Zionists" in such detestable terms by not presenting them a mirror, to see themselves for what they are, in much uglier terms: Zionophobic racists. These reptiles grew up without a mirror, believing themselves to be mankind angels.

2022-09-30 12:56:03 @MikeNew19512115 @AvivaKlompas The 1930's Judenrein Zones were established when Jews stood with trembling knees. Today's Jews can turn these racist enclaves into academic No-fly Zones. All it takes is leadership: https://t.co/eI2GeC48bZ

2022-09-30 10:08:07 My faithful calendar says we are two months ahead of November 29, the day in which history will celebrate the 75 yrs anniversary of its monumental miracle, the UN Vote to partition Palestine into a Jewish state and an Arab state.https://t.co/zJunAiQLjd Two months from now, 1/2

2022-09-30 09:28:30 @GaryMarcus @danbri @ylecun My experience has been the same, in the sense of realizing that scaling up without understanding can be disastrous and, simultaneously, that mathematics does have the power of bringing about a state understanding and spotting the pitfalls of would be disasters.

2022-09-30 06:52:36 @emilykschrader @Ilhan @Ilhan can't get it to her skull that, to most women in the world, especially to Iranian women, Hijab is an icon of "no choice," though she enjoys Western privilege of wearing "by choice". I therefore suggested she take it down for 60 second, in public, just to show solidarity.

2022-09-30 06:27:03 RT @emilyywu: Saturday! Music starts at 2pm. Sunshine all day. Free entrance! Outdoor fest w/ pizza, friends, beers, family. We'll be doing…

2022-09-30 04:57:13 An inspiring cover story about @GoldsteinBrooke and her grassroots org @EndJewHatred which are transmuting the character of the cognitive war against Israel, especially on US campuses. @EinatWilf @LekhtNaya @blakeflayton @CotlerWunsh @michaeldickson https://t.co/LRziGvDZNE

2022-09-29 19:57:59 @AlexTensor What is the input = Data + Judgment ??

2022-09-29 12:12:44 Dear @GovMurphy, Show me ONE Governor, or ONE University President who does not vow to fight antisemitism on US campuses. Only a handful succeed. Why? Because they are using outdated vaccines. Advice: Insist on Zionist representation in every EDI office on campus. https://t.co/UVqFvE57JO

2022-09-29 11:52:03 As frightening as these racist Zones are, our duty as academics is not to remain silent but to respond loudly and immediately: Any scholar scheduled to speak at any of the nine discriminating student groups at Berkeley Law School should receive hundreds of message from 1/2

2022-09-29 11:52:02 Berkeley Develops Jewish-Free Zones https://t.co/J3Zt1bVyH3 via @jewishjournal

2022-09-29 09:33:13 I was told in Kindergarten it's not nice to laugh when someone else is angry. I tried to live up to it, but had to crack up (forgive me teacher) when Fatah declared another "Day of Rage" yesterday, to be added to the Calendar below: https://t.co/SikpH6F7rJ

2022-09-29 09:11:39 @ylecun @artistexyz @pabbeel And here is another paper (w/ Yoshua as co-author) specifically characterizing the expressiveness of NN's https://t.co/61qez08cgm including, I believe, proposals such as yours, based on Next-state-predictive models.But #Bookofwhy is a good and gentle primer..2/2

2022-09-29 09:05:16 @ylecun @artistexyz @pabbeel Mostly the bulging works in CI (causal inference) which provide theoretical characterization of reasoning tasks in terms of the questions they can answer and the kind of data they must be trained on. Here are a few examples: https://t.co/KBGKcmJxM8https://t.co/HJJdZzjJHf1/2

2022-09-29 08:42:39 @KhaledAbuToameh I don't recall a single day in the past 100 years that was not a "Palestinian Day of Rage", starting with the killing of the author Y.H. Brenner (Jaffa, May 1921). Hey, brother Abbas, how about a "Day of Reason and Co-Existence"? Just to try something different.

2022-09-29 08:35:09 @blakeflayton I was about to retweet it, but had to pause. "Progressives" would say: No way! No "progressive" can ever be an antisemite", axiomatically. They are merely angelic "anti-Zionist". Why not name the racists for what they are?

2022-09-29 07:59:36 Perhaps not coincidental, @HillelIntl has a job opening for CEO of Campus Climate Initiative:https://t.co/97BssXYVqSMy advice, get some faculty of @SAFECUNY on the search committee, else you might hire another philosopher of antisemitism, susceptible to the tactic of inaction. https://t.co/fXKjSLygrB

2022-09-29 07:43:01 Thanks for the memories. https://t.co/7XX3UpmDfF

2022-09-29 07:34:26 @SAFECUNY @HillelIntl The same happened at USC https://t.co/vXxPyUFnQW. 85 prominent professors complained about Zionophobic discrimination and USC's President anointed a "committee" to philosophize on "antisemitism", to the cheering sounds of Jewish organizations, granting her a license for inaction.

2022-09-29 07:21:04 The honor were ours, and the LA Press Club who were looking hard for a Middle East Reporter with courage and integrity to report facts and avoid propaganda. Stay the course. https://t.co/eczXjjTi5I

2022-09-29 06:53:27 Strange new partnership: @SAFECUNY with Hillel International on Campus Climate.https://t.co/C59cEKJAHSWhy is it strange? Notice the absence of the word "anti-Zionism" in the press-release, and absence of consultation with Jewish faculty. Why it matters: https://t.co/y34QJmPUyH

2022-09-29 05:40:21 @artistexyz @pabbeel @ylecun The problem with this slide is not absence citations (Pearl has enough of them) but its list of challenges, some of which were given theoretical underpinnings in other disciplines of AI research, largely inaccessible to DL folks.

2022-09-28 17:14:29 I am getting signals that someone is trying to slander my colleague @KhaledAbuToameh. Let me state categorically that, having been reading his stories over the past two decades, I find him to be the most reliable and insightful source for all aspects of Palestinian society. https://t.co/hUcdQ7Pb50

2022-09-28 17:04:17 @stephensenn @learnfromerror I am sure that if the alchemists explain what they want to do, (i.e., what the purpose is of analyzing "(fairly) simple experiments", what is given and what is assumed) Chemistry can help them get there faster.

2022-09-28 16:48:39 @stephensenn @learnfromerror "Something must be wrong" may go both ways. Chemistry did not tell Alchemists how to turn lead into gold, but that does not mean that something is wrong with chemistry.["The Sceptical Chymist" Robert Boyle, 1661]https://t.co/94mNwWONe8

2022-09-28 16:17:23 I fully endorse @CotlerWunsh call "to add “Zionist” to the list of protected characteristics in existing hate speech policies, affording Zionists the very same treatment as any and every other protected characteristic." See why https://t.co/y34QJmPUyH https://t.co/NN7cjx9b56

2022-09-28 15:48:38 @MatthewEGunter @TuomasPernu @IntuitMachine @ylecun I like this distinction: "intuitive causal models" vs. "analytical causal models", though sometimes the distinction lies not in the model itself, but in how it is used. For example, a DAG is "intuitive", but to find out if a set Z is "backdoor admissible" takes analytical work.

2022-09-28 06:08:31 @omaclaren @DerriereLaLune @learnfromerror @stephensenn Nice and clear. I think I begin to see where your generalization is heading. Need to digest.

2022-09-28 05:59:52 @eoteromuras Thanks for correcting the link.

2022-09-28 05:25:16 @DerriereLaLune @omaclaren @learnfromerror @stephensenn is in translating scientific knowledge about mechanisms (structural equations) to statements about counterfactuals and interventions. Strangely, only a few enlightened economists/statisticians have come to appreciate the power of the First Law. See: https://t.co/etASSEk4Or

2022-09-28 05:11:34 @DerriereLaLune @omaclaren @learnfromerror @stephensenn My comments: It is a good and broad exposition of some problems in mediation analysis, uncontaminated by the dismissive attitude of the Imbens-Rubin school. But I do not see much difference between "mechanism" and "structural equations". In fact, the beauty of the First Law 1/2

2022-09-28 04:57:08 @instrumenthull @omaclaren @learnfromerror @stephensenn He surely was. But his potential outcome notation was devised for one specific task (confidence intervals in RCT)

2022-09-28 04:49:06 @instrumenthull @omaclaren @learnfromerror @stephensenn without invoking "ignorability" conditions which, as we have shown again and again, are thousands of miles away from the way scientific knowledge is stored. Indeed, potential outcomes penetrated social science only in the late 1980's. 2/2

2022-09-28 04:38:08 @instrumenthull @omaclaren @learnfromerror @stephensenn You are right, I should have said "traditional 20th century statistician", with which our discussion started by D. Mayo. Rubin is semi-traditional in that he introduced non-statistical notation, Y_x. Still, he could not articulate when the equality P(y|do(x))=P(y|x) holds,1/2

2022-09-28 04:05:46 @instrumenthull @omaclaren @learnfromerror @stephensenn counterfactual statements, in Neyman's notation and, accordingly, use them to answer many policy questions, e.g., How many patients who died from Covid would have been alive today had we started the vaccine a month earlier.It's all in the First Law https://t.co/etASSEk4Or2/2

2022-09-28 03:52:53 @instrumenthull @omaclaren @learnfromerror @stephensenn Sure! Neyman not only "talked" but also gave it notation. And Rubin (1974) further added the consistency Rule to Neyman notation. This is all in the history books and in #Bookofwhy. But it wasn't until the First Law was formulated that we were able to interpret your k's as1/2

2022-09-28 03:37:28 @omaclaren @learnfromerror @stephensenn (OLS = Ordinary Least Square). The answer is: Easy! It's known as "single-door criterion" see Corollary 1, https://t.co/veOJoVDode.For many other frills look up Section 3.2 of https://t.co/Jf1SkgV3CK.3/3

2022-09-28 03:14:31 @omaclaren @learnfromerror @stephensenn to go formal? We are happy with our mystery models and once we estimate the k's we are done. What do we gain by being clear about the k's? The answer has many frills, but here is one of my favorites. Looking at the model, can you tell which of the k's can be estimated by OLS?2/3

2022-09-25 10:58:33 Wishing all readers: HAPPY JEWISH NEW YEAR.May the new year make your wishes come true. My three personal wishes: (1) May our neighbors deem us worthy of peaceful co-existence. (2) May all oppressors be toppled by their oppressed constituents. (3) May we have the strength 1/2 https://t.co/9csnGgxL3i

2022-09-25 10:30:36 @stephensenn @learnfromerror Perhaps we can see the problematic assumption if you state it in the context of the model, where we have only two halls, each offers one and only one diet? I don't recall using Ancova

2022-09-25 09:43:45 @stephensenn @learnfromerror I don't see any assumptions made about regressions, only about causal relationships, specified formally in the graph. Perhaps you can state the problematic assumption in the context of that graph, where we have only two halls, each offers one and only one diet?

2022-09-25 09:06:07 @stephensenn @learnfromerror The #Bookofwhy solution to Lord's paradox is beautiful and general: https://t.co/38QW0n1JkC.It does not depend on any unstated assumptions, if we stick to the original statement of the paradox, free of added complications, as shown here: https://t.co/MSJJVipnXE

2022-09-25 08:46:50 @Alfrednguyen432 just posted a 2012 video of a seminar by Jim Heckman and myself https://t.co/Sx38f3sFyT which is highly relevant to our discussion of why statisticians can't let go of the statistical paradigm, and why economists may be able to do so. https://t.co/vN3ymMaIra

2022-09-25 04:11:00 @learnfromerror @stephensenn I see nothing contradictory here: "If you think reality for Cox was MORE than prob. dist., show me the mathematical object he used for the MORE." Rephrased: If you think Cox deemed prob. dist. insufficient to capture some aspects of reality, what did he use to capture them?

2022-09-25 02:49:47 @learnfromerror @stephensenn Quoting what I "alleged about what stat presupposes": "in traditional statistics, "reality" is equated with the probability distributions that govern the data we obtain." If you think reality for Cox was MORE than prob. dist., show me the mathematical object he used for the MORE.

2022-09-25 02:38:14 @cubic_logic @stephensenn As you can see from the first sentence, Box is dealing with "statistical model". What (some) statisticians find hard to internalize is that, even when the statistical model is declared "perfectly adequate", the model is still incapable of answering basic questions about reality.

2022-09-25 02:23:46 @learnfromerror @stephensenn Show me a mathematical object used by Neyman, Pearson or Cox, capable of answering the question: What is the likelihood my uncle would be alive today, had he not smoked. Pearson, in particular, said: "It's all in the contingency table." I quote him here https://t.co/hWDVC9lIED

2022-09-25 01:52:08 @learnfromerror @stephensenn I am claiming that, in traditional statistics, "reality" is equated with the probability distributions that govern the data we obtain. "Reality" is more than that. It includes "roads not taken", i.e., cause effect relationships that permit us to generalize and explain things.

2022-09-25 01:34:28 @ChristosArgyrop @stephensenn reformulation is "revolutionary" can be measured by the degree to which it is resisted or mocked by prominent statisticians to this very day. Why? See https://t.co/XFZYzAbLHy2/2

2022-09-25 01:26:32 @ChristosArgyrop @stephensenn He sure would, but the hypotheses he had in mind were all statistical. We are now in a new era, where our hypotheses are about causes and effects (including causes of effect) so all the beautiful ideas of Jaynes, Box, and others need reformulation. The degree to which this1/2

2022-09-25 01:05:47 @HL327 @stephensenn @RogueRad I think even Huxley would agree with us that you cannot slay a theory or hypothesis unless it is represented as a mathematical object. This is precisely what ML (and statistics) lack, a mathematical object representing reality, not merely data that reality graciously grants us.

2022-09-25 00:51:48 @stephensenn This distinction is an integral part of the scientific method which dedicates separate symbolic systems for (1) model of reality, (2) drawing conclusions from it and (3) testing it against empirical evidence. None exists in ML (or statistics) where data is mistaken for "reality."

2022-09-25 00:36:33 Thanking all readers who wished me well on my eye surgery. It went well, but it's still hard for me to return to normal activity, due to pain, patch and general discomfort. Will join the revolution slowly, so we can defend the faith together, and promote heresy when productive. https://t.co/tj1fyHbDgG

2022-09-24 17:18:31 The Israeli-Palestinian conflict in 8 words:Lapid to UN: Two states, equally legitimate.Abbas to UN: Two states, unequally illegitimate.https://t.co/DqJIoa9cJA https://t.co/qgL0tGk4Eu

2022-09-24 17:05:50 The "unassailable" mathematics tells us what it takes to climb from one rung of the Ladder of Knowledge to the next. Our brain has a model of reality, however flimsy -- most ML systems do not

2022-09-24 16:51:07 @raedhamed1 @IlhanMN @lsarsour The struggling women in Iran are begging women who wear the Hijab by choice to show solidarity with those who are burning it as a symbol of no choice.

2022-09-23 09:05:47 To all readers on this channel, I am undergoing a minor eye surgery tomorrow, so may be away for a day or two (hopefully not more). Stay the course and defend the faith. I gave the surgeon a copy of #Bookofwhy, of course, to keep his hands steady and his mind purposeful. Best, JP

2022-09-23 07:42:04 waste, dangers and thoughtlessness of the imbalance created by the trendy ML dominion over education, funding and media attention. It makes the revolution so much harder to take off and reveal the potentials of AI systems that reason and understand.2/2

2022-09-23 07:42:03 Luckily, I can escape the wrath against "AI critics". (1) I'm an "AI believer", I only critic AI'sslavery to popular trends in ML. (2) My heretical claims are not mine, they come from unassailable mathematics. I admire heretics like @GaryMarcus and others who decry the 1/2 https://t.co/FyC7l1s2EA

2022-09-23 07:11:31 @rravi Trying, but failing to parse the solution you propose, or any solution that is not based on uprooting deeply entrenched rejectionism to any form of co-existence that now dominates, to the last person, one side of the conflict.

2022-09-23 06:12:04 hostility against Israel and join the armies of Zionophobic thuggery, often to the tacit approval of Jewish organizations, and always to the willfully blinded eyes of University Administrators. Our only hope: pushback Zionist organizations like @newzionists and @SSI3/3

2022-09-23 06:12:03 Back to School With Antisemites https://t.co/drf4SAK80N via @jewishjournal

2022-09-23 03:57:59 discourse. The reason lies in the elephant:"We have only one condition: That a future Palestinian state will be a peaceful one. That it will not become another terror base from which to threaten the well-being, and the very existence of Israel." Here he states politely what2/3

2022-09-23 03:57:58 The ELEPHANT behind PM Lapid speech at the UN. Commentators express surprise that he explicitly backed a two-state solution. Rightly so, A decade has passed in which the 2-state dream has been simmering in the minds and hearts of most Israelis, yet absent from any diplomatic1/3 https://t.co/JTIEvpDkdy

2022-09-23 03:14:27 Saluting Christiane Amanpour for her refusal to comply with Iranian President Raisi's demand to wear a headscarf which, to the women of Iran, is a symbol of women subjugation. The empty chair which Raisi refused to honor speaks more about him and his regime than 1,000 words. https://t.co/qD33D1Cn7O

2022-09-23 03:06:42 We can do more for the Iranian people, certainly in the public arena. For example, Muslim feminist leaders @IlhanMN and @lsarsour could take off their hijabs publicly, in a moment of silence, as a gesture of solidarity with the Iranian women who are burning their headscarves. https://t.co/oAoZFhn0rF

2022-09-22 11:50:36 @slasherbarbie @blakeflayton Beware! The last guy who called me "colonizer" was asked "Cite one Philistine poem or one Canaanite hero". He never showed up again.

2022-09-22 11:24:10 @David_Leavitt @MassGov Glad you discovered their ploy before it was too late. They would have sacrificed you to their false Gods and used your blood to prepare Hanukah Latkes. Praise to the Lord you are alive.

2022-09-22 02:44:46 First they claimed sole ownership of Tel Aviv, then sole ownership of our history, now they claim sole ownership of progressive values. Soon@RashidaTlaib will demand sole ownership of mathematics and human biology. Can someone teach her the word "share"? https://t.co/iP8hq52JaW

2022-09-22 01:59:30 @slasherbarbie @blakeflayton Those who can't tell "homecoming" from "home- stealing" are likely suspects of the latter, and may end up homeless.

2022-09-21 19:58:19 @goldenrule108 @IlhanMN @lsarsour @RashidaTlaib While support for the "mandatory hijab" is zero in the United States, solidarity with those in Iran would be empowering, if offered by those whom the world perceive to be the voice of American Muslims. No axe on Palestine, just on Israel = my litmus test of logical consistency.

2022-09-21 17:02:32 @stu_frost @JohnHMorris @jhurwitz square a circle we need first to have a language in which to define a "circle". We can't define "causal effect" in the language of statistics.

2022-09-21 16:58:59 @stu_frost @JohnHMorris @jhurwitz The gap is not between me and ML folks, it is between Rung-1 and Rungs-2&

2022-09-21 10:25:32 @sdandach @ilanomar @RashidaTlaib To give their followers/voters an illusion of principled consistency

2022-09-21 10:19:43 @goldenrule108 @IlhanMN @lsarsour @RashidaTlaib The hijab is not a symbol of women subjugation to those who voluntarily choose to wear it, it is such a symbol to those who are denied that choice.

2022-09-21 05:52:01 A readers noticed a glitch in my tweet: The tag of Rep. Ilhan Omar is @IlhanMN, and that of Linda Sarsour is @lsarsour

2022-09-21 04:51:58 As Iranian women burn headscarves in anti-hijab protests, see https://t.co/xroht0lRpX, I encourage American feminist leaders, Linda Sarsour, @ilanomar and @RashidaTlaib to show solidarity and publicly take off their hijab, for ONE solemn moment of silence.

2022-09-19 10:53:21 Glad to see the first step accomplished: Demand that Zionism be recognized as "protected characteristic", replacing the fuzzy term "Jews"(even JVP claim they are jews). Next step: Declare Zionophobia a form of racism, of equal status to Islamophobia and White Supremacy. https://t.co/oEX8V8QvMg

2022-09-19 10:34:27 Good news! UCLA is hiring! See invitation from our chairman https://t.co/T6hYEkM6qN If you are a super researcher in any area of AI, or you know of one, please consider this opportunity. And if your research is in Causal Inference, I can't wait to hear your talk.

2022-09-19 10:16:46 Hilarious!Who said they don't teach selection bias in stat? https://t.co/oGot2f1O3K

2022-09-19 09:05:31 @blakeflayton This is partly because Zionist activists are still not using the word #Zionophobia, to let the "peacemakers" know how profoundly immoral BDSers are, much more repulsive than "academic or economic boycott".

2022-09-19 06:28:12 @skeenscene @Claire_V0ltaire @ReconRabbis Brant Rosen exist? I thought he must be a Hamas operative making fun of Jewish holidays. In my home town he would be excommunicated but, given that there isn't anyone here to do it, let's ask the Chief Rabbi of Gaza. She must have the credentials.

2022-09-19 00:17:33 @blakeflayton It's not nice to talk like that against my hero, Lina Sarsour, who is a known Jew-lover (donated money to repair vandalized Jewish cemetery) and all the Jewish leaders competed for a seat on her band wagon: https://t.co/2nCKQwTWFX. Please use her respectable title: a Zionophobe.

2022-09-18 23:46:46 @Claire_V0ltaire As everyone knows, Yom Kippur is an ancient Palestinian Holiday, started by Ishmael (a Palestinian prophet) and Jesus (another Palestinian prophet) then stolen by Zionistcolonizers, together with the holiday of Channuka, a National Palestinian Holiday commemorating their victory

2022-09-18 23:38:40 I am going to watch CNN soon (5 pm Pacific), it's a painful topic I cannot easily dismiss. It bears similarities to the British blockade against Jewish refugees, 1939-1948. https://t.co/tpjWIsOGnj

2022-09-18 13:22:35 This is very kind of you. I'm retweeting with the hope that some Trialists will take the time to read a few chapters of #Bookofwhy. I was shocked the other day to find out that some DID NOT, then learned that MOST did not. https://t.co/fFpiSRko2t

2022-09-18 13:14:07 @LuisaBorrell Why? Because the media, academia, and even some Congresswomen have bought into Palestinian logic: Israel is guilty of horrible crimes, Jews are likely to support Israel, hence every hoodlum is morally justified in acting to silence this criminal support! It's elementary logic.

2022-09-18 13:02:59 Thanks @TriciaRoth18 for reposting this famous video, with Zahra Billoo (Dir. of CAIR-SF) warning us (38:00 min.) against "polite Zionist" organizations, ADL, Hillel, Zionist synagogues, who will befriend you today and betray you tomorrow (e.g., by advocating a 2-state solution) https://t.co/vXoevpNXcn

2022-09-18 11:48:42 @AviBittMD If so, what is your criterion for "correctness".I, too, have no dog in the fight

2022-09-18 11:41:32 @AviBittMD Of course people have spoken jiberish about confounding since it was noticed (By Pearson (1890's) and Yule. Of course people proposed, for more than a century, all kind of wrong criteria for confounding, and continue to do so today. Are u trying to correct their thinking? If 1/2

2022-09-18 10:07:22 @AviBittMD You seem to imply that "how confounding is being defined in book of why" is somehow unique to #Bookofwhy, and that there are alternative ways of defining it, perhaps more meaningful". I would be eager to hear! Can we start with a formal definition of "confounding" or "c-effect"?

2022-09-18 09:50:55 @MikeNew19512115 @ChancellorCUNY @EricDinowitzNYC @InnaVernikov @NYCMayor @GovKathyHochul @KathyHochul @nypost @AdamMilstein @lauder_ronald @JGreenblattADL @ZoaKlein @MortonAKlein7 @KalmanYeger @KalmanYegerNYC Thanks @MikeNew19512115 for recognizing the one, and only one way of unveiling the racist face of "diversity" officers in our universities. To my knowledge, no Jewish leader has dared doing it, which is beyond me: Why? Don't they understand that a Zionophobe can't deny its credo?

2022-09-18 09:23:25 @nagaraj_arvind What happened? The media, academia, and even some Congresswomen have bought into Palestinian logic: Israel is guilty of horrible crimes, Jews are likely to support Israel, hence every hoodlum is morally justified in acting to silence this criminal support! It's elementary logic.

2022-09-18 09:02:45 @AviBittMD #Bookofwhy , i.e., P(outcome | do(treatment)) then, only then, we can ask whether randomization answers the question asked.#Bookofwhy pages 143-150 show that it does, and I do not know of a better proof (or a better conceptual argument). Disagree? So, let's start with the proof.

2022-09-18 08:52:31 @AviBittMD I think any explanation of "what the goal of randomization in RCT is" must start with "what the research question is that RCT attempts to answer". Arguments about "balancing covariates" or "equalizing distributions" must begin with definition of "causal effect", as given in1/2

2022-09-18 08:18:35 @DavidDeutschOxf I think clinging to Kings and Queens as symbols of commonalities that unite members of a culture explains the longevity of those cultures. Abstract ideas such as "common values" quickly lose their punch, unless they are attached to names of individuals, and turn "common stories".

2022-09-18 08:00:00 Sad to learn the passing of Saul Kripke, a brilliant philosopher. I once tried to interest him in the necessity of causation, but he insisted on "Naming and Necessity". A loss to philosophy and to history of ideas. https://t.co/cMdWEzCWQ4

2022-09-18 07:04:19 This great piece by Bill Maher need be enjoyed together with this article on him: https://t.co/iJ9J3OBnPvI'm surprised the Woke Forces allow him to stay on the air. Stay the course @billmaher https://t.co/XkP15CDw8s

2022-09-18 06:18:36 RT @TheLeadCNN: A new documentary examines America's response and inaction during the Holocaust. Co-directors @LynnNovick and @KenBurns tal…

2022-09-18 06:12:49 Interesting thread. My two cents: None of the Journals mentioned has invited a survey article on Causal Inference or Graphical Models, thus constraining economics to another decade of [...][I would let red-blooded economists fill in the appropriate adjective] https://t.co/wrZQjq4TUY

2022-09-18 06:02:58 RT @_zhuyuchen: If you are in Tuebingen on the 20th September, come and hear me talk about causal inference for social sciences! Abstract…

2022-09-18 05:38:21 Great read! @ShMMor! I've also written a piece about Beinart, "Jews of Discomfort", on his first visit to UCLA: https://t.co/lz8kD8UMvU."Mother, I hate you, for making me feel so inadequate, unable to defend your honor except by joining the bullies in amplifying your blemishes.” https://t.co/caG4BYjiKi

2022-09-18 05:22:51 After refusing accreditation of Israeli envoy, Chile reschedules ceremony https://t.co/047pU2TYKt

2022-09-18 04:48:10 Watching this 12-16 hours queue on BBC today made me ponder where nationhood and peoplehood reside in the human brain. Plus, it made me ask: How did we, Jews, survive without a monarch for so many centuries? Are the Biblical stories of King David a sufficient substitute? https://t.co/fMzxAxaH7p

2022-09-18 04:40:33 RT @Carthica: Is it safe to assume IID when you know that data are not IID? What conditions contribute to bias when estimating causal effec…

2022-09-18 04:38:56 @DavidHirsh @centre_as I wish I could take part in this important conference. I am hearing only good things about it. Stay the course.

2022-09-18 04:36:20 RT @AJCGlobal: History was made 44 years ago today, when Menachem Begin and Anwar Sadat signed the Camp David Accords. Since then, Israel…

2022-09-18 04:35:02 RT @HillelNeuer: History will record that on this day Twitter suspended the account of Masih Alinejad—Iranian-born journalist who is the mo…

2022-09-18 04:28:38 I would not blame any President for child exploitation in his country, because I havn't studied the issue in sufficient details to attribute blame to one factor or another. But I can confidently charge a President with Zionophobic racism, a topic I have studied for a few decades. https://t.co/1g43MdNh7Z

2022-09-18 04:04:15 Kudos to Professor Lax @CUNY_Prof for so clearly stating the charge: "discrimination against Zionist students and professors." To exonerate herself from allegations of partiality, all Saly Abd Alla has to do is to publicly state her views on Zionism. Challenge her to the task!!! https://t.co/DknNxVaUCW

2022-09-18 00:58:57 @Drg_elness @gabrielboric I the era of Putin and Ebrahim Raisi, we should not fall for the words "democratically elected." We should look at policies, norms and attitudes. President @gabrielboric failed the latter and "democratically elected" arguments do not make him less of a racist.

2022-09-18 00:16:26 @robwilliamsiii Sorry, Twitter tells me "This Tweet is unavailable"Evidently someone blocked me.

2022-09-18 00:12:45 @DavidDPreiss @gabrielboric Yet your countrymen claim it's shameful to compare @gabrielboric to Maduro. Can't they see where he is heading with his animosity towards Israel, and his determination to please the Zionophobic voting-base in Chile?

2022-09-18 00:07:12 @Drg_elness @gabrielboric Tyranny begins with dogmatism, and dogmatism can be detected by how one treats the plight of both Israelis and Palestinians for a day of normalcy, not the latter alone.

2022-09-17 23:41:03 It is time for the FDA to assemble a task force to examine these (theoretically sound) proposals, and remove the language barriers between "trialists" and "causal inference" researchers.Step-1, read #Bookofwhy. Step-2, read https://t.co/B0TKO64IzR, the rest will follow.2/2

2022-09-17 23:41:02 Last week I had a Zoom session with a higher-up in the FDA and the same question came up. I cast my answer in a positive mood: Solutions to chronic weaknesses of RCT have been proposed: Selection bias, integrating diverse studies, generalization to target population, etc.1/2 https://t.co/hV88hEOxPh

2022-09-13 03:43:41 Remember the conversation I had with @MomentMagazine, about Zionophobia, antisemitism, AI and more? The video is available now, and I am happy to share: https://t.co/qHtORmprjmWe didn't get to the AI part - coming shortly. https://t.co/ZrnTwEOEHw

2022-09-13 03:14:34 Time to examine this week's production of causal inference articles: https://t.co/qt7ZDDPmMe. Some of the titles are extremely enticing, but I do not have the time to sort them all. Please alert us if you find new breakthroughs or new perspectives.

2022-09-12 19:06:19 @stuz5000 @csabaveres Is this a confession or an accusation? Please unpack.

2022-09-12 19:03:05 @hohl_jeremy @truth4_today @Ostrov_A @benshapiro @UN_HRC @UNHumanRights @NikkiHaley @ShaharAzani @RepRitchie @MeghanMcCain @Kasparov63 @EnesFreedom @The_ILF We are not talking "history", we are talking daily practices towards their own citizens, this day and age.

2022-09-12 13:18:17 No, NOT a joke! I often think about the #HRC51 delegates from enlightened countries, say UK or France. What do they tell their children when they come home, and are asked: "Dad, did you sit with these guys in the same room and play the game without puking?" I feel for them. https://t.co/eSdsjkf5fR

2022-09-12 12:59:52 RT @yudapearl: @PavlosMsaouel @WvanAmsterdam @arthur_alb1 @soboleffspaces @GelovenNan @PWGTennant @Mo_Safavi @Tufts_PACE @ESteyerberg This…

2022-09-12 12:56:02 @PavlosMsaouel @WvanAmsterdam @arthur_alb1 @soboleffspaces @GelovenNan @PWGTennant @Mo_Safavi @Tufts_PACE @ESteyerberg This is a paper even I could understand, though it is billed as "predictive", in the PATH category. I hope every oncologist reads it, including some from the PATH enterprise

2022-09-12 10:43:05 @haaretzcom Sorry, but you cannot brush off this idiotic subtitle so lightly, without informing your readers what steps were taken to prevent its author from repeating. @Ostrov_A

2022-09-12 09:55:06 @WvanAmsterdam @arthur_alb1 @soboleffspaces @GelovenNan @PWGTennant @Mo_Safavi @Tufts_PACE @ESteyerberg Suggestion: Some readers (like me) do not know what "Offset" is, and it would be helpful to defineCR-CATE earlier than bottom of page 3. It is, after all, the essence of the paper.

2022-09-12 09:33:10 @WvanAmsterdam @arthur_alb1 @soboleffspaces @GelovenNan @PWGTennant @Mo_Safavi @Tufts_PACE @ESteyerberg I may have been misled by the term "clinically meaningful HTE" which, for poor me, connotes "causally meaningful." Re-thinking, I still believe doing correlational modeling outside causal modeling is not wise, because we lose tools for fusion, selection-bias and missingness.

2022-09-12 09:11:37 Strange, but the only thing that makes these speeches different from the ones I heard in my childhood (say 1940-1949) is the word "occupation". Change "occupation" to "Yahud (Jews)" and you reconstruct the 1947 texts of Azzam Pasha and Fawzi al-Qawuqji . What's new? https://t.co/Z8JursQBJq

2022-09-12 08:42:43 RT @IDF: The whole world was put on hold when the twin towers were attacked on 9/11.Today, we remember the 2,977 innocent victims that we…

2022-09-12 08:34:22 RT @michaeldickson: The 1st victim of #September11 was the exceptional Danny Lewin, an elite IDF unit veteran.Seated aboard American Airl…

2022-09-12 08:26:03 @WvanAmsterdam @arthur_alb1 @soboleffspaces @GelovenNan @PWGTennant @Mo_Safavi @Tufts_PACE @ESteyerberg I've been searching for such issues, that are particular to health care, and could not find any which would not show up in a general causal model. Even with purely RCT data, causal interactions among covariates need be modeled if one seeks criteria for clinically meaningful HTA.

2022-09-12 07:52:54 Whenever I hear an anti-Israel diplomat regurgitating: "The two-state solution has reached a dead end" I am reminded of that legendary fisherman talking to the fish in his net: "I guess, buddies, our swimming exercise has reached a dead end". @Ostrov_A @EinatWilf @LawfareProject https://t.co/qYyxhWri1p

2022-09-12 07:28:24 Two-state solution has reached a dead end, former Jordanian ambassador to Israel says https://t.co/L7xVvrNws1

2022-09-12 06:58:55 @Claire_V0ltaire Yester-Jews may give away anything THEY own to whomever they wish, but not the home that my father owned, and his father, and his...

2022-09-12 02:46:10 @arthur_alb1 @soboleffspaces @GelovenNan @WvanAmsterdam @PWGTennant @Mo_Safavi @Tufts_PACE On a 2nd reading of the Tuft's paper, its main claim seems to be: "Accounting for multiple variables simultaneously, .. is foundational to the concept of personalization in EBM". This task is standard since 1995 so, perhaps one of the 15 authors can explain: What have we missed?

2022-09-12 00:07:31 @arthur_alb1 @soboleffspaces @GelovenNan @WvanAmsterdam @PWGTennant @Mo_Safavi @Tufts_PACE I find here 15 authors &

2022-09-11 11:07:21 And the educators among us would value this introduction: https://t.co/eKGIHTfZyP. where @smueller and A. Forney share their experience teaching Causal AI for undergraduate classes. https://t.co/m7fCchp3OG

2022-09-11 09:40:43 Many new readers ask this question and, as I've done in the past, I would recommend Primer https://t.co/XofByb7KEO which is available free (chapter by chapter) and logically coherent, starting with the First Law and ending with Tool Kits for Mediation and Attribution. https://t.co/zD5vs0kd4F

2022-09-11 09:21:41 Hilarious! They actually sing it in the UK:"O Lord our God arise,Scatter her enemiesAnd make them fall

2022-09-11 09:07:34 "He is a Sabra"! My aunts and uncles said when they had to apologize for my spiteful behavior. Superheroine Shira Haas, so I hope, would need no apologies. https://t.co/PzYWbxHNZY

2022-09-11 08:47:19 RT @AviKaner: “May the God of peace bless this region and all the world with peace.” - William, Prince of Wales https://t.co/ghz6N9kJCL

2022-09-11 08:45:58 It may not make Eleanor Roosevelt happy, but it would surely exhilarate @hrw and @amnesty. And @KenRoth will sing Halleluiah, in high falsetto!!! https://t.co/YbywgxVnXm

2022-09-11 07:19:48 The clock has just struck midnight, and has brought with it the gravity of September 11, a day that refuses to match our understanding of homo-sapiens. I cannot find a better expression to 9/11 gravity than the monument below, built in Israel, honoring every victim by name. https://t.co/ggmbu863Wt

2022-09-11 04:34:33 A wiseman once said: "A nation without history is a nation without a future." Shortly thereafter, scholars at Al-Quds University discovered that Adam and Eve were two Palestinian philosophers who were ethnically cleansed by white colonizers while fighting for a Right of Return. https://t.co/UIBnQFAFdN

2022-09-10 21:35:09 @AdamMilstein It's not Marouf fault, who was just doing what he learned to do at home and in school. It's Ahmed Hussen's responsibility, who hired Marouf, and who is orchestrating the agenda of this whole department.

2022-09-10 12:07:40 it is erroneously supposed to do no harm.” My thought: Had Russell been with us today witnessing the rebirth of the science of causality, he might perhaps have changed his opinion about the usefulness of the monarchy. 2/2

2022-09-10 12:07:39 As we are watching the coronation of King Charles III we should remember the poetic connection between the monarchy and causality, as orated by Bertrand Russell (1911): "The law of causality, I believe, is a relic of bygone age,surviving, like the monarchy, only because 1/2

2022-09-10 06:05:03 That's not the end of the story. The sad part is that as soon as we hear a University choking on: "These labels are antisemitic" we find dozens of Jewish organizations praising them for showing such consistent and forceful commitment to fighting antisemitism.@StopAntisemites https://t.co/84toueoHWT

2022-09-10 00:30:20 This double standard is written into the code of every university that I know, and into the charter of every Equity Diversity and Inclusion office that I have had a chance to work with. https://t.co/FXNaWoPFqq

2022-09-10 00:17:55 RT @EinatWilf: For a while I was intrigued by our fascination with The Queen, especially in TV shows like The Crown. I now realize that in…

2022-09-10 00:12:59 @blakeflayton Where do you collect these draggers?

2022-09-09 23:33:14 @Shayadler2 Sure, but one of Britain's narrow interests was to appease the Mufti who was leading the 1936-39 riots in Israel, and organizing the anti-British uprising in Baghdad 1941. "Enlightened intellectuals" can't accept the idea that Zionism if a home-coming endeavor - not in textbooks.

2022-09-09 21:16:06 @WvanAmsterdam @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi I was under the impression that "making sure the prediction models actually estimate the interventional distributions (p(y|do(x),w)) " is a Rung-1 exercise, so, we can call the resulting distribution "estimand". Why'predictimands'?

2022-09-09 13:25:02 @stephensenn @jwbelmon @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi My hope is that someone who understands both SCM and "practical real-life" RCT jargon will be able to articulate the latter in the language of SCM, and things would fall into place. This was done in (real life) epidemiology, social science, economics and more. Why not RCTs?

2022-09-09 13:14:44 @jwbelmon @stephensenn @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi I could not see any methodological difference between prognosis and diagnosis. Willing to accept practical differences, but why do we have more patient-specific data available in diagnosis? Is it because the patient is already in a hospital environment?

2022-09-09 13:09:09 @stephensenn @jwbelmon @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi I can't outguess what you mean by "new formulation" I need variables, events, data available, quantities needed to be estimated, etc.SCM offers us a flexible language to express any research question I can think of. Why not use it?

2022-09-09 13:01:22 @Shayadler2 Balfour did a lot for the establishment of Israel,(see https://t.co/9XOCOs9hGT) but then Herbert Samuel (trying to appear neutral) appointed Amin El-Hosseini as Mufti of Jerusalem which turned Britain against the Zionist project and started a 100 year conflict.

2022-09-09 12:47:51 @stephensenn @jwbelmon @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi Because I still can't cast the research question in a language that would permit scientific analysis: variables, events, what is given, what is needed, etc.

2022-09-09 12:38:08 @stephensenn @jwbelmon @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi "How does this work?" is not a research question. You probably mean: "Can we eliminate the difference between experimental findings and what we wish estimated." This is the "selection bias" problem solved. If u have a different problem in mind, articulate and it will be analyzed.

2022-09-09 12:18:29 Ettlee, Bevin, and King George VI were my swornenemies (for blocking Holocaust survivors from reaching Israel). Yet, I feel only reverence, warmth and admiration towards Queen Elizabeth II, and sadness on her departure. She worked for a cause higher than herself. Gun Salute!

2022-09-09 12:18:28 I was born in the British mandate. My father traveled on a British Commonwealth passport. We screamed our heart out: Down with the British Government! Down with White Paper! (in the bus to summer camp). British paratroopers searched our apartment (for illegal weapons).1/2 https://t.co/jxxsJmN2yx

2022-09-09 11:52:36 RT @Ostrov_A: Mossad to CIA: #Israel will not stand by as #Iran defrauds the world.https://t.co/J3UvRz0ewZ

2022-09-09 09:57:20 @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi Given that you are familiar with both SCM and "predictive models", can the problems tackled in the paper above be cast as problems in SCM? Why not? For example, can the notion of "predictimand" be translated into SCM? I did not find it well defined (formally or conceptually).

2022-09-09 09:34:42 @stephensenn @jwbelmon @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi pertains to a task called "selection bias"

2022-09-09 09:20:00 @stephensenn @jwbelmon @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi I assume you want to find out whether tools developed in causal inference are applicable to tasks in your areas of interest? The general answer is: We define each task scientifically and appeal to SCM to find out if it can be accomplished formally. For example, question 51/2

2022-09-09 08:58:21 Excited to see this paper tackling the century-old problem of finding "the root causes of outliers" using Structural Causal Models. My only question: Why not use "probabilities of causation" to rate the many "root cause" candidates? e.g., https://t.co/quhPwUiQII https://t.co/S5e0zEO6AA

2022-09-09 08:42:32 RT @Isaac_Herzog: HM Queen Elizabeth II was known far and wide simply as The Queen. Her passing is the end of an era. Together with the Isr…

2022-09-09 07:19:24 @elderofziyon As they once said in our places: History with no truth to people with no history.

2022-09-09 06:49:53 @MaartenvSmeden @soboleffspaces @PWGTennant @Mo_Safavi This paper re-enforces my interpretation here: https://t.co/dbmwL3bWiX. Am I wrong? I can see a few points of intersection with causal inference: 1. explanability, 2. meta-analysis and 3. missing data. Do you see more?

2022-09-09 06:33:14 @jwbelmon @MaartenvSmeden @soboleffspaces @statsepi @PWGTennant @Mo_Safavi Thank you for illuminating us with this book. Would I be wrong to characterize this field as "classical statistical classification (Rung 1) applied to medical problems"? What do you think would the leaders of this field like to ask researchers in causal inference?

2022-09-09 06:23:46 RT @AbigailShrier: “If the problem we’re facing is despair, the cure may be hope, that precious metal that is best mined wherever a sense o…

2022-09-09 06:09:02 @SaraReyi This is what Israelis are proud of. I was born in one of those cities, Bnei Brak, which was named "Ibn Ibrak" till 1924, and Bnei Brak till 140 CE. Believe me, you wouldn't want to grow up in Ibn Ibrak, nor did the few shepherds who welcomed the water pump my family installed.

2022-09-09 05:53:28 @blakeflayton Why is it only "gaslighting" and not a "license for genocidal plans"? as I argue here: https://t.co/K37zGYzt8w

2022-09-07 11:27:56 Another week loaded with papers on Causal Inferences: https://t.co/ptIeujekrh. I've examined two "survey papers" and found them wanting. Authors of such papers should take a (reasoned) position and express preferences, rather than quote other authors with opposing views.

2022-09-07 07:31:34 My 86th birthday has lured some readers to ask which paper, lecture, or interview I am most proud of. It took me two days to decide, and I now dare to share my choice: https://t.co/b4b7LEVdKl. It is a 2007 commencement speech I gave at the University of Toronto, 1/2 https://t.co/XQfiCfHSrV

2022-09-06 17:40:29 @PhilDawid There is no gap, because the goal was not to reach consensus but to minimize the number of things people need to agree on for full consensus to prevail, and to make those things MEANINGFUL. I'm not the one who labeled them "meaningful", it was human language, from Abraham to Hume

2022-09-06 09:12:56 @PhilDawid Of course it must be "parial". We know from the ancients that Rung-3 claims cannot be tested by data alone. But having testable implications is merely a bonus on top of the real miracle: having a parsimonious and meaningful object to agree on before we agree on everything else.

2022-09-06 06:35:25 RT @ShaharAzani: Yossef Romano was an #Israeli weightlifter butchered 50 years ago by #Palestinian terrorists at the 1972 Summer Olympics i…

2022-09-06 06:31:56 @DKedmey This question belongs to the "Probability of Causation" variety https://t.co/QQa4sduT11 also known as "Causes of Effects" , formalized through PS PN PNS and a few more.

2022-09-06 04:29:31 This bookshelf is even more impressive than the one I have at home and, confessedly, I too find some of the old papers not easy but extremely enjoyable. What a strange combination! Thanks for this unique birthday gift. #Bookofwhy https://t.co/tELcv4QHqK

2022-09-06 01:58:18 @soboleffspaces I would say "causal effect must be defined in other ways...". They can continue and argue that those "other ways" are built on assumptions they do not need in RCT. Fine. We are dealing with definitions, not estimation. This distinction is basic to the scientific approach.

2022-09-06 01:53:32 Good question, the answer to which revolves around personalities and ego trips, so it would take 100 tweets. Imagine what a trauma it must be for Rubin to see all his Potential Outcomes, their properties and intricacies emanate from something as simple as a causal model, 1/2 https://t.co/8ZcyoGrJyb

2022-09-06 01:28:54 @soboleffspaces Agree. The backdoor criterion and the adjustment formula do not need the details of SCM. They emanate from the DAG and the DATA alone.

2022-09-06 01:21:19 It's a great historical page by @Martin_Kramer , to which I can only add that for us, who grew up in pre-State Israel, the name was as natural as mother milk

2022-09-05 22:24:26 Trite? Saying that "two people who accept Euclid axioms of geometry would also agree on all theorems about triangles" is not Trite but a Miracle. Same as saying "two people who accept Kolmogorov's axioms of probability would also agree on Dawid's axioms of conditional1/2 https://t.co/p1MMydFeSm

2022-09-05 22:09:50 the power of logic to produce answers to such a gigantic space of counterfactual questions from such a parsimonious, transparent and (partially) testable representation. Look: Y_x = Y_{M_x}the rest is just math.Halleluiah!2/2

2022-09-05 22:09:49 The statement may appear trite, once we accept the First Law of Causal Inference https://t.co/etASSEk4Or.But given the number of prominent researchers who resist the First Law, or fail to appreciate its power, the statement is the opposite of trite

2022-09-05 12:33:14 50 years to the Munich Olypics massacre compel us today to remember 7 under-reported facts about the event that has changed how we view terrorism. https://t.co/tiPLORtWZO

2022-09-05 08:23:18 @AndrewDGarrett "hence the trial" is what we are trying to explicate.Can we examine more closely what the trial tells us about "safety features"? Can it be articulated in scientific language? Is it in any way related to Prob(harm) defined in https://t.co/rGHIc9kH0O? or is it another animal?

2022-09-05 00:53:26 RT @KhaledAbuToameh: Hamas on the recent violence: "Our people in the West Bank have launched a new phase in the confrontation (with Israel…

2022-09-05 00:46:12 @SAFECUNY CUNY is not entirely at fault. "Jew hatred" and "antisemitism" are (purposely ) slippery concepts.Zionophobia is clear, well defined and undeflectable. CUNY should be called out for what they can't deny.

2022-09-05 00:31:08 @DingWynn Glad you undertook this broad survey of causal inference. My only reservation is its breath. By trying to give a voice to anyone who expressed an opinion about causation the impression is created that disagreements rule the day when, in fact, the field enjoys widespread consensus

2022-09-05 00:17:18 @AndrewDGarrett What is so hard about causation? The definition or the estimation? I think the latter should follow the former, which is simply the research question motivating the study. Do you know any research question, be it individual or population level, that cannot be articulated in SCM?

2022-09-05 00:10:11 @AprileBernardi I was skeptical too, till I was told that RCTs have so many imperfections that all they can deliver is a "comparison" between groups eligible for comparison, and all attempts to rectify imperfections are based on assumptions known to be false (in the imperfect world)

2022-09-04 23:35:10 @stephensenn I think even a stupid guy like the one you are mocking should be listened to when he/she points out a few terms in the RCT jargon that could benefit from scientific interpretation. In particular, what RCT findings promise the public, aside from lists of individuals compared.

2022-09-04 22:04:17 @EpidByDesign @robertwplatt @RWJE_BA "May Differ" does not mean "differs". S expresses subjective suspicion in possible differences. More importantly, we are dealing with "definition", not identification. If it so happens that S is too large to allow identification, researchers need to know it too.

2022-09-04 21:51:48 @EpidByDesign @robertwplatt @RWJE_BA In our Fusion paper https://t.co/E1ap1N4ZCh, the "target population" is defined by a set S of characteristics by which the population USING the study MAY DIFFER from the population from which samples are selected for the study. It is thus well defined and formally managed.

2022-09-04 19:41:30 @artistexyz @soboleffspaces It's a little hard for me to envision how system 1 can be more accurate that system 2, when the former must sprout from the latter.

2022-09-04 19:12:11 @stephensenn Some of the "critics of RCTs" are not really "critics" but "enrichers," seeking, for example, to facilitate fusion of several RCT studies. Moreover, if there are aspects of RCT that the enrichers find to be ill-defined, imagine how confused the public must be, the end user of RCT

2022-09-04 19:02:41 @AndrewDGarrett Can such labeling be communicated without the word "causal effect", or "cause", or "effect" ??

2022-09-04 18:59:14 @dataengines @stephensenn Tiny clarification. Pearl does not say answering the other questions is not worth RCT expenses. He says absent Q5 guarantees, it behooves trialists to explicate why answering the other questions is sufficiently relevant to justify RCT expenses.

2022-09-04 08:46:44 Readers seeking additional proofs, references and background material for our "Crash Course on Good and Bad Controls" https://t.co/GkAcnFnCys may find it in this extended survey of "Confounder Selection" https://t.co/M5zW6Xob2y by Guo, Lundborg and Zhao (2022).

2022-09-04 08:05:57 Agree, the cartoon does not show the line up of the other 10%, the Giordano Brunos in our community, who are driven not by thirst for confirmation, but by an ingrained suspicion of conventional wisdom. https://t.co/ZQB3DknXrJ

2022-09-04 06:19:37 @AmbDennisRoss @Ostrov_A I wonder what was it that we have done to give the Iranians the impression that we are so gullible? And.... aren't we still doing it?

2022-09-04 03:20:32 @IsabellaGhement Amazing flower. And every leaf thinks it is i.i.d. !!!

2022-09-04 03:16:24 @blakeflayton This will go on and on till...till we learn to charge @rogerwater and his likes with Zionophobic racism, not deflectable antisemitism. Even @rafaelshimunov would not be able then to say: "He ain't one"

2022-09-04 01:26:09 @artistexyz @soboleffspaces @stephensenn @LGHemkens @dylanarmbruste3 @PWGTennant @JClinEpi My first reaction to the "missing data" myth was: Show me a problem that is not "missing data", eg is Chess a "missing data problem"?My latest reaction: "missing data" is a causal, not statistical problem, as shown here: https://t.co/1BNZSxBPCwit's hopeless w/o causal graphs.

2022-09-04 01:15:00 @AdanZBecerra1 I would be enlightened with much less, e.g., if a trialist could just say who should pay attention to the trial results.The fact that "trial relevance" isn’t rigorously addressed amazes me - but doesn't surprise me given the notional and conceptual deficiency in the field.

2022-09-03 23:07:05 @AMULETAnalytics @uclaextension "Tough read"? How about de-toughening it with a softer book? It shouldn't be difficult once you get the main message: speak science not data.

2022-09-03 22:59:27 @artistexyz @soboleffspaces @stephensenn @LGHemkens @dylanarmbruste3 @PWGTennant @JClinEpi Rubin does mention graphical models, 57:00 min into the video: "The graphical guys, they mix everything up... using observed values.. confusing..where is the insight?.. where is the science?..." My wish: May his words be available to historians of 21st century statistics.

2022-09-03 20:38:30 This construction gives us X=const1 Y=const2,but the CBN: X Y says: X is a random variable distributed as P(X) (same with Y). We need a repair. https://t.co/c9POfz3mow

2022-09-03 15:08:32 @artistexyz @gottfriedmath @soboleffspaces @stephensenn @LGHemkens @dylanarmbruste3 @PWGTennant @JClinEpi I don't see how we can define SCM from CBN.Here is a CBN: X Y no arrows.What is the SCM associated with it?

2022-09-03 08:26:37 My My! What a wonderful gift for my 86-th birthday tomorrow. Knowing that some bookshelves on planet earth carry my thoughts makes me feel immortal and doubly committed to spending the next 34 years of my life searching for the calculus of our mind. Thank you! https://t.co/8gXfvgbAMD

2022-09-03 08:13:25 Clarifying. Of course student groups have a legal right to be selective. But University administrators have a legal obligation to set norms of discourse on campus and tell them publicly: "You are acting against the values of this great university". Care to see it in action? 1/2 https://t.co/izSHfizA9U

2022-09-03 08:09:20 @Claire_V0ltaire Roger is right, he is just a Zionophobic racist, not an anti-Semite. What allows him to continue and fool the world with this mask is Jewish leaders refusal to treat Zionophobia as an uglier form of racism.

2022-09-03 07:37:51 Bad analogy, @Ostrov_A . It invites Ken Stern to argue: Aha! Race and gender are immutable, Zionism is just a political belief. But change "Zionism" to "Palestinians" and watch how haven and earth roar with indignation, as Ken Stern is barred from speaking again on any US campus. https://t.co/jegNbmH1c1

2022-09-03 06:21:12 RT @JewishNewsUK: Lecturer David Hirsh on why he's launching the London Centre for the Study of Contemporary Antisemitism to lead academic…

2022-09-03 05:36:34 I have often been thinking how strange it must be for a Palestinian pupil to sit in a "Palestinian History" class and listen to teachers who never had such classes. https://t.co/HHAcq81ruS

2022-09-03 03:50:38 I would agree with this analogy if you can present a mathematical alternative to SCM. I am not aware of any such alternative, unless we count an exhaustive list of all counterfactuals as an alternative. https://t.co/1ow9e1Y09C

2022-09-03 00:19:43 I think the source of the confusion lies in assuming that an SCM is defined for one individual u. It is in fact defined for EVERY conceivable u. So, nothing is missing nor hidden. https://t.co/aSi8DVeUGZ

2022-09-02 23:50:22 @artistexyz @soboleffspaces @stephensenn @LGHemkens @dylanarmbruste3 @PWGTennant @JClinEpi Mohan &

2022-09-02 21:37:38 I agree. I don't know of any causal quantity that remains undefined in SCM, assuming of course that we take the relation "listens to" as a primitive for defining SCM. As for RCT, the proof in #Bookofwhy is the only defensible proof I know that RCT delivers (asymptotically)1/3 https://t.co/ndyhAyX3oE

2022-09-02 13:20:03 @cristina__m16 I am not sure what PCA is. But a prequisite for causal interpretation of regression coefficients is either (1) physical intervention or (2) causal model.

2022-09-02 10:34:29 I remember how this "3-No's" resolution triggered the West-Bank settlement movement: If we are destined to live by the sword to the end of time, we better do it from a position of strength. Official Arab League renunciation is way overdue. https://t.co/VwDw7dQzc6

2022-09-02 10:11:05 RT @HillelNeuer: On January 1, 2021, Vladimir Putin was elevated to the United Nations Human Rights Council, joining fellow tyrants Xi Jinp…

2022-09-02 09:55:09 RT @analisereal: For the instructors out there—code to simulate the models of the Crash Course in R, Python and Julia!

2022-09-02 02:19:28 @YanChan98780882 Which napkin? the old or the new?

2022-09-02 02:18:16 @EpiEllie @AmJEpi @tweet_epi @TheEconomist @causalinf Not sure I see the confusion. [By definition, confusions are hard to see]

2022-09-01 03:00:50 Hurray, our Crash Course is finally published: https://t.co/Jv8C4b6u94, which makes it not only"looks super useful" but "truly super useful". Confessedly, I often wake up wondering how perfectly good people can walk through life without it. (Names deleted: reviewer-phobia) https://t.co/YxfzQSAxAZ

2022-09-01 02:34:59 @angelamczhou @artistexyz @pablogerbas Is there anything special about the form of your density functions that would prevent you from using the method to estimate ANY density function from its samples? If not, what makes your method "causal" or "counterfactual", rather than "statistical"?

2022-09-01 02:24:18 @angelamczhou @artistexyz @pablogerbas I am not familiar with the term "influence function" though I've seen it used in a couple of papers I could not understand. What is it? Is it not the distribution P[Y(1)=1]? If the "inputs are probability density estimates", what remains to be estimated?

2022-09-01 02:15:00 @angelamczhou @artistexyz @pablogerbas I missed this part in your paper, partly because I could find no graphs there, nor an ignorability assumption usually invoked by graph-avoiding authors.

2022-08-31 22:27:54 I heard Biden no longer reads Fox News, so I've hoped CNN would echo this urgent letter. All I see on CNN is Trump Trump Trump, as if the Ayatollahs promised not to attack Israel till the US elections. https://t.co/FAzbNCmuA6

2022-08-31 21:58:53 Hilarious! I'm pretty sure the guy on the right is a comedian, impersonating Palestinian mentality. And the guy on the left (with the hat) is the producer of the clip: You mean Ukraine?Hilarious! https://t.co/2U7eBFRdjx

2022-08-31 11:20:05 @CotlerWunsh Why? Do you know a West Bank Mayor who thinks differently?

2022-08-31 11:11:46 I once presented this question to a reasonable colleague I've chanced to meet. His answer: Pedagogically, it's seductive to start with "If I treat someone, he either dies or not. Same if I don't," The rest is just missing data. https://t.co/6B1rZjO25I

2022-08-31 10:34:53 RT @JasonMBrodsky: #BREAKING: Nearly 30 House Democrats have signed onto a draft letter expressing fresh concerns about the #Iran nuclear d…

2022-08-31 10:28:41 The economic situation in Lebanon must be super rosy for its Energy Minister to take time off and play kindergarten near the Israeli border. Let's hope he is allowed back to office by Nasrallah. https://t.co/uj8uQ1NcaT

2022-08-31 08:15:40 @soboleffspaces @artistexyz @angelamczhou @pablogerbas To make it even clearer, I would say:"L3 answers another question: How likely is it that *THIS SPECIFIC* outcome would have been different if the treatment had been different?Once we condition on the observed outcome, we can no longer express the question in do-notation.

2022-08-31 06:07:59 RT @eitansgarden: The murder of Eyal Haddad wasn’t antisemitism, it was criticism of Israel.

2022-08-31 06:00:25 Are they considering twinning with Shderot? After a decade of Hamas rockets they can surely use some Scottish support. https://t.co/cEpbdqFCJR

2022-08-31 05:27:58 @stuz5000 @artistexyz @soboleffspaces @eliasbareinboim Thanks for clarifying. I get many annoyed tweets which I cannot decipher.

2022-08-31 04:10:26 @angelamczhou @artistexyz @pablogerbas I just noticed that you are a co-author on this paper, so perhaps you can resolve our puzzle. What is the input to your system? Do you estimate ATE from RCT data? Or from some other kind of input. If the former, what makes it "causal"? Isn't it standard stat. estimation?

2022-08-31 03:50:19 @stuz5000 @artistexyz @soboleffspaces @eliasbareinboim I don't get why you are annoyed at me. Have I been careless in my treatment of counterfactuals or in emphasizing their importance, or in not calling out careless authors?

2022-08-31 02:28:12 @stuz5000 @artistexyz @soboleffspaces @eliasbareinboim I don't even recognize the 1980's. There is no d-separation, nor graphoids

2022-08-30 22:20:25 One more mini-tweet: "And the U.N. expects humanoids to takes its other "commissions"seriously!" https://t.co/Sdz0K0nwby

2022-08-30 22:04:57 @RabbiWolpe The other one is "Chutzpa". Which only Larry King was able to define: "Hadassah organizing a fund-raising event in Khadafi's Libya", not as a provocation, but because it's a really "noble cause."

2022-08-30 18:22:15 @artistexyz @pablogerbas @angelamczhou Even Rung-1 estimators love to say they are doing "counterfactuals," a word that connotes modernity and forward thinking. That is why I urge people to reserve this word to task that cannot be accomplished by Rung-1 or Rung-2. For example, estimate E[Y(1)|Y(0)].

2022-08-30 17:43:10 @analisereal @artistexyz @soboleffspaces Can you give us a 280 characters description of what "influence functions" are, where they come from, and what ordinary mortals should think about when someone tells them "influence functions".

2022-08-30 17:00:44 Beg to disagree. Pearl does not claim that "you can only do counterfactuals using SCM." He claims that "if you want to understand what you are doing, e.g., defend or test your assumptions, then you need to know that counterfactuals come from SCM." See https://t.co/etASSEk4Or https://t.co/cPIkgks8TR

2022-08-30 16:48:00 @artistexyz @soboleffspaces ATE too is Rung-2. To avoid such confusions, I recommend using do-expressions for Rung-2 and subscripted (or parenthetical) expressions for Rung-3. But isn't the paper really Rung-1, given the identification is assumed away?

2022-08-30 06:49:37 The long abandoned children of Shderot truly deserve a visit of their heroine Gal Gadot. She grew up among them (figuratively), we all did, and we all know what they had to experience past year. Still, the NYT considers their plight unworthy of a mention. Thank you, @GalGadot https://t.co/W2N6hrYvKG

2022-08-30 05:09:09 Glad to receive a confirmation from someone who understands the paper. I didn't, because even the Interventional mean in Example 1 is ill defined

2022-08-30 04:59:55 It's linking to Causality chapter 1 https://t.co/1h9ZT0CuCx https://t.co/9eT3m06F91

2022-08-28 12:47:41 @MariDTakashima @ChristinABuelow Welcome to the island of commonsense. The food is not exactly on the gourmet side, and the sharks take their toll, but the good company makes up for it.

2022-08-28 12:07:30 @ChristinABuelow Keen observation and, true, I don't think I would ever get to ask the questions I did ask if I were say a statistician or economist. It was quite early that I was shaken by the realization that we are causal machines and, therefore, AI should become a causal technology.

2022-08-28 07:45:29 To me it does not sound antisemitic.To me it sounds like saying "People of wisdom, moral integrity and deep commitment to peace control the US government" which is a blessing of good fortune for America. I wish it were true. @AdamMilstein https://t.co/vNM97O7oA2

2022-08-28 07:13:17 Great! I was always suspicious of "Degrees in Conflict Resolution". All the failed "peace makers" (John Kerry, Denis Ross, etc) had such "degrees". But with Maitha graduation I begin to think: Perhaps there is a solid body of skills and knowledge deserving of the title! https://t.co/1kyVtog5uI

2022-08-28 06:07:01 @elderofziyon @jjz1600 We must add @jjz1600 to the list twisted minds who rush to label Israel "apartheid" and who fail to understand the harm they are causing to those who truly experienced it. Sad to see a word once stood for human struggle for equality turn into a word soiled with twisted minds.

2022-08-28 05:14:40 This is a profound statement that has to do with the monotonic character we wish information to have. The things that a message excludes increase monotonically as more evidence accumulates, not so the things that a message permits. https://t.co/aETYzCHPOn

2022-08-28 05:06:42 @DrMikeH49 @AvivaKlompas @Claire_V0ltaire Thanks for the link. Chemerinsky's semi-encouraging words can be explained by the kind of Zionist he is. Instead of saying "I am a proud Zionist myself, who isn't?" he had to bow down and beg for legitimacy saying that he “condemn[s] a lot of Israel’s policies". Now he is Kosher.

2022-08-28 04:12:31 @AvivaKlompas @Claire_V0ltaire Do you have the source?

2022-08-28 04:04:27 @blakeflayton How proud He would have felt.

2022-08-28 02:19:18 @harrysfrand @ilovemilfrryy Some people can't grasp that "Peace sign for Palestinians" is an insult to the "Palestinian narrative" and a death warrant to the "Palestinian struggle". "From the River to the Sea, Peace will not replace us!" @Eve_Barlow

2022-08-28 01:20:17 Agree. Taking counterfactuals out of history books would be worse than taking Algebra out of science. https://t.co/yct7P743YG

2022-08-27 22:57:01 @tdietterich @khademinori @ChrSzegedy The question is: What have we learned from the outcome of the first game, that we did not know before hand? If the answer is: That the opponent has weakness W, then the question is counterfactual. Otherwise, the first game is redundant and our question is: find effect of move M.

2022-08-27 22:47:00 the ultimate aim of all assumptions is to encode “the reasons for missingness” which is a causal, not a statistical concept. "We've added this paragraph to pacify a reviewer who claimed: "I don't need graphs to think about conditional independencies". Yes, reviewer, you need!

2022-08-27 22:46:59 To explain the virtues of graphs in missing-data problems, here is what we write in our paper:"The transparency associated with graphical models stems from three factors. First, graphs excel in encoding and detecting conditional independence relations, far exceeding the 1/3 https://t.co/AhESv0BtmB

2022-08-27 22:30:19 @tmorris_mrc @dailyzad @stratosinit If it's true, then you would be the first humanoid I met who claims to feel more comfortable thinking about conditional independencies (among observed and unobserved events) in his head than with causal graphs. Can we test it on some missing-data problem which we both understand?

2022-08-27 22:13:46 @yisongyue @tdietterich @khademinori @ChrSzegedy This is "effect of action" not a counterfactual question, unless you learn something from your actual move that you did not know before hand.

2022-08-27 21:56:38 @SamelDavid @PeterBeinart There is a typo in your tweet. The belief that Jews should have equal rights with others in called "Zionism", also going by the slogan: "Two states to two peoples, equally legitimate and equally indigenous", which Zionophobic bigots resist tooth and nail.

2022-08-27 21:42:38 @tdietterich @khademinori @ChrSzegedy What would be a typical counterfactual question that AlphaZero would be able to answer?

2022-08-27 21:34:55 Any debate related to #LordsParadox is worth watching, especially one that concerns inference about "change". #Bookofwhy https://t.co/5rxXLd60Nc

2022-08-26 10:42:53 A beautiful summary of what being a Zionist means, and why the word "zionophobia" drives the conversation from what Israel does to what Israel is, casting the former in the context of its neighbors' rejection of the latter. https://t.co/StXdp9XPL7

2022-08-26 09:53:49 @WickedViper23 "ML = stat" is nasty because it reduces ML to tasks achievable by statistical methods. This rules out many ambitious goals on the aspiration list of ML researchers, for example, causal analysis, data fusion, transfer learning, explainability etc https://t.co/WXQaA4NVjH

2022-08-26 08:21:18 Chapter 6 of Causality should be instructive to our discussion https://t.co/h04jmLN7v8It is titled: "Why there is no statistical test for confounding, why many think there is, and why they are almost right" https://t.co/i7XBlo6zAc

2022-08-26 07:32:47 This overview confirms a nasty allegation that machine learning is just statistics on steroids.I hope someone from #datasciece rebels. https://t.co/AG93my3C3g

2022-08-26 06:43:16 Insightful analysis of a recent Count decision.My take: employees who object when the boss peddles disparaging comments about their race should argue rather than sue.Not finished: Students who find a teacher doing it should demand the President's resignation, rather than sue. https://t.co/D8KFbbizt2

2022-08-26 05:50:10 RT @elderofziyon: The road to redemption starts with only one word. https://t.co/aUrrDcRs56

2022-08-26 04:10:28 @osazuwa @soboleffspaces @spphubc @eliasbareinboim "Inductive bias" is a term used by some deep learning folks to say: We know that certain tasks must be done by causal modeling but, since we feel awkward doing causal modeling, let's call it "inductive bias"

2022-08-26 00:18:17 RT @Kasparov63: Important article, especially for those who believe Putin will never face justice. https://t.co/ymcTGYOHLB

2022-08-25 23:34:07 Just Like with Afghanistan, Biden Must Have His Deal with Iran https://t.co/DuPSmv3MS8 via @jewishjournal

2022-08-25 23:21:27 @osazuwa @soboleffspaces @spphubc @eliasbareinboim Last time we discussed "inductive bias" I felt that it is a catch word that does not covey more information than, say, "be careful" or "some theories are a priori preferred to others"or "it's all magic". I prefer therefore to talk SCM, and identify features needed for each task.

2022-08-25 20:29:25 The twisted minds who decided to label Israel "apartheid" did not foresee the harm they were causing to those who truly experienced it. Sad to see a word standing for human struggle for equality turn into a word soiled with twisted minds. https://t.co/yogvLgpNXw

2022-08-25 18:48:54 Students asking for "empirical rule for the presence of confouding" are probably seeking some pattern in non-experimental data that detects confounding. Such a pattern does not exist, since "confounding" is a causal, not statistical notion. Eq. (8) is only valid if we have a 1/2 https://t.co/dS8d7rql6q

2022-08-25 08:29:26 As we remember the 1929 Hebron massacre, we should not forget that it was ignited by the Mufti Haj Amin El-Husseini when he came up with an ingenious theory "The Jews are destroying Al Aqsa Mosque." The Mufti is forgotten, but his invention remains with us to this very day. https://t.co/UpR9iX0O3B

2022-08-25 08:19:56 It's the first time in my life that I wish I could believe Saudi sources. https://t.co/wb4K4MHpIl

2022-08-25 08:16:31 @GadSaad @NatureHumBehav I've been laboring for two decades to define what "harm" and "benefit" means in toy problems, with 3-4 variables. Now, these editors claim they can predict "harm" and "benefit" for the entire human race, for the foreseeable future and beyond. I wish I was an "Editor".

2022-08-25 07:44:59 @Eve_Barlow Deplorable? Yes, but they have been saying it for the past 100 years! There is virtue in openness - others plan in sobriety.

2022-08-25 07:32:05 Glad the USC story made it to the press and, reading through the responses of Jewish leaders, I seem to be the only one to focus on the core issue: "Zionophobia". Moreover, as I read the convoluted way Dean Soni Varun defined anti-Zionism, I get the strong suspicion that 1/2

2022-08-20 11:13:33 RT @HananyaNaftali: There is nothing like the streets of the old city of Jaffa in Israel. https://t.co/D0ZqqPOOFj

2022-08-20 10:53:44 @Dam_Nuwen @Sjnjerak Shocked? Why? It's just commonsense. For next step, I'd recommend Primer https://t.co/M5bbaH64GBStats textbooks? For haven sake, not in 2022!

2022-08-20 09:57:33 @DChristensen73 @ProfMattFox Most SEM researchers are still stuck in the linear universe of SEM's founding fathers, so they use product terms as nodes in the graph. Non- parametric analysis assumes all possible interactions are possible a-priori, and goes ahead to quantify their magnitudes, IF NEEDED.

2022-08-20 09:48:43 We are a week away from celebrating 125 years to the First Zionist Congress, Aug. 29, Basel, 1897. Time flies, but @GilTroy's story will make your time freeze: https://t.co/kJypoPyKqVand my humble take will add an unexpected color to that pivotal event: https://t.co/ICqKCBgrOh

2022-08-20 09:15:58 @KhaledAbuToameh The supreme arbiters of right and wrong have developed a quantitative measure for crime severity, and they can't understand why some people chuckle.

2022-08-20 07:36:24 @b_surial @doaalsaleh @ProfMattFox Extremely unlikely is not a reason to leave an arrow out

2022-08-20 07:32:22 @JonMinton @ProfMattFox Every DAG imply the existence of some function Y=f(pa(y)) behind every parents-child family. If all are binary, f() must be a Boolean function, namely, a logic gate.

2022-08-20 07:28:34 @MD3_Phd @WillBall12 @MKellyIrving @ProfMattFox Its a solid paper, though a bit ambiguous on the meaning of "within levels of", i.e., do we hold the level constant? Or condition on it.? More generally, why label a variable "effect modifier" instead of estimating the quantity of interest, say, P(Y|do(X), do(Z)) ?

2022-08-20 07:16:17 @MKellyIrving @ProfMattFox Interaction? Yes, with a hidden, unnamed variable."How best to feature that in a single DAG?" Retain the arrow Z-->

2022-08-20 07:09:58 @DChristensen73 @ProfMattFox DAGs do not depict interactions because interactions are assume to hold by default between any two parents of any child.

2022-08-20 07:08:08 @DChristensen73 @ProfMattFox It is "interaction", but between A and a hidden variable that may not have a name, but characterizes some unit, say those having an unknown type of allergy.

2022-08-20 07:05:30 @sweichwald @ProfMattFox No need to invoke/exclude higher moments. See example on page 35-36 of Causality.https://t.co/PqATlrRh91Causation exists even when Prob(Y=y|do(A=a)) = Prob(Y=y|do(A=a')) for all y, a and a'.

2022-08-20 06:55:51 @PWGTennant @AndersHuitfeldt @ProfMattFox Another such example, involving three binary variables, is discussed on page 35-36 of Causality. It goes even farther: 'causation does not necessarily imply average causal effect (ATE)'

2022-08-20 06:45:58 @EpiEllie @ProfMattFox "faithfulness" is assumed in causal discovery (ie, constructing the DAG from data alone), not in causal inference where a DAG is constructed from expert knowledge. The knowledge needed in the construction is individual-level, which some call "sharp null hypothesis".

2022-08-20 06:37:19 @soboleffspaces @ProfMattFox Glad Elwert got it right, saying: "to anyone"

2022-08-20 06:35:30 @ProfMattFox I see that my reply did not make it to your thread, so I am retweeting. The answer is: Yes. See why: https://t.co/QZVVZYVGcw

2022-08-20 06:22:46 Just checked @bellahadid Twitter. Wow, she hasn't block me! Though I explained to her clearly https://t.co/s8rrv4Nabf what it would take for her to be invited again to Friday dinners - chant the holy oath: "Two States for Two Peoples, Equally Legitimate and Equally Indigenous." https://t.co/wu6V9tJAfJ

2022-08-20 05:42:21 @mishtal @Edinburgh_CC The way to fight it is to propose an amendment:Twin with two cities, one being Shderot. This reasonable amendment might draw some support from the decent camp of the City Council, and adamant resistance from the Zionophobes, shelving the motion, and exposing its shady motives.

2022-08-20 00:02:37 We are in agreement that "Terms such as benefit, harm, kill, cure are not defined properly" in medical practice. I attribute it to the language in power, i.e., statistics, where those terms are indeed provably undefinable. You attribute it to "separate outcomes". I need examples. https://t.co/TLGqWP6rdX

2022-08-19 23:17:45 for a practical explanation of "why": Assume we have a name to that characteristic (say allergy) which we would like to recognize as a variable C in the DAG. Wouldn't we like the DAG to tell us that Y depends on A given C? It won't if we deprive it of the A--->

2022-08-19 09:47:56 The Western Press was so happy to dismiss President Abbas "Holocaust speech" as a slip of the tongue. And here comes this massive parade and tells them: "No tongue and no slip, these are out deepest unshaken beliefs, our mother milk and our water supply. This is who we ARE!" https://t.co/t6vINiXIo4

2022-08-19 06:08:15 Students booted from sexual assault support group for being Jewish | World Israel News https://t.co/XyU5aldHfF

2022-08-19 05:12:29 An important and captivating paper on a Turing Test for "agency". Still reading, but it occurred to me that a more credible counterexample to agency would be a thermostat, not a rock.Would a thermostat meet your definition of a "non-agent"? Still reading. https://t.co/6k3paLh1EJ

2022-08-19 05:01:18 It's my signature, I confess. And "Chazzak V'Nitchazek" (get strong and we all will) used to be my code for Jewish solidarity. I didn't think anyone would take me seriously and make Aliya

2022-08-19 01:42:53 @kareem_carr My best tip: check if the book (video, course, totorial) starts with the Ladder of Causation. If not, I would hesitate, because you would be missing the Rosetta Stone of data-science.

2022-08-18 22:13:58 An unbelievable story! It's the most inspiriting solace I've received since Danny's death. Thanks. As Malcolm Hoenlein once told me: "He chose life." Now I see what he meant. Danny is still living up there, still inspiring strangers to take a look at where the planet is heading. https://t.co/6GnzruM30I

2022-08-18 11:55:08 I tried, but your system says I am in already - can't vote twice. Too bad. In Poland, before WWII, professors could vote twice (so my mother told me). https://t.co/Qrpt7n2wAU

2022-08-18 11:42:47 @HL327 saves others. If you are concerned about "Can this treatment harm my patient" (a counterfactual question) you would act differently under (1) vs. (2). Conclusion: You would act differently from "the usual", assuming "the usual" is acting according to what RCT tells us.2/2

2022-08-18 11:31:35 @HL327 A verbal example of how PNS leads to different decisions from "the usual" is give in p. 3 of https://t.co/rGHIc9kH0O. We see an RTC with zero treatment effect and PNS analysis that decides whether (1) the treatment has no effect on any individual or (2) it kills some and 1/2

2022-08-18 09:24:21 The question that @HL327 is asking troubles many practitioners that I've talked to: How canmathematical analysis of hypotheticals (eg counterfactuals) assist us in making everyday decisions in the real, not hypothetical world? My answer: The questions you ask in real world1/4 https://t.co/iEGIkosUKy

2022-08-18 07:33:23 @akbarzamir "why was it necessary to put a direct causal link between Sex and Final Weight?" Because we cannot rule out the possibility that the effect of W_I on W_F differs for boys, compared with girls.Oh, Lord, Kumbaya!

2022-08-18 02:34:13 Cover Story: AIPAC Doubles Down on its Mission https://t.co/0tUJZBdgyr via @jewishjournal

2022-08-17 19:56:31 @elderofziyon I don't think you get Abbas arithmetic. If the Holocaust is "the most heinous crime in modern human history" and Israel has committed 50 Holocausts, go figure out how evil Israel is. "It's elementary" says President Abbas. "It's elementary" say mama, papa, teachers, clerics..

2022-08-17 19:23:03 Israel, Turkey return to full normalization https://t.co/sSvq4Tjjlk

2022-08-17 18:45:58 @elderofziyon Amazing! From his photo, one would swear he did finish high school. He is making a stupid mistake writing in Arabic

2022-08-17 18:25:00 @PHuenermund I always ask people "where are you from?". In that way I convey to them how proud I am, and how proud they should be as carriers of cultures that add color and luster to our lives.

2022-08-17 10:19:16 I'll weep for you tonight, Malki Roth. And when I weep for you, I weep for my son Danny too - your brother in pain. Malki and Danny, two treasures of humanity, crushed in Barbarian claws, to the cheering silence of those who can't spell "Barbarians". https://t.co/rxK3pyxhB7

2022-08-17 09:48:45 RT @elderofziyon: I sense a pattern here. But it is apparently too subtle for Western media and human rights organizations to grasp. https:…

2022-08-17 09:38:36 RT @Ostrov_A: Hard to tell these days, which of the two is the bigger antisemite and Holocaust denier? https://t.co/51TX6XAxoH

2022-08-17 09:02:43 I love the slogan "All regressions are 'true'". How true! https://t.co/MLYKYcqNn6

2022-08-17 08:40:42 Same goes for this man's advice (2) "Listen to voices outside the foreign policy mainstream." In our region, this amounts to listening to what Israelis are saying: Mutual acceptance first, diplomatic horizon second. https://t.co/gdl1ANLrBH

2022-08-17 08:17:50 @JFronczekMD @smueller The example on page 3 of https://t.co/rGHIc9kH0O demonstrates that such reversal is not unlikely. Note that the RCT study not discarded, but *combined* with the observational study

2022-08-17 08:06:33 @davidboxenhorn @soccerdhg @EinatWilf The most "defining moment" of all is talking to any Palestinian intellectual today, and observing the certainty with which they hold Israel's demise.

2022-08-17 07:54:47 @davidboxenhorn @soccerdhg @EinatWilf For me, and for many Israelis of my generation, the defining moment was the famous "3 No's" of the Arab League Summit in Khartoum (Sep. 1967) in response to Israel's "Land for Peace" overtures.

2022-08-17 07:37:50 "Dangerous consequences" will not deter PA President Mahmoud Abbas from future Holocaust distortions. What will? Reminding him of the causal connection between Palestinian's actions of 1939 and the Holocaust consequences of 1942. https://t.co/5cN0S28XEK https://t.co/6I44TwLQT0

2022-08-17 06:48:29 The word "Holocaust" spoken by Mahmoud Abbas always reminds us of Palestinian's role (1936-1939 riots) in preventing 6 million European Jews from fleeing the Nazis, blocking their escape routes (White Paper, 1939), and sealing their fate in the death camps. https://t.co/ts8Hqbqa6u

2022-08-17 01:58:58 @HL327 @soboleffspaces As I read your introduction to diagnosis reasoning, I can't help by asking what the research question is. IOW, are we searching for a factor X that maximizes the causal effect of X on Y? Or, perhaps, one that maximizes the conditional probability P(x|y), or PN, or PS of X on Y?

2022-08-16 20:39:56 @soboleffspaces If #causal reasoning cannot capture clinical reasoning, what can? Is clinical reasoning beyondcapturing?

2022-08-16 20:34:54 @duffy_maths For "causal inference in economics", I recommend: https://t.co/Jf1SkgV3CK. And, if you can take the time to solve the 12 toy- problems presented (and solved) there, you'll be ahead of 85% of all living economists. @causalinf @econtweeter, @VC31415 , @instrumenthull

2022-08-16 12:21:27 @soboleffspaces @prem_k The first sentence raises the question: What practice is not "evidence based". Do you find it satisfactory?

2022-08-16 12:14:14 @JFronczekMD @smueller This thread is longer than a Govt manual. Can you summarize the doubts and concerns that you find compelling?

2022-08-16 12:04:52 @stephensenn Unfortunately, no progress yet. Travel, conferences, deadlines etc. hijacked my thoughts. It's simmering there, and will I hope yield some resolution.

2022-08-16 01:17:18 If any of us was kept sleepless, wondering why days go bye with no word from CAIR (Council on American-Islamic Relations) on the Rushdie attack, @TheIPT has gotten to the bottom of it:https://t.co/zYLRqNt6R8Put simply: CAIR is on the side of the attacker, spiritually, of course

2022-08-16 00:48:34 @artistexyz @esMktDigital @MatteoCourthoud @soboleffspaces Can you describe the "synthetic control" setup in the language of Bayesian Networks? Such description would be very helpful for communicating with economists.

2022-08-15 23:03:11 learn from an RCT alone?Resolving conceptual puzzles is one of the most satisfying advantages that causal modeling has over black-box methodologies. 2/2

2022-08-15 07:42:07 Palestinian bus-shooter Amir al-Sidawi was incited by Palestinian Islamic Jihad (PIJ), the rockets of which were not condemned by mainstream media (eg, Mehdi Hasan of msnbc), thus conferring legitimacy onto PIJ's ambitions and practices. https://t.co/TfDtyVinrG

2022-08-15 07:09:57 Congratulations! And I thought good math professors, like the one I had in high school, no longer teach math in Israel. So proud to be proven wrong! I am eager to see the questions. https://t.co/VvuQhdhW78

2022-08-15 05:45:40 @Valerie_Harper1 The credibility of Amnesty's "fact finding" in Ukraine, parallels the credibility of their "fact finding" about Israel. Both are products of same institutionalized culture in which facts are enslaved to ideology.

2022-08-15 05:14:11 That was in the old days. No more. Come to college campuses and you'll get unambiguous descriptions of what these terms mean. UC Irvine pioneered this disambiguation with the a week-long celebration of "A World w/o Israel", followed with "Allah is a Terrorist"

2022-08-15 04:55:37 I have friends who've, for years, been supportive of Amnesty International - I'm often wondering if they still do. I've other friends who are telling me: It's impossible that such a noble-aimed NGO could mislead its donors-supporters for so long, so many of them. IT IS POSSIBLLE!

2022-08-15 04:55:36 False Accusations and Ideological Bias https://t.co/fiWDGEMLlX

2022-08-15 00:45:19 Great thread! To be expanded perhaps with "probabilities of counterfactuals"

2022-08-14 20:29:24 Highly recommended!#causaltwitter, #causalinf, #EconTwitter, #Statistics,#instrumentals , #Bookofwhy, #causality, #MLB, #DeepLearning https://t.co/BlRop928jI

2022-08-14 19:23:54 There is no understanding/teaching of causality without an understanding of the Causal Hierarchy. So much so, that I've started calling it Knowledge Hierarchy. @eliasbareinboim etal chapter formalizes and exemplifies the hierarchy to make it understandable to all researchers. https://t.co/wTacFYRArd

2022-08-14 18:17:02 @PavlosMsaouel @gary_lyman @eliasbareinboim @gonen_mithat @f2harrell @Merregan @stephensenn @AndrewPGrieve @Richard_D_Riley @ESMDcan123 Looking forward to read the complete paper.

2022-08-14 17:52:52 RT @realbassemeid: Several children who were injured by misfired PIJ rockets over the weekend have been transferred to Israeli hospitals fo…

2022-08-14 17:43:22 Antizionism/Zionophobia is a despicable form of racism even it were a totally new phenomenon, unconnected to ‘traditional’ antisemitism. https://t.co/vClbByC8U3

2022-08-13 22:41:02 @soboleffspaces @stephensenn @AndrewPGrieve @PavlosMsaouel @f2harrell @smueller @ChristosArgyrop This is one of the greatest victories of scientific thinking. I am wondering what AI architecture could generate such an argument, and from what kind of experience/input. This exercise should be taken seriously by those who argue about symbols vs neurons.

2022-08-13 22:23:08 Senior UN official loses her post after tweet condemning PIJ rocket fire at Israel https://t.co/0P0AHO7PK3 via @timesofisrael

2022-08-13 19:51:57 @soboleffspaces @eliasbareinboim @artistexyz @johnmark_taylor @ylecun @GaryMarcus I don't get it. Who are the "algorithmicists believing in data"? What are 3B-parameter models? Lost in guesses.

2022-08-13 19:45:26 @eliasbareinboim @soboleffspaces @artistexyz @johnmark_taylor @ylecun @GaryMarcus I agree. The paper cited is strictly level-1. The authors state so explicitly: ANN is used for estimation and BN for "interpretation".

2022-08-13 00:17:08 Let's not overlook the fact that "distribution drift" has been solved completely in the microcosm of causal inference which, under normal circumstances, would trigger the curiosity of Titans from the AI macrocosm to ask: How? Can it be generalized? Can DL do it too? https://t.co/g7Eb0mimF2

2022-08-12 23:57:57 @matloff @MichaelALewis10 I've given this trend more that a serious thought. Can you imagine me using the terms "revolution" and "New Science" without givingit a serious thought? Additionally, CI is not a "Trendy New Thing", it is a New Thing turned trendy after 3 decades of hard struggle with stats.

2022-08-12 22:12:00 @matloff @MichaelALewis10 Please, do not "leave it at that". Please tell us what exactly is "unfairly dismissive of stat". We need to understand why statisticians feel offended when offered tools to express what they wanted to express but couldn't. See my interview with D. Hand https://t.co/XFZYzAbLHy

2022-08-12 19:56:01 @MichaelALewis10 @matloff I would add to it: (1) causal assumptions is what's needed for CI, nothing else would do. (2) causal assumptions are defensible, because we think in causal, not statistical, assumptions, and definitely not "ignorability" assumptions.

2022-08-12 18:07:59 @MarielzaTalks @GaryMarcus @ylecun Which raises two interesting questions: (1) Why are people interested in debates more than, say, new polynomial time algorithms? (2) Looking back at your personal research, how much have you learned from debates compared to solving intellectually challenging problems?

2022-08-12 13:40:21 @johnmark_taylor Bottom line: Identification = Symbolic, Estimation = Deep Learning.More details: https://t.co/dxqdoqMsQmhttps://t.co/KBGKcmJxM8Also:https://t.co/61qez08cgm

2022-08-12 13:22:48 Thanks for correcting the link. https://t.co/jfkAGuvfZU

2022-08-12 13:17:37 @DialSquareNZ @idatheleader @mishtal The number gap has always been against us, yet the world came to see our side in Nov. 1947, when we proudly touted our cause and exposed our enemies intentions. Today, we are moaning: Oy Vey! Antisemism! (which no one takes seriously) Instead of shouting: Zionophobes! Get out!

2022-08-12 12:57:59 An interesting debate between @GaryMarcus and @ylecun on the tension between neural vs. symbolic architectures in AI: ttps://twitter.com/GaryMarcus/status/1557768248841703426For us, in causal inference, the symbiotic harmony and division of labor between the two is self-evident. https://t.co/iTCDqyX9f5

2022-08-12 04:52:41 @artistexyz @soboleffspaces @ylecun @GaryMarcus When the Titans are fighting up there, on Mt. Olympus, it's a perfect time for us, mortal sailors in causal inference, to sail the revolution ship forward, many miles forward.

2022-08-12 03:47:50 @kareem_carr And thank you, for giving us a proof that racists can still be handled as in the old days.

2022-08-12 03:30:11 Great! I think every respectable university should hire a couple like this, to help professors advertise new classes. Gee, just think what they could do with Stanford's "Foundation Models". https://t.co/pQobIqRawD

2022-08-12 01:53:24 @robwilliamsiii I would second your recommendation of #Bookofwhy which, btw, is not very charitable to Reichenbach's principle of common cause -- he forgot about colliders.(He is forgiven though, since, like most philosophers, he didn't have graphs to check ideas)@PessoaBrain

2022-08-11 21:39:38 Very interesting paper, thanks for shouting out to me. I am wondering whether all these features: "causality, role, knowledge, objective foreseeability, capability, intent, desire, autonomy, and character" are independently needed for computing "responsibility." Poor Jury. https://t.co/dgOShcIZUC

2022-08-11 21:22:25 @EliKohn3 @isaacdecastrog Gee, and my Bible goes all the way to Abraham: "To the land that I'll show you.." I bet it was printed by AIPAC.

2022-08-11 21:10:11 @mishtal When there is no speaker calling "Zionophobes out of London" passerby-s think these thugs are human rights advocates, and then some even join them. Words count, and we fail to use them.

2022-08-11 18:57:11 @dataengines Sorry, I did not understand what your argument was from the "thermometer" metaphor. I scanned your paper and its not clear to me whether your PNS (prob of necessity and sufficiency) is the same as in https://t.co/Asxfs1E2vr and how you compute its bounds from data.

2022-08-11 10:47:08 @holdmytowel @stu_frost @jhurwitz This trend towards "Inflated Expectations" was also noticed by @YiMaTweets https://t.co/lt2UH1GqAL who called it "the explosion of new names". But I'm not sure the choreographers of "Foundation Models" really have inflated expectations

2022-08-11 10:23:28 I meant to say: Thank you @SadhguruJV.

2022-08-11 10:20:38 @elderofziyon I don't believe @CAMERAorg captured the real difference between @MehdiHasanShow and @AvivaKlompas . He calls her "racist" 3-4 times, and she treats him as a rational unbiased reporter, not once labeling him "Zionophobic racist"

2022-08-11 09:59:32 @TsviArii @Claire_V0ltaire What do you think has gotten us out of Egypt, being Bundists?

2022-08-11 09:54:19 @holdmytowel @stu_frost @jhurwitz It's the peak of "Inflated Expectations".

2022-08-11 09:50:58 Thank you @JFREJNYC for reminding us of things we tend to forget about Israel. And to imagine that all this was created under 75 years of relentless assaults by Zionophobic neighbors and other haters, @vinojininair. https://t.co/Nbio3glcnO

2022-08-11 09:27:48 @skeenscene @juliajassey @JGreenblattADL @JFREJNYC @AdamSerwer Those who understand what BDS stands for, especially its Zionophobic aims, are Ex-Jews, the rest are "useful idiots" - exonerated by reason of thoughtlessness.

2022-08-11 09:22:49 @skeenscene @juliajassey @JGreenblattADL @JFREJNYC I looked at @JFREJNYC's Pinned Tweet and I am not sure the slogan "all Jews are Jewish" makes any sense. Zionophobic Jews are certainly Ex-Jews because they renounce the central component of Jewish peoplehood. What about BDS-supporting "Jews"? Good question. Those@AdamSerwer

2022-08-11 08:43:57 @VivianBercovici @Claire_V0ltaire The titlte "Jew Hater" is bad because no "Jew Hater" would admit to being one. Compare with the title Zionophobe, of which they are all proud.

2022-08-11 08:39:19 @DrewBHeadley Gelman's book could be a starting point for writing the badly needed "Regression Analysis: a causal perspective" . My opinion of Gelman's style of causal inference is summarized here: https://t.co/j7IDEsMklc

2022-08-11 08:09:57 @stu_frost @jhurwitz It's wonderful to see "causal ai" being noticed by technology futurists but, for some reason, I do not feel comfortable seeing "causal ai" placed on the same curve with "foundation models."

2022-08-11 01:08:47 Speaking of Zionism, Judaism, Jewishness and Israel -- here is the most profound lecture on the topic, by Yossi Klein Halevy: https://t.co/zzWH84FWt1Main motto: "We didn't survive the Holocaust, we defeated it!"@JGreenblattADL @michaeldickson @Ostrov_A @EinatWilf @blakeflayton https://t.co/E2w6wOknhY

2022-08-11 00:53:10 Now, suppose they are right. Suppose "Zionism isn't Judaism". Does this make Zionism less noble? Less inspirational? Does this make Zionist Jews less deserving of rights to dignity, normalcy and self determination? Less deserving of admiration for what they have accomplished? https://t.co/E2w6wOknhY

2022-08-11 00:38:51 It's time to examine this week's harvest of papers in Causal Inference: https://t.co/LeG6B2dBK7I se increased interest in "fairness" and "responsibility".

2022-08-11 00:28:43 Time to this week's harvest of new papers in Causal Inference: https://t.co/LeG6B2dBK7.I see increasing interest in "fairness" and "responsibility".

2022-08-10 23:53:46 @juliajassey @JGreenblattADL This is all very true in theory

2022-08-10 23:42:40 @Claire_V0ltaire This is partly due to our timidity and unwillingness to use a fighting word, Zionophobia, to assert our readiness to defend the nobleness of our cause and to expose the moral deformity of anti-Zionism.

2022-08-10 23:34:14 @lackboys3 @bennettjake1 @IlhanMN The way to test it is to donate "out of area" money to defeat @ilhaMN and see if it turns "shadowy &

2022-08-10 22:23:42 Jokes aside. Statistics is in dire need of a textbook on "Statistical Regression Analysis: A Causal perspective". It should be easy to write, just take the most popular regression textbook written, and explicate the thought that the author has been trying to hide. A best seller! https://t.co/laHAYxCvcx

2022-08-10 21:31:12 I have been waiting for a re-run of Moshe Vardi's Lecture: "How to be an Ethical Computer Scientist." It is now available on youtube: https://t.co/JhYqtJnvzMThe painful lessons for us, computer scientists, begin about 20 min. into the video.

2022-08-10 19:08:43 @causalinf When was this workshop?

2022-08-10 18:37:14 Beautiful quote by Leonardo! Which I would use to argue for SCM:"understand the model and you will have no need of the experiment, be it "controlled", "natural", "quasi" "targeted" or "well-defined". https://t.co/as9PhfXSh6

2022-08-10 18:28:19 @shiwalimohan Lots of wisdom in your advice

2022-08-10 18:23:58 @AnoopRKulkarni Welcome to the causal revolution and, as Churchill once said: “I have nothing to offer but [ ... ] commonsense and victory" (revised to fit the circumstances.

2022-08-10 14:01:58 @YiMaTweets I suspect you overlooked the causal revolution, where you will not find explosion of new names, but explosions of new ideas and new solutions of problems that were lingering for generations. Including, BTW, ML problems such as transfer learning, data fusion, and explainability.

2022-08-10 10:49:33 @causalinf It was an economist (Arthur Cecil Pigou) who first noticed collider bias (see #Bookofwhy, p198) but Economists lost direction once they accepted PO philosophy.

2022-08-10 10:34:26 @axel_red_third @HumzahAlkindi @MehdiHasanShow @RashidKhalidi1 @marwanbishara @FranceskAlbs It was me. I call Zionophobic Jews Ex-Jews orYester-Jews, stressing the observation that Zionism is a key component of Jewish identity, perhaps THE key.

2022-08-10 10:28:56 @HumzahAlkindi I forgot where we stand on the 3rd Edition.Cambridge U. Press was essentially closed during the pandemic. I'll try to check again.Thanks for your interest.

2022-08-10 07:49:03 I'm not sure if I shared with you my March 8 interview with Roi Yozevitch, who believes that Bayes Rule in the key to scientific thinking. It is available now on youtube (In English). https://t.co/IXlh9sKnvmWe cover probabilities, causation, Hume, Eratosthenes, and more.

2022-08-10 06:42:15 @robwilliamsiii Blaming is a dangerous occupation. I would rather blame it on commonsense, which #Bookofwhy tried to capture to some extent, and which so many researchers hoped would sprout on its own, given more data.

2022-08-10 05:46:34 @HumzahAlkindi @axel_red_third Of course it doesn't mean that 2+2 is 5. But it means that @MehdiHasanShow is a Zionophobe, and @RashidKhalidi1 is, and@marwanbishara is ...(BTW, name a Palestinian who is not a Zionophobe), and so is @FranceskAlbs.I can identify them from miles away.

2022-08-10 02:50:31 I remain cold, silent, skeptical and somewhat surprised at any attempt to do causal analysis without (1) the First Law (2) the Ladder of causation (3) Back-door criterion etc. etc. It's like shopping in a strange city without a city map. You may find a few bargains, but ... https://t.co/DYoScYlge9

2022-08-10 02:27:32 This is a more comprehensive survey, which covers some of the latest advances in the field, like transportability, sensitivity analysis and fairness.An important addition to the Causal Machine Learning avalanche would be https://t.co/61qez08cgm https://t.co/uM0smNJU7x

2022-08-10 02:13:32 Short but good paper to wet the appetite of RS readers for the emerging field of Causal Machine Learning. Missing ingredient: The First Law, i.e., that all potential outcomes can be derived from SCM, without which the impression is created that SCM and PO are non-intersecting. https://t.co/vqZOQQCnB6

2022-08-10 01:11:04 @FabItMart @EpiEllie @pietrobiroli @PHuenermund @instrumenthull @paulgp @dlmillimet @eliasbareinboim I've glanced at the paper. It seems restricted to linear analysis, so the basic issues we have discussed on the definition of natural effects do not surface.

2022-08-10 00:41:57 RT @DavidDeutschOxf: This seems to be the Prof Weizenbaum who wrote ELIZA, worried about AI, and wrote a book saying that AGI is impossible…

2022-08-10 00:37:49 Hey, Marwan, are you willing? I like you conclusion: Israel could still turn all that hatred into tolerance,... if only it has the courage to atone for its violent past...and start treating [Palestinians] with the respect and honor they deserve as equals,.. in their homeland.2/3

2022-08-10 00:37:48 I'm not sure how much of it is conspiratorial,and how much is deep genuine beliefs. The writer, @marwanbishara, is a great source for learning what Palestinians truly think about Israel. He might even be willing to engage with us to examine what myths we should both shed.1/3 https://t.co/2UZe3wMnjM

2022-08-09 20:46:01 RT @AvivaKlompas: Hamas ordered journalists not to report on Gazans killed by misfired Palestinian rockets and were told to blame Israel fo…

2022-07-27 08:25:09 @PHuenermund Yes. The problem Manski complains about is "missing data", not "measurement error", but his aversion to using graphs is universal, stronger even than Imbens and Rubin.

2022-07-27 07:40:33 Thanks for setting up the audio-visuals for my talk tomorrow, and for alerting me to update my slides. See you at @IJCAIconf 2022. https://t.co/LnCHuOkprz

2022-07-27 07:32:29 @PHuenermund Manski does not like Rubin's imputation, for reasons different than ours: https://t.co/1BNZSxBPCwBut when he talks about "credible assumptions" you can tell he hasn't seen "missingness graphs" yet, without which one cannot judge the credibility of one's assumption. Try it!

2022-07-27 02:05:01 @erikbiz @JewishJournal Not only Arab countries are "threatened", every Palestinian person is personally threatened by the presence of a sovereign Jewish state, no matter how far, no matter how small, no matter how meek.

2022-07-27 01:21:58 Education Dept. to Investigate USC’s Handling of Antisemitic Harassment Against Rose Ritch https://t.co/3UDrAs42gY via @jewishjournal

2022-07-26 17:04:44 @papa_cientifico In a specific example? Or in a formal model?

2022-07-26 12:03:22 They highlight the difference between a professional journalist, who never forgets the human face behind the news, whatever the news, and a propagandist, who consistently dehumanizes one side in a conflict. I have humbly asked: In her 25 yrs of covering the Middle East, 2/3

2022-07-26 12:03:21 Yesterday's article by @DrMichaelOrenhttps://t.co/d1eYiDj0Gbmentions a comparison (made in @TheAtlantic) between the death of #ShireenAbuAkle and the premeditated murder of my son, Daniel Pearl, in 2002. It compels me to reiterate my reaction to such outrageous comparisons:1/3

2022-07-26 01:28:19 This week's harvest of Causal Inference papers is listed here: https://t.co/rvKkkMi9OP, for your scrutiny, enjoyment and panoramic overview of where the field is going.

2022-07-25 17:49:41 Why I am helping to start a new Jewish Studies Network? Adam Fuller speaks for me. Plus, I have seen 1st hand what happened to Jewish Studies at UCLA and neighboring campuses - our students deserve scholarly alternatives.https://t.co/EgqsBYQsji@GillTroy @EinatWilf @blakeflayton

2022-07-25 16:41:49 The celebration cannot be complete without Steve Zipperstein's in-depth analysis of the legal and historical forces at play: https://t.co/BGy6n0act0 https://t.co/iP4QNNc71L

2022-07-25 16:31:18 Naïve me! And I was taught listening to the other side is a moral duty. https://t.co/SDTvpz7OS5

2022-07-25 13:57:44 @DKedmey That's a tough one. Partly because I have zero experience with Times Square campaigns, and partly because passengers speak Rung 3 routinely, e.g., "Are you sure, doctor, that this surgery is worth my risk?" Nothing to be intriguedby.

2022-07-25 13:47:23 @artistexyz @maxkasy @HarvardEcon It's worse than geometry without figures. We can do geometry without geometrical figures because our assumption are defensible (eg. Euclide's axioms). In Potential Outcomes, the assumptions are conditional ignorability assertions, the plausibility of which humans cannot judge.

2022-07-25 11:25:06 @artistexyz @maxkasy @HarvardEcon Nothing wrong with DAG aversion, if you can substitute its gold mines with comparables.Can Potential Outcomes substitute for transparency, selection bias, covariate selection, data fusion, testability, more? I've left it as a home-work exercise in 2012: https://t.co/qnvu54CxNg

2022-07-24 17:23:40 @DavidDeutschOxf @artistexyz @paul_snively Confessing ignorance: Where can I find what Everett says about "Potential Outcomes"?

2022-07-24 16:22:11 @artistexyz @paul_snively @DavidDeutschOxf I once lamented thus: Probability function P is a powerful thing to have, it even tells us, on its own, how P ought to change when an observation X=x is received. Unfortunately, it does not have the power to tell us how P ought to change when an intervention do(X=x) is enacted.

2022-07-24 14:30:04 @NeilHudsonC_R @ToKTeacher @DavidDeutschOxf The notion of "intervention" is not defined in probability theory. If it were, then causality would be reducible to statistics and "correlation does not imply causation" would be false.

2022-07-22 11:10:55 @stephensenn @smueller We do understand the problem intuitively, but when we speak about "calculus", we need to translate our understanding into formal specification. That's why I am interested in how Nedler specifies the two designs. BTW, would things change if we replace "centers" with "age-groups"?

2022-07-22 09:37:08 @stephensenn Very little progress, I confess. Partly due to my Spain trip, partly due to remaining misunderstanding of the problem. Here is where we could use some help: How would you describe the 2 designs in Nedler's calculus? Not the solution (which you gave us) but the PROBLEM setup.

2022-07-22 09:03:05 If you are attending the ICML conference, hungry to learn about non-iid data, don't miss our paper:https://t.co/XBFL1Rsz9nIt will be presented by @zcccucla, at Session P15.

2022-07-22 08:52:54 I haven't mentioned other gold mines, because I assumed RCT trialists are interested only inAverage Treatment Effect, nothing else. For enlightened trialists interested in distinguishing "harm" from "benefit", we have a 3rd gold mine -- Rung-3 analysis: https://t.co/mm9LnNtBCk https://t.co/3LxRg38d7s

2022-07-22 03:07:24 New Poll: Andy Levin was not very smart when he endorsed Rashida Tlaib and thus turned this election into a one-issue race: Whether she is fit to serve in the US Congress. He miscalculated how strongly voters feel against Zionophobic racists. https://t.co/UoHRfjRyKJ

2022-07-21 21:37:35 @AbigailShrier Do you think it's worth asking @womenmarch if Zionist women are women? Or simply remind them how Linda Sarsour decided that my sister was not pure enough to march with them? See https://t.co/2nCKQxaZHX

2022-07-21 21:15:57 @GadSaad @JoeBiden @KamalaHarris As an American and an uncompromising atheist, I pray to the All-mighty! Oh God! Not @KamalaHarris. This country needs a Golda Meir!

2022-07-21 21:03:29 @AndrewPGrieve Well put! Carefully avoiding the phrase "the true value".

2022-07-21 20:56:17 @exactsake I don't know how to answer your question, because you are using probabilistic terms and, we know, they are orthogonal to causality. "No causal in, no causal out" (Cartwright)

2022-07-21 18:34:37 @mishtal This was one land auction I always kill myself for not bidding on.

2022-07-21 14:33:17 @GedigMoritz You start by defining what you want to "find out". Note that C is conditional dependent on A, given B. Note also, that this example illustrate the boundary between Rungs 2 and 3. (see Causality page 24)

2022-07-21 14:26:54 You are right! It was July 21 2002, 20 years ago that Bernard Lewis wrote this Essay for our "I am Jewish" book. It may have been his influence that made me avoid the word "Judaism" in favor of "Jewishness" - the former is overly theological. https://t.co/l8uVfIWwe2

2022-07-21 14:14:17 @Meh3625555 @PosenIzzy My problems with EDI: (1) No list of "ethnic groups" is available. (2) "Jews" are not mentioned in ANY of their official documents. (3) If we demand Jewish representation, they will choose JVP, which is worse than no Jews. Campus discrimination targets Zionism, not Judaism.

2022-07-21 13:17:55 @RachelMiniK @PosenIzzy You will be surprised to know that ALL Diversity officials in the US think that way. I haven't met ONE (I mean ONE) that understands what "diversity" is, namely, that excelling &

2022-07-21 13:04:39 @Meh3625555 @PosenIzzy You are lucky that UK Govt lists "ethnic groups" on an official document, the Equity Diversity and Inclusion (EDI) enterprise in US keeps the list secret, and I know (from insiders) that "Jewish" is not on it. Our effort now is to keep "Jewish" out, but add "Zionist" in.

2022-07-21 04:23:51 Tomorrow morning, 9am Pacific, Steve Zipperstein will be giving a talk on the 100th anniversary of the British Mandate for Palestine (july 22, 1922). Here are the details, in case you're interested in a thorough understanding of the ME conflict https://t.co/gAoEQFfLYX

2022-07-21 02:27:29 Congratulations to @bschoelkopf on a well-deserved recognition and, more importantly, on a breakthrough 2012 paper that has built bridges between causality and machine learning. https://t.co/W1ExAjg33x

2022-07-21 02:10:44 RT @HillelNeuer: Emotional moment today as one of the longest-serving former members of the U.N. Human Rights Council (left) meets one of t…

2022-07-21 02:00:12 Engineers on the brink of extinction threaten entire tech ecosystems https://t.co/UQahFlsVQG via @theregister

2022-07-18 14:14:14 @GinoDelFerraro Progressives do not count wounds, we count choices.

2022-07-18 14:05:46 The ship that opened the world's eyes to British Govt treatment of Holocaust survivors, and swayed public opinion in favor the UN vote (33:13) for partition of Palestine (Nov 29, 1947) despite Arab's promises of "War of Extermination" (Azzam Pasha, Oct. 11, 1947). https://t.co/a3Q235ce4P

2022-07-18 13:50:11 RT @Ostrov_A: #OnThisDay: 1994: Suicide bombing of #AMIA Jewish community center in Argentina (85 dead). 2012: #Burgas (Bulgaria) b…

2022-07-18 13:45:49 @MissDiagnosis @IWF Pres. Roosevelt died in 1945.

2022-07-18 13:39:26 @mishtal @freemonotheist Those who have no history must fake one, even at the cost of being caught.

2022-07-18 08:28:07 @maradonasattva @BillDonovan1234 Judaism without Zionism is like Christianity without a Cross. Not entirely infeasible, yet entirely a-historical and pitifully crippled.

2022-07-18 08:17:20 @POTUS main achievement: allaying fears that leaving #Afghanistan was phase one of leaving #MiddleEast. Saudi "no normalization before X" is actually hopeful, b/c X was phrased as "agreement" on 2-states, not implementation thereof. https://t.co/30pZeDhN2r

2022-07-18 07:53:39 RT @EinatWilf: Read the thread and re-read it in a decade. This is where we are heading (aka what I have been describing as the emergence o…

2022-07-18 07:44:59 @soboleffspaces @IJCAIconf Interesting observation! Both trialists and ML folks try to avoid domain-knowledge, but the motivation is slightly different. Trialists fear "subjectivity bias", while ML fear the labor of elicitation. CI alleviates both fears by offering friendly &

2022-07-18 01:11:03 RT @mishtal: The @BBCNews has a long history of demonising IsraelThere is no doubt that the vile lies and distortion we saw on Thursday –…

2022-07-18 00:54:24 @soboleffspaces @IJCAIconf Who are the algorithmists?

2022-07-17 20:38:18 @BillDonovan1234 @maradonasattva I do not know why you are honoring the @maradonasattva with the clean title "Jew Hater" when he has earned the label "Zionophobe", in the deepest meaning of thisugly disease.

2022-07-17 17:16:47 @Wurmserscribit I know that Biden knows who the oppressor is in the Israel-Palestinian conflict. I assume therefore that he chose this rhetoric as the only common denominator he could find with his audience, who, btw, also knew what he thinks of them.

2022-07-17 17:10:43 Extremely interesting paper: https://t.co/hPrC89titoBTW, what is the current status of the statistical marginal problem? https://t.co/YjAS09CrN1

2022-07-17 14:40:06 That's why we call it "data science" - a two-body enterprise - the data "suggest" and the science "reasons". https://t.co/YeOcFWk7Oj

2022-07-17 14:05:16 @KPMDas Many tell me: "OK, so the NYT has an obsession with Israel but, by and large, it is still "good journalism"". I say: I can only judge them in areas that I've some competency, Israel and AI, and if they are glaringly so anti-professional in the former, I can't trust them anywhere.

2022-07-17 13:50:23 @maradonasattva I love it when eliminationists take off their masks. Abbas forgot, and told us: "We want peace", but that was in English, so he knew everyone in the room understood what he meant.

2022-07-17 13:41:41 @HillCitizen "Saying" and "documenting" quantitatively are too different things. I have some fairly rational friends telling me: "I read NYT and haven't seen anti-Israel bias". Seeing the word statistics may convince them.

2022-07-17 13:37:03 @EvanHoffman Unfortunately, terrorism has been normalized in "freedom fighters" land, so it carries zero weight in public discourse.

2022-07-17 13:33:22 @adamalmo These two papers summarize the main thrust of my talk: https://t.co/dxqdoqMsQm, https://t.co/KBGKcmJxM8

2022-07-16 01:33:12 @jeffreywatumull @DavidDeutschOxf My first reaction: Bacon, Hume and Popper tried to capture the way science works, this paper tries to capture the way machine learning works. I do not see the connection given how far machine learning is from emulating scientific thinking. (e.g., counterfactual thinking)

2022-07-16 00:51:08 RT @IDF: RAW FOOTAGE: This is the moment innocent Israelis had to run for immediate shelter when terrorists in Gaza fired 2 rockets toward…

2022-07-16 00:49:57 RT @MelMitchell1: Interesting take by @AlisonGopnik : Large language models are "cultural technologies" -- they are not themselves intelli…

2022-07-16 00:27:42 RT @TOIAlerts: Live update: Biden announces removal of Red Sea island observer force, advancing steps toward Saudi-Israel ties https://t.co…

2022-07-15 20:54:44 @GadSaad Satirical or not, this outspoken protects the honor and credentials of a tenured UC professor. Hats off, outspoken!!!

2022-07-15 16:27:58 In Bethlehem, Biden pledges to keep up efforts for Israel Palestine peace https://t.co/HeoyQQW4GK

2022-07-15 11:40:07 Try reading some of the graphical causal inference literature and I bet your enthusiasm will increase 10 fold. E.g., https://t.co/PehuRiIg7u, https://t.co/JnPnqezBsGSpecifically, introducing matching and propensity score w/o backdoor admissibility is not advisable. #datascience https://t.co/oXR2Ep5hpS

2022-07-14 23:01:48 @Eve_Barlow Here is the irony. Jews get their strength from history, and their history has culminated in Israel. The poor souls that reject Israel have lost the strength to stand up to their bullies -- a vicious circle. I empathize with their dilemma here: https://t.co/nzitIG6UoU

2022-07-14 20:41:52 @artistexyz @soboleffspaces I tried to follow it, but I couldn't. Do you have agentle tutorial to the method?

2022-07-14 19:04:30 @DKedmey I think it worked

2022-07-14 16:46:51 @WelshBrendan2 I understand. The question is whether a peace-seeker like yourself would be willing to accept the harsh facts as learned by another peace-seeker (myself) who has been searching for a moderate voice the past 75 years -- there isn't any, and it ain't "climate." Hard to swallow!

2022-07-14 16:33:51 @WelshBrendan2 @CotlerWunsh So you, personally, believe Israel "operates under an apartheid model". Am I right? You do not buy the possibility that the "apartheid model" was attached to her out of pure malice?" Am I right?

2022-07-14 16:24:24 @WelshBrendan2 @CotlerWunsh As soon as I read your question: "How can moderates on both sides be empowered?" I was about to offer my help, only to pause and ask: Who? Do I know moderates on both sides?Do you? Can you name ONE Palestinian leader/teacher/cleric/journalist willing to say:"equally legitimate"?

2022-07-14 16:12:52 @soboleffspaces @JohanDH2O @Lizstuartdc @nickchk For me, getting meaningful definitions of NIE and NDE was more important that preserving the additivity property, which I thought to be a disposable feature of linear systems.

2022-07-14 14:56:51 Certain breed of Jews, mainly those who believe that Israel is a colonial apartheid state, get extremely upset when called "Ex-Jewish". Some even call me "anti-Semite" for suggesting their exclusion from the Jewish fold. I'm willing to reconsider and call them:1/2 https://t.co/zZOCgiN0Eu

2022-07-14 13:50:01 Finally, the Jerusalem Declaration signed! The key sentence: "The United States stresses that integral to this pledge is the commitment NEVER TO ALLOW Iran to acquire a nuclear weapon, and that it is prepared to use ALL ELEMENTS OF ITS NATIONAL POWER to ensure that outcome." https://t.co/BQjgxRhlVI

2022-07-14 11:03:51 Confession, I could not hold my tears when President Biden listened to these two ladies and, ignoring all the "time is up" hints from his entourage, including Blinken, continued to listen to what his visit meant to people who survived Auschwitz and have made their home in Israel. https://t.co/S7OqpHTW02

2022-07-14 10:44:48 @tomgoldsteincs This is a legitimate question for every acronym in machine learning. I've asked it on Stanford's "Foundation model" with no answer in sight.The problem is that acronyms are attached to successful programs, not to scientific principles.#datascience, #Bookofwhy

2022-07-14 10:30:39 @hari_hari I am searching for the fundamental ideas, and all I can find are different labels of the same problem, coined by different groups in DARPA .

2022-07-14 10:18:42 I don't know much about pregnancy, borders or obesity, but have had some fun in science, till yesterday, when I read that it's more racist than Trump, w/ all its equations named after white male adults

2022-07-14 09:19:28 RT @soboleffspaces: My lecture "Causal epistemology: Questions" is available as a preprint https://t.co/ig08s1fU3wAs a bonus, you’ll find…

2022-07-14 03:40:02 @shagunsodhani To those versed in causal inference, what you are describing seem to be minute differences of emphasis in several versions of "Data-Fusion" theory https://t.co/E1ap1N4ZCh. I wonder if lifelong folks looked into the solutions offered by Data Fusion.

2022-07-14 02:23:20 President Biden has given Jewish students their first line: "You don't need to be a Jew to be a Zionist."Their next line should be: "Dear Chancellor, You can't bring yourself to state that Zionist students are more than welcome on campus? Aren't you the one who sets norms here?" https://t.co/3pdjCDo2vH

2022-07-14 01:34:49 I've always wondered, but was afraid to ask: How does "lifelong learning" differ from “transfer learning,” or “domain adaptation." In other words, would there be a problem of "lifelong learning" if domains stayed constant? Or,won't LLL be solved if we solved "transfer learning? https://t.co/Ocqg5yIPiO

2022-07-14 01:11:18 RT @MomentMagazine: The Daniel Pearl Investigative Journalism Initiative is in memory of Daniel Pearl, investigative journalist and son to…

2022-07-14 01:10:44 Congratulations to @tgjelten &

2022-07-14 00:31:16 @malickheydar @Harvard Me too. I just gave a copy of #Bookofwhy to my dentist, dedicated with: "Easy on my wisdom tooth, doctor, it ain't dead."

2022-07-14 00:18:11 Thanks for re-twitting this timely post. As more economists enter the world of DAGs, I sense their genuine appreciation of DAGs as a communication language, but insufficient appreciation of DAG's computational power. This post shows https://t.co/6dIyc06BM0that even in 1/2 https://t.co/IwB4Wsyl93

2022-07-13 22:54:43 @RitchieTorres I salute you @RitchieTorres for articulating so clearly what the nature of BDS is, and what your vision is of ME peace, based on "co-existence", a word that no BDS crony has ever uttered. You have my back.

2022-07-13 22:42:37 @mynameisjro @RitchieTorres I salute @RitchieTorres for speaking for our people, because those EX-Jewish New Yorkerswho believe that Israel is an apartheid state, and that BDS is not eliminationist, have abandoned us, in need of a non-Jewish Congressman to counter our predators in the US Congress.

2022-07-13 14:19:14 I once put it this way: "The landscape has become a scripture". Lord Balfour put it even nicer: ".. intimately bound up with a petty territory wedged in between States more powerful far than it could ever be". See https://t.co/9XOCOs8JRl https://t.co/GW5mnWIvUR

2022-07-13 07:50:11 @GadSaad @AyannaPressley I do have the capacity of answering this question, but I'd rather keep mouth shut for fear that "my line of answering" would render me vulnerable to accusations of transphobia (or truthphobia) hence indirectly responsible for suicides among bald white men, a protected minority.

2022-07-12 23:25:44 @adamalmo These two papers summarize the main thrust of my talk: https://t.co/dxqdoqMsQm, https://t.co/KBGKcmJxM8

2022-07-12 23:20:18 You caught me off guard. I was going to jump out of my skin and shout: "Peace! finally!" before I saw the humble #WishfulHeadlines. Back to the real Mahmoud Abbas. https://t.co/eI3FzlwXJh

2022-07-12 23:15:15 This red carpet is longer than River JordanIt's purpose to generate media focus on the real issues of ME, keeping @AJEnglish and other noise makers out of the conversation. My question: Will Biden treat Palestinians as responsible agents, or just reinforce their passiveposture. https://t.co/OUn8cudmUa

2022-07-12 15:57:25 Here are some facts that even Israel's worst enemies know to be true, and which they find to be just nasty obstacles in their heroic mission to prove the opposite. https://t.co/IRulbsSMFO

2022-07-12 14:52:51 @richard_landes A Palestinian never says or "I recognize Israel", it's always: "Arafat did, in 1990", forgetting that Israelis have learned something in the past 30 years, especially to demand a fresh Arabic copy of the PLO charter, w/ the words "right to exist" spelled out for children to read.

2022-07-12 13:56:03 Glad the Academy of Management is about to enter the age of causation, through the very practical and omnipresent question of Good and Bad controls, a question that graph-averse economists have chosen to avoid rather than address. https://t.co/sZPpL7Ifmg, https://t.co/UGqywX2kgF https://t.co/77o5wa9gQl

2022-07-12 13:31:49 This paper, though written in 2014, should be read today in other fields undergoing methodological transformations:https://t.co/f29ZrkXboMespecially those laden with paradoxes or controversies (who isn't?)

2022-07-12 11:09:09 The market is waiting to see how innovation economists circumvent the infinite inertia of mainstream economics. https://t.co/dix3SSHH0s

2022-07-12 10:30:39 Welcome to peace-dodging land! Here is a billboard that would be torn down immediately by our neighbors, yet is put up repeatedly as if we do not know

2022-07-12 08:27:13 How can one tell when Causal Inference (CI) goes mainstream?(1) When universities dedicate buildings as CI-Centers.(2) When recruiters demand CI-qualifications. (3) When US patent office grants patents to CI-ideas.Beware! The latter just occurred: https://t.co/g4CNHEmJpg

2022-07-12 01:45:00 For an academic organization (Association for Israel Studies (AIS)) that counts Zionophobes like @Ian_Lustick among its former presidents, breaking ties with MESA must be a traumatic catharsis. MESA can safely change its name now to IMESA - the Irrelevant side of ME Studies. https://t.co/XTTr0gXJMw

2022-07-11 20:21:54 FYI, this week's harvest of causal inference papers: https://t.co/GSqbg47hAF

2022-07-11 16:35:38 This is a fair comparison. Biden should meet the parents of Malki Roth in Jerusalem. They have been waiting 21 years for justice, and Malki, their slain daughter, was a symbol of goodness and joy. https://t.co/HaY59TXKEn

2022-07-11 16:22:00 Informative narrative of the history of Czech-Israel relations. Thanks https://t.co/aos1rTmMiB

2022-07-11 14:29:21 @DerriereLaLune @rcguo Interesting! First time I see it. But I don't buy the argument: "The derivation does not use an intervention on the mediator, which in many situations would not be possible in reality." Derivational step need not be "possible in reality". Best example: Complex numbers in math.

2022-07-11 14:05:57 This is all true. I'm honored to be speaking again at IJCAI (first time was in 1983 https://t.co/4L78pwd0hc) and to challenge the direction "big data" is taking. No, I'm not flying to Vienna, but will make my Zoom screen as big as possible. See you there. https://t.co/2nsFbjLEGX

2022-07-10 21:12:56 @HaimHabulai Thanks for the correction, I didn't know it.Kol Halomed MeChavero Davar ....Rabbi

2022-07-10 15:41:06 @rcguo Deterministic relationships violate the positivity assumption which is implicit in many identification tasks. Combining X and Y may solve the problem, but be careful, you certainly don't want to ask for P(z|do(x,y)) for any (x,y) pair thatis never observed.

2022-07-10 15:15:58 Israeli president to meet with Czech counterpart in Prague https://t.co/jsbPeBjoKF via @JNS_org

2022-07-10 14:35:55 Jewish father files lawsuit over discrimination at Los Angeles private school https://t.co/WtzS6Q8C68 via @JNS_org

2022-07-10 12:58:43 @elderofziyon @BenWeinthal @ShuratHaDin I'm glad nevertheless that a serious complaint has been filed with the International Criminal Court against the PLO

2022-07-10 12:42:31 Refreshing sight, even to a Californian like me. My mother used to take us (me and my younger brother) to this very beach every week. You see, she was Polish, and believed religiously that sun rays and sea waters is good for your health. Wow she was right! https://t.co/CrrWA7EQFr

2022-07-10 12:28:28 @kareem_carr What about for the kind of data scientist that's more like a data-scientists?

2022-07-10 12:16:31 Most economists use potential outcomes in conjunction with instrumental variables, not realizing that both are derivatives of SCM https://t.co/etASSEk4Or, Some begin to use DAGs but, to make them look home-grown, strip them of their main powers. https://t.co/UGqywX2kgF https://t.co/7LMrdOIdlq

2022-07-10 11:53:48 @blakeflayton We got here (partially) because even Zionist writers use the term "anti-Zionism" as if it was some sort of a legitimate political opinion, instead of using "Zionophobia", which highlights its racist, genocidal and pathological character.

2022-07-10 11:42:23 Our next heroic effort is (again) in economics.My attention is captivated by a BA thesis on the mediation fallacy and the analysis of police violence: https://t.co/09RtTQQ1IM. Considering the graph-averse culture dominating economics, and the risk-averse, sheepish style of1/2

2022-07-10 06:44:36 @EricJGreenberg1 @mishtal @Ixar_Bargains That was easy. I asked him to sing me a Philistine song, or tell me a Philistine story, or name a Philistine hero he would name his son after, or a Philistine holiday he enjoys celebrating, or, ... or.. It was easy.

2022-07-10 05:13:58 @rickwsint @skdh Got it, thanks!

2022-07-10 05:11:34 @rickwsint @skdh I didn't get the better source.

2022-07-09 18:58:46 @mishtal @Ixar_Bargains A Palestinian colleague once argued that Palestine was there first, because Philistines are mentioned in Egyptian records, not Israel. He dropped this argument once I proved to him that his connection to Philistines is weaker than my grand-mother connection to the grand-canyon.

2022-07-09 15:42:05 It's called science-centric paradigm: https://t.co/dxqdoqMsQm https://t.co/IbabQOhFLb

2022-07-09 15:12:29 @Ostrov_A @Israel @TelAviv @yiftahc @GeorgeDeek @AvivaKlompas @EmmanuelNahshon @GalGadot @ShaharAzani @IdoDaniel @IsraelinUSA @TravelLeisure The one on the bottom left reminds me of Shuck Hapishpeshim (the bed-bugs market) where I bought my first acoustic guitar, in 1976. Glad the place is still striving.

2022-07-09 14:57:50 @roydanroy "Swallow", of course, by claim of ownership (called "inductive bias"), but not before ML folks learn to articulate what it is they claim to own.

2022-07-09 14:34:54 @rasbt Jobless - not yet, but outdated, on the mark.

2022-07-09 14:31:21 Surprise! Ali Khamenei Slams Western Values:https://t.co/ILPneZcSwjA perfect candidate for Dean of Social Science at one of our finest US Universities. The salary and faculty enthusiasm would surely match his current employment benefits.

2022-07-09 13:40:55 When ‘Racial Justice’ Means Antisemitism: A Takeover of the Yale Postdoctoral Association https://t.co/yffKwWIyKm

2022-07-09 13:07:12 @thinkyparts "Reasoning" entails decision making under murky waters.

2022-07-09 13:04:37 This paper by Grosz etal, "The Taboo Against Explicit Causal Inference in NonExperimental Psychology," https://t.co/A4W2dSeHm0explains much of the research environment in that field, and why I consider the works of @dingding_peng &

2022-07-09 11:36:32 Every country on the right side of reason is a fortress on the right side of humanity https://t.co/fF8ul3GGre

2022-07-09 11:04:36 @blakeflayton You are giving readers the impression that Zionism just "does not say they dont have a right". I truth, Zionism explicitly says they DO have this right "even if it compromises Zionism". Why not quote the sources, Ben Gurion, 1931 https://t.co/PxDjqnprDp

2022-07-09 10:51:49 @eIumax Problematic is the mind that sees the sun, the moon, and living organisms as "problematic"

2022-07-09 10:35:48 RT @philipcball: The great thing about these images of how the brain works is that they show us this is not the way to understand how the b…

2022-07-09 10:18:16 @blakeflayton Why? Has anyone mistaken it with the civil rights movement of another nation? French? Egypt?

2022-07-08 21:24:22 RT @JamesGrace111: @yudapearl Judea, it is available online (except for the references). https://t.co/TtgddR05Bb

2022-07-08 16:21:40 @JohanDH2O Excuse my ignorance, but what ideology does Jordan Peterson represent??

2022-07-08 16:16:22 RT @thomaskipf: Happy to announce that there will be a workshop on Neuro Causal &

2022-07-08 16:08:09 @tkaiser_science Very encouraging, but what is the NAME of this journal? We need to publicize it and commend the editors on being forward looking. (unlike some fields I won't mention).

2022-07-08 15:53:32 Our next hero welcome goes to an economist (who would ever imagine). Nick Huntington-Klein has a new book which, according to its TOC,https://t.co/zrAIFwJedkpromises to deliver economics students from the graph-less torture that their professors went through, and which1/2

2022-07-08 14:23:01 @MiriamElman @GeraldNGOM @AndrewPessin @UCLA Miriam, I need no FYI's. Here is what I posted two days ago: https://t.co/PUTQuiHLouand, you are right, Joe was on the UCLA committee that fought the Zionophobic racism of the Asian-American Studies department, and testified before the Regents. I know we will prevail.

2022-07-08 13:00:16 The next heroic effort is in psychology.https://t.co/Q4HFdV6E7aBased on a 2018 paper by Julia Rohrer @dingding_peng https://t.co/DZrGC4hlQEMy first reaction was: Who is this community of Psychologists who were not convinced by Julia's article? It's hard to believe1/2

2022-07-08 12:10:38 @IllyseusDasein @SamHarrisOrg @jordanbpeterson @BretWeinstein I do not understand what you are trying to insinuate, but I've known Joe Manson for 20 years. He ain't an oppressor, nor a "self-styled victim", nor any of the adjectives you associate with people you don't like. Please do not use this smear technique on this platform.

2022-07-08 11:48:55 fail to mention that many (if not all) of the diverse types of causes they discuss are now unified through SCM. I never understood those who prefer pluralism on monotheism. See https://t.co/WsIfsMFM4T 2/2

2022-07-08 11:48:54 I would like to dedicate the next few posts to heroic readers who are struggling to introduce causal clarity to fields where such clarity has been lacking.I will start with https://t.co/pJScBlqjcMwhich focus on neuroscience.My only gripe: they push me into "statistics"and1/2

2022-07-08 11:02:55 RT @kareem_carr: when you contribute to the paper but don’t get authorship https://t.co/qlfW3WbfXv

2022-07-08 10:59:35 This film taught me quire a few things I didn't know. But most of all, I liked the dedication to feminists like Linda Sarsour and other fake leaders of the Women March, who continue to peddle their dishonesty to the young and unsuspecting. @ZionessMovement @bariweiss https://t.co/fVQNn8l4u8

2022-07-08 10:13:47 RT @MilsteinFF: RIP to one of the leading movie stars of the 70s, James Caan, who passed away today. He is known for playing Sonny Corleone…

2022-07-08 10:10:14 @SAFECUNY Great! As the first step. The next steps are:1. CUNY's administrators learn to spell "Zionism"2. They learn to pronounce it, loud, w/o watching if anyone is listening.3. They tell the campus what it contributes to diversity and education.

2022-07-08 09:22:41 Populist slogans are not supposed to mirror facts or figures

2022-07-08 06:30:20 The next take belongs to the @latimes, whom Californians trust to spot such stories and turn them into a public outcry: "Are we losing our universities?" They used to listen to me, before the woke takeover. Perhaps they would listen to @bariwies. https://t.co/7nsEavTEEh

2022-07-08 06:05:41 @KabirCreates @ProfLaurenRoss @KordingLab @shansiddiqi The beauty of Hume's counterfactual definition of "cause" is that, for 2 centuries, people thought it's final, and no one asked: "So how do you define counterfactuals?". Why? Because counterfactuals are so entrenched in our mind, we thought they're irreducible primitives.Wrong!

2022-07-07 23:23:02 @shansiddiqi @ProfLaurenRoss @KordingLab If by "objective" you mean "based on data alone" we know it is impossible. If you mean "based on both experimental and non-experimental data" you are guilty of circularity, because you need it to "design effective intervention". The philosophers I read are guilty of the same.

2022-07-07 07:15:41 I can't dive too well, but if anyone is swimming already in the Indian Ocean, I would be interested to examine the earliest counterfactual expression in ancient Indian literature. Similar to the one spoken by Abraham: "What if there were 50 righteous people in the wicked city?" https://t.co/Wf02Ltprwg

2022-07-07 02:34:46 A confession of a repentant Zionophobe. https://t.co/MoplqlGnLS

2022-07-07 00:45:59 The New Zionist Congress - 2022@blakeflayton @newzionists https://t.co/VK5gGoDUKE

2022-07-07 00:05:29 Photo of the leaders of Palestine and Hamas, very symbolic, but this will not improve the situation https://t.co/2fBT8oTa2f via @ePrimefeed

2022-07-06 15:23:57 An important page of history. The event that led to Israel's "defensive shield" campaign, which some regard (including me) as the only effective"war on terror" by any country, though still continuing. https://t.co/R4ChcOyZn5

2022-07-06 14:36:29 @EpiEllie @KordingLab @ranilillanjum @SDMumford Beautiful introduction to a Short Introduction, and very useful, for every CI scholar, to collect this arsenal of examples to communicate nuances and variants of causal thoughts. Thanks.

2022-07-06 10:26:25 @MechncOfMeaning I'm only concerned about the last line. do(c) always appears under a probability expression.e.g., P(r|do(c)) = P(r)

2022-07-06 10:19:37 @intrinsic_motiv @ylecun What do you expect to get by implementing this strategy? I am asking, because a theory is now available that can tell you what the strategy can and cannot do. I would recommend that use the theory to guide your implementation. e.g., https://t.co/KBGKcmJxM8, #Bookofwhy

2022-07-06 09:45:28 @JGreenblattADL @almog_doron @JewishAgency Congratulations, Almog, on assuming this leadership. You can count on our help.

2022-07-06 09:31:50 RT @blakeflayton: Putin: We are liberating Ukraine from Nazis Also Putin: We are banning the Jewish Agency from Russia

2022-07-06 07:31:43 Yes, you can get the Solution Manual if you write to my assistant kaoru@cs.ucla.edu https://t.co/zupr0YOnku

2022-07-06 07:03:48 Can you help me climb this paywall? It tells me: "Access to this article via University of California Los Angeles is not available." Why? I think I've lived a clean life! https://t.co/j61ckJNypS

2022-07-06 06:52:52 Agree full-heartedly. Unfortunately ML folks are busy building Gothic Cathedrals for a handcuffed God, funding them, and convincing investors that Gods should know how to liberate themselves. https://t.co/dkscaksQMM

2022-07-06 06:37:49 I would recommend Primer https://t.co/XofByaPBqG which is available free (chapter by chapter) and logically coherent, starting with the First Law and ending with Tool Kits for Mediation and Attribution. https://t.co/dXuwCFs6Oj

2022-07-06 06:30:25 @giulioprisco @SMcfarnell @WiringTheBrain @DavidDeutschOxf @mpigliucci @BobbyAzarian @erikphoel @PsychToday @Sara_Imari @philipcball @danieldennett @skdh @SamHarrisOrg @bgreene Interesting conjecture which has skipped my #FreeWill optics. I am inclined to surmise that the philosophical debate is orthogonal to the political, with the latter misunderstanding the former.

2022-07-06 06:23:24 @intrinsic_motiv @ylecun Before a system can acquire a blueprint of itself it ought to know how to acquire and use a blueprint of its environment, also called "world model". I am not sure DL is ready for the former, having skipped the latter.

2022-07-06 06:14:03 I was right then in predicting that Imbens writings and prizes will entice people to study causal inference the scientific way https://t.co/sZPpL7Ifmg,starting with the First Law: https://t.co/etASSEk4Or@causalinf #Bookofwhy @VC31415 https://t.co/oszgFk1or7

2022-07-06 05:46:41 @animesh1977 Viewed as a threat, yes, but I don't recall the "clowns" era.

2022-07-06 05:41:08 For readers and teachers of Causality (2009), viewgraphs and exercises are available here:https://t.co/xTetlHllUhAlternatively, Primer is loaded with exercises: https://t.co/XofByaPBqG, &

2022-07-06 05:17:53 @texan_marshall I suspect if such a system turns successful, it would spawn a movement to "defund AI," which no AI researcher wishes to see happening.

2022-07-05 23:47:13 RT @AvivaKlompas: There are no words. Both parents of this lost toddler from the #Highland Park mass shooting - Irina Levberg and Kevin McC…

2022-07-05 23:42:24 When an activist quotes @AlJazeera and says: "It truly appears to be XYZ" you can bet on two things: (1) He knows nothing about what happened, and (2) what happened is definitely not XYZ. https://t.co/uISXhJ2rvi

2022-07-05 23:09:03 Defeat Overture INT o2 PCUSA - Sign the Petition! https://t.co/JFnxga3d9r via @Change

2022-07-05 22:13:04 @equilibriumcow Yes, this is the page. I'll tell readers to go there.A more gentile technical introduction is Primerhttps://t.co/XofByaPBqGIt has exercises and solutions.

2022-07-05 18:40:45 @equilibriumcow ill post it, thanks

2022-07-05 12:57:21 @equilibriumcow Thanks for the comments, but which "causality" do you have in mind, and which website are you referring to?

2022-07-05 11:02:59 I'll second your recommendation. I truly have no idea how many ML folks read #Bookofwhy. There are days when I think: "zero". And then, when I see them struggling to emulate causal thoughts, I think: "99%". Hard to tell. https://t.co/uvgzAol626

2022-07-05 08:11:59 To complete my story about the Kielce Pogromhttps://t.co/69HoNUvsEdbelow is a picture of my aunt, Mania (an Auschwitz survivor), during the funeral, consoling families of the victims.(Courtesy of her son, now living in Israel, a proud graduate of my own college, The Technion.) https://t.co/r60wxeRpnm

2022-07-05 04:19:46 @Wetassprior @KeurigBad No, Z is a son of W, so, conditioning on Z opens the backdoor path X<

2022-07-05 04:12:56 Dan Rather: "The U.S.A. is and always has been an idea more than a place" "https://t.co/c0EeQkgnoIMade me wonder: How many societies think like that about their countries? I know Israelis do.But I don't get it from my Canadian granddaughters. Switzerland? Egypt? Anyone knows?

2022-07-04 21:07:18 I've recently took the liberty of calling it "Ladder of Knowledge" rather than "Ladder of causation"

2022-07-04 20:43:58 On July 4th we realize that the words "defending freedom and democracy" are more than a broken record. It hits you that certain freedoms and certain democracies do need defending. https://t.co/ZBnQHrWQLf

2022-07-04 20:23:59 This hits close to home. My mother's family is from Kielce. And her sister witnessed the pogrom in her own eyes, miraculously survived, and later eulogized the victims in the cemetery, as was recorded on movie camera. Only one year after liberation. https://t.co/naoS0xpoCM https://t.co/kN6pwUnTFj

2022-07-04 18:12:35 What a coincidence for July 4th holiday.An important addendum, for American Jews to remember. In 1997, Herzl said:“Zionism is a homecoming to the Jewish fold even before it becomes a homecoming to the Jewish land.” We can see it being incarnated today, in the US. https://t.co/p9nbVPo6Jt

2022-07-04 17:30:38 I'm still surprised that no one in the media raises the question of motive: Why would Israel aim to kill a journalist when she has everything to lose and nothing to gain from such act? https://t.co/fzdJuvHoz1

2022-07-04 16:27:10 Happy 4th of July!On this day, it is meaningful and relevant toread aloud Longfellow's poem:"Thou, too, sail on, O Ship of State!Sail on, O Union, strong and great!Humanity with all its fears,With all the hopes of future years,Is hanging breathless on thy fate!"How true!

2022-07-04 07:44:50 @arbutuspointBC You can tell from his logic that he ain't likely to bea fan of Israel or of other islands of commonsense. @EnglerYves

2022-07-04 07:34:51 I have elaborated on these Eight Myths in "The Causal Foundations of Structural Equation Modeling", an updated Chapter for the upcoming R. H. Hoyle (Ed.), Handbook of Structural Equation Modeling. New York: Guilford Press, https://t.co/veOJoVDode https://t.co/xjakfS3d1E

2022-07-04 07:22:36 Toronto logic at its glory: "Irrespective of the veracity of accusations, rallying for the accuser is reasonable!" Imagine rallying against Ukrainian borscht at TO store called “Taste of Ukraine” because Putin called Zelensky Naziand deemed Ukraine "stolen land." 1/2 https://t.co/hgZChELafo

2022-07-04 06:45:23 @arbutuspointBC Toronto logic at its glory: "Irrespective of the veracity of accusations, rallying for the accuser is reasonable!" Imagine rallying against Ukrainian borscht at TO store called “Taste of Ukraine” because Putin called Zelensky Naziand deemed Ukraine stolen land. Grand-Logic!

2022-07-03 23:10:30 @robertclab I doubt people who shun DAGs even for simple problems would fall in love with SWIGswhen problems get harder.

2022-07-03 14:42:16 @anon61845948 @jrootham @MelissaLantsman Beg to differ. An ethnic group that equates its liberation, loudly and unequivably, with the elimination of another ethnic group is not equally deserving of support and respect. Please listen to how THEY define liberation, not how Westerners wish them to define it.

2022-07-03 08:37:24 PM Lapid spoke for me when he said:" It was here before us, and be here long after us". I will vote for him, if I'm so lucky as to be in Israel November 1st. https://t.co/1EmD8sJqzq

2022-07-02 19:20:52 Every time a reader retweets this interview I am tempted to re-listen to it and then wonder how the interviewer, Guido Imbens, could possibly continue to work with Potential Outcomes instead of switching to graphs. An resolved mystery of scientific evolution. https://t.co/4bxgt9jw9N

2022-07-02 18:41:09 Israel is the only country whose future depends on its neighbors textbooks more than it depends on their guns. https://t.co/cQEDbIdgcK

2022-07-02 14:42:17 @hdjnicklin Your logic is right, but missing one fact: Palestinian riots of 1936-1939 forced the British Government to stop immigration from inferno Europe, thus sealing the fate of 6 million Jews who sought refuge in Palestine, the only country where they were welcome as brethren.

2022-07-02 14:30:26 @dizzdangles @MelissaLantsman You got a point, if it weren't for the death-threat encoded by the "Free Palestine" slogan, these "protester were shouting.

2022-07-02 14:17:51 @jrootham @MelissaLantsman Not sure if you seriously do not know what "Free Palestine" means to those who shout it.

2022-07-02 14:11:16 @MelissaLantsman, you are talking as if anti-Zionism is an excuse for this form of protest. Is it? Suppose it was purely anti-Zionist, soaked with angelic love for non-Zionist Jews. Would it be less racist? Less morally deranged? Less despicable? https://t.co/UiThywux29

2022-07-02 13:46:18 COVID-19 hero: Pfizer CEO Albert Bourla reflects on the past, future https://t.co/wMaQwtfDkn

2022-07-02 07:53:08 Only a Zionophobe like @KenRoth would retweet this misleading statement. It was not the Presbyterian Church that voted, but its lopsided Committee on International Engagement. The overture will be voted on by the entire membership next week. Hoping Presbyterians defeat INT 02. https://t.co/EGkR5YlO6S

2022-07-02 07:29:01 RT @ADL: We condemn the PCUSA’s International Engagement Committee passage of a resolution that falsely labels Israel as an apartheid state…

2022-07-02 07:25:27 I wish left extremist would be as generous and accuse us of things we are proud of. https://t.co/JsO64G1JUt

2022-07-02 07:15:43 Glad the conversation at CUNY is converging toward the real issue: Welcoming Zionist students as a valuable minority group. Sadly, CUNY adm. has not passed the litmus test for open-mindedness: Spelling the word Zionism.

2022-07-02 07:15:42 Jewish Groups Criticize CUNY Chancellor for Not Appearing at NYC City Council Antisemitism Hearing https://t.co/3mgecpqeVs via @jewishjournal

2022-07-01 20:18:33 @stephensenn By "asymptotic" I mean "as the number of samples in each center increases indefinitely".Isn't ACE an asymptotic concept? (assuming the 'A' in ACE stands for 'average', not sample 'mean.'

2022-07-01 20:09:41 I welcome the emergence of the new Jewish Studies Zionist Networks. From now on, anti-Zionist professors will not be able to say "We represent THE Association for Jewish Studies" without specifying WHICH ONE

2022-07-01 19:31:22 @kareem_carr Great! But didn't they rule last year that machine learning is statistics on steroids?

2022-07-01 18:10:15 @stephensenn Will try to digest - give me a couple days (or weeks??). Question: Do the two designs give same asymptotic estimate of ATE?

2022-07-01 09:16:03 @elderofziyon We call it "The Tectonic shift of John Kerry."

2022-07-01 08:06:12 Those who equate conspiratorial theorizing with drug addiction understand why this poor fellow feels so "high" about discovering Jewish black magic and Israel's Book of Solomon. Others get high on "apartheid" and "occupation" - popular modern "fixes" to anxiety and ignorance. https://t.co/YgP2JSi4Xe

2022-06-30 23:51:35 As a holder of a dual citizenship, Israeli+American, I am proud to introduce readers to the new Prime Minister of Israel, Yair Lapid, a man of vision and integrity, who just took office today, and is being congratulated by my President Joe Biden. To their partnership!- LeChayim! https://t.co/VYcJMHVEbh

2022-06-30 23:22:29 @AvrahamAdler @mishtal @benandjerry The malice comes from Anuradha Mittal. But @benandjerry remind me of so many Jewish American uncles, who just want to please everyone (including Mittal) as long as you dont ask them to study more carefully what's going on in ME, or what the Palestinians say about Israel's future.

2022-06-30 22:35:26 @mishtal Again, you are too harsh on @benandjerry who always say "we believe in the right of Israel to exist" and just happened to forget that "existence" in certain neighborhoods necessitates some measures of defense, including military presence. No malice intended, just forgetfulness.

2022-06-30 20:44:10 @mishtal You are too harsh on innocent boycotters like @benandjerrys who simply wish to signal virtue,peace, love and ice-cream and, not knowing any better, picked up the most populist slogan on the internet, "occupation of Palestinian land", and made it their banner of thoughtlessness.

2022-06-30 12:40:38 @LekhtNaya And I want to be an "oppressor" - teach me!

2022-06-30 09:16:24 @PMinervini @kerstingAIML I would ask it more provocatively: I have a large model that someone found on the pavement outside. We know nothing about it except that it is "large", in fact huge, humongous. Is it a "Foundation Model" in Stanford sense"? In anyone's sense?

2022-06-30 08:54:17 @PMinervini @kerstingAIML All we know thus far about Foundation Models is that they are "Large". Has anyone found any other quality to FM's but "Largeness"? BTW, if I was a reviewer and thought the paper is from Stanford HAI, my first impulse would be "reject" -- tired of "largeness".

2022-06-30 08:26:40 @kerstingAIML A calculus of intuitive causation would be great to have. Any aspirations as to its eventual structure? Here is an intuitive causal inference: If a drug helps men and helps women it must help people.Can "Foundation Models" reach this conclusion?

2022-06-30 08:14:40 @kerstingAIML What is LLM? Couldn't find it in your paper, nor in Wikipedia.

2022-06-30 07:58:09 Orwellian. But I don't believe President Pollack, a no-nonsense lady, would have approved of such stupidity. It tells more of how Wokism works: Some looney complains, a clerk gets scared of being accused of insensitivity, another hurries to his protection -- soon Lincoln is gone. https://t.co/32q3voLzWg

2022-06-30 06:39:12 My list of recent articles on causal inference:https://t.co/IQ7UFAU8jyOne paper catching my attention is "Can Foundation Models Talk Causality" https://t.co/GXfTbqqMKVa topic discussed on our platform which made me wonder: How can one ask "Can X do Y?" when X is undefined?

2022-06-30 05:56:14 @MESA_1966 decision to boycott #Israel academic institutions amounts to stripping the Middle East from the one country that has ushered the region to modernity-- in art, science, business and democracy -- thus committing the region to another century of hopelessness. https://t.co/IfdrUKOWGx

2022-06-29 23:10:14 This betrayal thrives on disinformation: Donors to "Israel Studies Centers" are kept ignorant of what leaders of those Centers teach or advocate. I therefore link here to the original statement, https://t.co/gPJSVcBuID to help current &

2022-06-29 20:00:55 An amazing building! They mushroom on so many campuses. Naughty me, I once called them "Gothic Cathedrals for a handcuffed God." I still do, while hoping that the ratio of "science" to "data" would justify the hype. https://t.co/zwdqhcdYIy

2022-06-28 16:00:24 @StaceyEBurke @Vandalay_Inc @ShMMor @AIPAC This may be the reason why JStreet, as well as @NewIsraelFund , spend their time attacking other Jewish organizations - impressing their donors with some action and uniqueness. I once wrote an oped about the Jews of Discomfort: https://t.co/lz8kD9bPxUStill relevant today.

2022-06-28 15:12:25 @ShMMor @Vandalay_Inc Both @AIPAC and @Jstreetdotorg are honorably listed on the Boston Mapping Project, yet it is the latter who finds the time to fight the former, as if there haven't been any Roman legions approaching the walls of Jerusalem.

2022-06-28 14:37:41 Samantha, those who harassed you expect you to beg for protection. Try to address them as "Zionophobic racists" and see how they lose balance. Example: https://t.co/7jn0pnLspP.Zionophobes are shocked when their ideology is examined as a moral deformity, by facts &

2022-06-28 13:57:41 @AleksanderMolak The closest I got to dynamic systems was in a paper with James Robins ". "Probabilistic Evaluation of Sequential Plans" https://t.co/WxOYYrNECswhich computes the effect of time varying actions, in discrete times, each depending on the outcomes of its predecessors.

2022-06-28 11:33:33 Thoughtful analysis of the Boston Mapping Project, especially on why Jewish groups have a hard time naming the ideological roots of progressive antisemitism. https://t.co/GwlHdZb6rZ

2022-06-28 08:30:48 @GadSaad @KamalaHarris Finally, someone surpassing @AOC understanding of the Middle East.

2022-06-28 08:05:54 @stephensenn @snoble It's not a general rule. It's specific to this particular problem which we have cycled dozens of times, with zero progress, despite my honest attempts to penetrate the verbal curtains of RCT's discourse. It should also be helpful for trialists to see their words cast in 2 Eqs.

2022-06-28 07:53:39 @Martin_Kramer @MESA_1966 @DovWaxman Hard to believe my letter pushed UCLA to rescind its membership. I assume there were stronger oppressive and colonial forces behind the process, and no trumpets, to avoid backlash. That should only encourage Israel Studies Centers at other universities to try.

2022-06-28 07:14:05 Illuminating article and fascinating demonstration https://t.co/Ho37uyW33g

2022-06-28 06:48:50 @Martin_Kramer @MESA_1966 That's not enough. Each should write to demand that their university revokes its affiliation with MESA. I wrote to UCLA Chancellor, but it didn't go very far, considering that I couldn't say: "MESA activities undermine the legitimacy of the Center I was hired to run." @DovWaxman

2022-06-28 06:09:39 @stephensenn @snoble I am curious to see TWO different kind of RCTs that cannot be distinguished by SCM and that can be distinguished in some mathematicallanguage, using 2 x 280 characters if possible, but no links to other papers.

2022-06-27 18:50:00 I became friendly with the Presbyterian Churchwhen they asked me to go to Alabama to debate BDS cronies. Today they sent me an interesting video which I would like to share here. It's an interview with Palestinian HR activist Bassem Eid: https://t.co/jnHgYYkyeJWorthy watching.

2022-06-27 17:18:10 The next step is direct flight, Mississippi - TelAviv,paid for by confiscated Iranian bank accounts. https://t.co/67EU4P2ez0

2022-06-27 17:04:06 And after the awards comes the yield: List of last week's articles on @causalinference:https://t.co/QtBgZvR1EpThe one paper that caught my eyes was the thesis by Nick Pawlowski on Probabilistic and Causal Inference in Imaging https://t.co/O47fdIjWQg https://t.co/HP11UtPe9Y

2022-06-27 16:17:51 It has been good year for @causalinference. Congratulations to Laureates James Robins, @_MiguelHernan, Thomas Richardson, @RotnitzkyAndrea and Eric Tchetgen Tchetgen for the King Baudouin Award.All that remains to do now is to agree on what @causalinference is about!!! https://t.co/qTWFe80hUa

2022-06-27 15:33:31 Sorry to see @sharon_nazarian leaving @ADL at this crucial time, and wishing her good luck and much success and satisfaction in her next endeavor. https://t.co/N2D2wiUWHA

2022-06-27 15:07:13 The FDA was formed in 1906. I wonder when was it that they allow counterfactual vocabulary to enter, perhaps even dominate their official documents? https://t.co/lDcd5DgYnF

2022-06-27 10:20:34 UN: Look at what you have done with a truly noble idea! How could you let it slide to such lows? https://t.co/wpyNdpLfLN

2022-06-27 08:51:53 RT @StandWithUs: Today is #Holocaust Survivor Day. Perhaps you know a Shoah survivor or have heard one's story. We invite you to keep this…

2022-06-27 08:45:41 Speaking of Human-Centered AI, I can't think of a more human-centered project than this. https://t.co/dszKNQLC20

2022-06-24 23:32:52 RT @SethAMandel: I want to be shocked that the BDS Mapping Project--something out of mid-20th century Europe

2022-06-24 23:26:43 @GnomeSortingHat I would march it to its obvious conclusion:"Breathing has no effect on death"And, to sound statistically savvy, I would add:"On the average".

2022-06-24 14:44:48 A significant by rarely mentioned fact: The first book to be printed in Western Asia was printed in the city of Tsfat (upper Galilee), in Hebrew, in 1577, by the Kabalistic community that flourished there following the expulsion from Spain (1492). https://t.co/FkL8H9ko7E

2022-06-24 14:16:12 the rationale for opposing boycott (putting aside the whimsical and baseless nature of the allegations raised) should rest on the fact that business licenses are granted by society with the understanding that equitable services will be provided by the grantee, in according 2/3

2022-06-24 14:16:11 This is a major legal victory against BDS, agree, but now we need to move to the moral dimension and explain to people concerned with First Amendment why boycotting Israel differs from maligning Israel, or even lying about Israel, both are protected speech. I believe the 1/3 https://t.co/hAUgck7kvy

2022-06-24 12:54:49 This reminds me of the possibility of demonstrating "Reverse Regression" and the argument whether we should compare salaries of equally qualified men and women, or, rather,compare qualifications of equally paid men and women. See https://t.co/xXl4OMhduq https://t.co/XVR6DPd6UT

2022-06-24 12:16:21 @le_lummy Sure, but I couldn't cite all Biblical/Mishnaic cities Tzfat deserves a special mention, I agree. Why? Because the first book printed in Western Asia was in Tzfat, in Hebrew, in 1577. Thus, it was the Hebrew language that ushered the ME to modernity, not Greek, Latin or Arabic.

2022-06-24 10:17:10 @johnpaulpagano @blakeflayton We can mock Lady @codepink, but she commands over 100k Twitter followers, each one of whom shares her mentality and can decide to follow the "mapping project" and undo some "Zionists &

2022-06-24 09:59:27 Investing in UNRWA means investing in self-sustaining bureaucrats whose jobs, relevance and mission is to perpetuate the conflict, and providing them with the means to do it: Education to Passivity and Hate. https://t.co/gG9niTDBlJ

2022-06-24 09:40:41 Good morning, Tel Aviv!The first Hebrew speaking town in the entire universe (barring ancient Jerusalem, Beer Sheva and, of course, my own home town of Bnei-Brak) https://t.co/EQtvbgo2hA

2022-06-24 08:52:00 @grbradsk @IfNotNowOrg I think it is more of a "battered wife" syndrome in the case of @IfNotNowOrg and @NewIsraelFund

2022-06-24 08:30:22 @soboleffspaces @MatthewFe94 Good paper, agree. And just in case there are changes, here is the revised, published version of this paper https://t.co/gvwnPKDGKH

2022-06-24 07:59:28 I could accept that kind of "controlling" behavior from reviewers if they themselves felt more comfortable with Imbens-Rubin notation. But they don't, b/c no mortal does. So what could be their motivation, aside from the cultish preservation of the Imbens-Rubin torture?Gee,1/3 https://t.co/nz10zYdCFP

2022-06-24 06:55:25 @MairavZ The days of such accusations are over. Today we label "Zionophobes" those who rush to blame Israel with hearsay evidence, and who refuse to grasp that Israel has everything to lose and nothing to gain from targeting journalists. @pressfreedom

2022-06-24 06:30:01 What an honest confession of the confused, misinformed, rootless and spineless culture that spawns converts to @IfNotNowOrg . What an indictment of Jewish education in America. https://t.co/7OCzxhDXOe

2022-06-24 06:05:50 RT @emirates: Touchdown. We've officially landed in our newest destination, Tel Aviv. https://t.co/2ZgEnJVdVV

2022-06-24 05:39:01 @soboleffspaces Where do you find statisticians "controlling"?Those I know are exhausted trying to explain to their students what "unconfoundedness" is - an insurmountable, super-human task according to Imbens and Rubin (Cambridge, 2015)

2022-06-23 15:01:31 @EinatWilf Trial balloon, yes. But, realistically, I can't see Israelis going through the civil strife of uprooting X number of settlements without a more meaningful ideological commitment to "end of claims" on the part of the Palestinians.

2022-06-23 14:26:47 @TheAcsMan @BDSBoston @BDSmovement The same can be said about some Jewish organizations who discovered BDS only when their names appeared on the "Boston Mapping Project." Two decades of BDS delegitimization of Israel only made them yawn

2022-06-23 14:16:17 @DKedmey Both Wingate and Patterson are inspirational! Perhaps my poem will inspire someone in the 22nd Century to salvage the remanences of our great civilization and rebuild the walls of Rome.

2022-06-23 14:05:03 @neuroviana @FundacionBBVA I've nothing but praises for this book - enjoy!

2022-06-23 12:11:06 RT @HenMazzig: Israel’s commitment to empower and protect #LGBTQ people — including Palestinians makes me proud.LGBTQ Palestinians f…

2022-06-23 12:09:00 RT @EVKontorovich: BREAKING: Constitutionality of state anti-BDS laws vigorously affirmed in first full federal Ct of Appeals decision on t…

2022-06-23 08:01:07 Now mother @BDSmovement is telling them: Drop it. You know that we count on the world not to take seriously what we say. You spoiled it. Americans are too sensitive to "mappings", so pretend the map is only for ordering Kosher food, and go back to advocate "justice".2/2

2022-06-23 08:01:06 Poor @BDSBoston, they simply tried to explicate what mother @BDSmovement has been telling them for years

2022-06-23 06:38:44 @patrickmesana @ylecun @GaryMarcus I would like to believe that that mysterious life force is called "the scientific method".

2022-06-23 06:24:28 Begging to differ. If the only reason Israel should take existential risks is the moral burden of "55 years of occupation is unjustifiable," then I would venture to suggest that 74 years of denying a neighbor's right to exist is many many times more unjustifiable. https://t.co/NsgdDOpGBR

2022-06-23 05:50:33 Derek Penslar, unsurprisingly, gets it totally wrong, as would any author of the Van Leer "definition" of antisemitism. What's the point of wasting energy on scholastic distinctions between "anti-Zionism" and "antisemitism" when the real issue (which the Van Leer "scholars" 1/2 https://t.co/eZx1ZXQO3s

2022-06-23 03:44:33 Thanks for the reminder. I've lost my poetic muse lately

2022-06-23 03:34:39 @slowhoneybee @blakeflayton to replace the term "anti-Israel", which allowed them to ignore the anti-coexistence aims of the so called "pro-Palestinian activists", eg, the nice guys from @JVPBoston and @IfNotNowOrg, who are not antisemitic, heaven forbids, just Kosher eliminationists.2/2

2022-06-23 03:22:11 @slowhoneybee @blakeflayton You are analyzing the Forward article as if it was a piece of news when, in effect, it is written to drum up sympathy for the poor semi-Jews of Boston who are being forced to examine their identity, something they haven't done since they coined the term "pro-Palestinian" 1/2

2022-06-22 16:10:21 @blakeflayton The laws of arithmetic are a Zionist invention, created to justify Jewish aggression. Palestinian science has different rules. It predicts about 100 million refugees in 2030, and says that the longer one refuses to recognize its neighbors the shorter the conflict would last.

2022-06-22 15:39:20 @kareem_carr Apropos, Why I am only a Half-Bayesian https://t.co/gqeQb2lfLJ

2022-06-22 15:10:25 @AndrewPGrieve Congratulations! For my curiosity, can you share a simple problem the Hybrid method helps solve, which a half-Bayesian like me would find difficult to solve?

2022-06-22 15:04:39 @SAFECUNY @ChancellorCUNY I don't get it. Your Chancellor declares anti-Zionism a legitimate form of campus hate and you applaud him?

2022-06-22 14:54:14 The Boston Map is just another "pound of flesh" anti-Israel "progressives" are requested to give for social acceptance, awakening some of them to notice that there is no more flesh to give -- dry bones is all that's left. Maps have virtues after all - happy awakening. https://t.co/iwNijbrhsD

2022-06-21 12:09:12 I've just returned from Bilbao, Spain, where they staged quite a show for the Laurates of the BBVA Frontiers of Knowledge Award. See https://t.co/93GgdfyxCB, and https://t.co/uL2iSJQTK2I will post a transcript of my acceptance speech later.

2022-06-21 11:48:53 Please note, all Jew-haters deny that they are "Jew-haters" (even Linda Sarsour), while no Zionopnobe has ever denied having this uglier variant of the virus, thus making it easier to detect and irradicate. https://t.co/LGdz1yDFII

2022-06-21 10:03:36 Rapists always argue that a woman's short skirt is responsible for her rape.Our panelists at Van Leer Institute are now claiming that its not her skirt, but her very body. They're right: If Jews were not what they are at heart- lovers of Israel - there would be no antisemitism. https://t.co/Fz0iBrQuXJ

2022-06-21 09:35:23 This smallpox example was my first exposure to the "dark world of causation", as related here: https://t.co/4gu19vPmzN https://t.co/YG11YEuFdf

2022-06-21 07:35:12 RT @SAFECUNY: https://t.co/Tf0d81kkM1

2022-06-21 07:30:24 @HenMazzig If over 90% of Jews view Israel as the homeland of the Jewish people, what's the point in whining "antisemitism" instead of protesting an uglier form of racism: Zionophobia.

2022-06-21 07:12:41 @Unesco own site on Masada is also worth visiting: https://t.co/8INgYnOnJPIt was inscribed as a World Heritage site in 2001,and has not been taken down yet, after Unesco decided (2016) to erase Jewish connections to the land. https://t.co/kw2NZU8sO5Hurry! https://t.co/oXPwgoM03p

2022-06-19 15:39:11 RT @Israel: Masada: The ancient fortress in Southern Israel and final stronghold where our Jewish ancestors stood bravely against the Roman…

2022-06-19 15:34:12 RT @AdamMilstein: The NYPD has arrested two teenagers in connection with an assault on a Jewish man in Brooklyn last month during which the…

2022-06-19 15:29:43 I would not confer academic credibility to people who signed the Van Leer definition of antisemitism. Let history judge their thoughtlessness.

2022-06-19 15:29:42 Fighting antisemitism: 'Pound of flesh' in 2022 is never enough - opinion https://t.co/r6yIi2KzBz

2022-06-19 14:08:20 @blakeflayton A culture that has produced no knowledge should not be mocked for trying to produce something.

2022-06-16 03:54:01 RT @GalitPeleg: Now that #COVID19 is almost behind us it's time to improve our health systems. Today was the 1st day of "Emergancy Manage…

2022-06-16 03:51:53 RT @FundacionBBVA: Concierto extraordinario en homenaje a los galardonados en los XIV #PremiosFronteras del Conocimiento. https://t.co/Z3…

2022-06-12 09:37:03 May his prediction match his knowledge of history. "Make believe" about the future is cute kids game, "make believe" about the past is hardly cute. https://t.co/LXOxZFtw7m

2022-06-12 09:14:23 RT @StateofTLV: “Everyone understands,” a senior diplomatic source explained to me recently, “that there’s nothing to do. The Palestinians…

2022-06-11 20:13:07 @charleswangb Every universal machine, be it cellular or no cellular, network or no-network, can simulate the do-operator if given the information about the underlying DAG. The issue is getting the DAG structure and knowing what to do with it when asked a question about actions and policies.

2022-06-11 19:57:37 @BifarinTheFifth I love the beads on your hand, each singing a hymn to good taste!

2022-06-11 13:00:44 @terrible_archer @arya_amsha @ruchirsharma_1 I wasn't aware of such anti-Hindu sentiments in America. But I've only been here 62 years.

2022-06-11 05:38:45 @artistexyz @rspadim @amt_shrma @ylecun @GaryMarcus Do you think I would be able to understand what "transformer" is, given my limited vocabulary? You know what my limitations are, I start with a research question, I must know what information is provided in addition to the data, &

2022-06-11 02:05:05 @rspadim @amt_shrma This mapping will be very hard, because some regression folks do have a causal model in mind when the say "regression," because that's all they can say, and some really have "regression" in mind and mean it.

2022-06-10 22:16:25 RT @yudapearl: @AlJazeera: This is another Zionist pink-washing trick, to divert attention from Israel's blood drinking ceremony a few mile…

2022-06-10 21:39:39 @AlJazeera: This is another Zionist pink-washing trick, to divert attention from Israel's blood drinking ceremony a few miles away. https://t.co/5nnrcj6xBM

2022-06-10 21:08:08 A rare visit by Pakistanis to Israel has caused furor in Islamabad https://t.co/WJMz6nHWDO

2022-06-10 20:47:33 Agree! Congressmen need love too. Especially for the courage it takes to remember Jerusalem in a Squad infected Congress. https://t.co/I59LIPi6Cv

2022-06-10 19:31:36 @GaryMarcus @slatestarcodex @MelMitchell1 @mmitchell_ai @timnitGebru @JanelleCShane @katecrawford @merbroussard @ErnestSDavis @ylecun Can we have a link to the main arena, where the punches are thrown among ideas?

2022-06-10 12:52:20 @mariotelfig @philipcball @newscientist It seems that this is the example I was looking for, many thanks . Now I need to swallow and digest.

2022-06-10 12:38:30 @_fernando_rosas I'm purposely trying to avoid man-made interventions which will open the door to all kind of "free-will" speculations.

2022-06-10 10:26:15 @IntuitMachine Yes, this is the kind of example I am looking for, starting with ideal gas and proving that the atmospheric pressure (a macro variable) "causes" the barometer reading (also macro variable) and not the other way around. Here "causes" is defined by counting configurations.

2022-06-10 10:15:42 @philipcball The reservations cited in your article merely express my inability to follow the arguments when all measures of causal strength are expressed as probabilities. My comments now beg for an example that goes from symmetrical '=' to asymmetrical '--->

2022-06-10 08:32:04 Few comments on @philipcball article @newscientist. (1) That causation is an emergent property is indisputable. Both classical and quantum mechanics are governed by the symmetric equality sign '=' at the microscopic level, hence could not account for causal asymmetries 1/3 https://t.co/Aejh9mWceC

2022-06-09 19:05:55 University of Toronto to Withhold Thousands in Graduate Student Union Fees Over Israel Boycott https://t.co/jxqCmbH5yR

2022-06-09 18:52:42 @HumzahAlkindi The journalists Hippocratic oath is not "being found credible by both sides" but "showing the human face of both sides". Since the latter is a prerequisite to "normalization", it is resisted by all anti-normalization forces, e.g. @AlJazeera, @en_emamkhomeini, @RashidaTlaib

2022-06-09 13:08:37 @HumzahAlkindi By: "Humanity not-shown is humanity denied" I mean: A journalist that covers the ME for 25 yrs and does not once show any human aspect of Israeli society is, in effect, telling his/her readers that Israelis have no human qualities, hence, are legitimate targets of assassinations

2022-06-09 11:15:07 @HumzahAlkindi Her death occupies most/all of the discussion, not her journalistic credentials. I believe I was the only one to raise questions about the latter --standing to be corrected.

2022-06-09 11:08:59 This linguistic transition has not been linear but circular, returning to the vocabulary of the Mufti in the 1920's who stated openly "Not even the size of a postage stamp". It was socially unacceptable for a while (after the Holocaust) but regainedrespectability lately. https://t.co/bkZFoqWins

2022-06-09 10:39:49 Are people still doubting whether Twitter posts by Khemeni or Rashida Tlaib may be inspirational to "resistance fighters" such as Sirhan Sirhan? https://t.co/ZS9yTnba60

2022-06-09 10:09:39 A serious question indeed. We can joke about the mentality of @khamenei_ir and his ilk but, considering that he has 900K followers, each a potential killer, Twitter should seriously consider suspension. No one would be allowed to make such remarks on say Mexican nationals. https://t.co/qflQJuHfjs

2022-06-09 09:39:07 @HumzahAlkindi The question I raised was independent of the circumstances under which she died. Nothing justifies "lethal force", yet, a totally different question is whether she was a "reporter" or a "propagandist". I am separating the two issues, why confound them?

2022-06-09 09:28:32 A second list of newly published paper touching on causal inference is compiled here: https://t.co/JnlZGyB2vyI hope to continue on a weekly basis. https://t.co/f5cTGFkU4y

2022-06-09 07:52:24 @rjherber @memarkelliottme I am not a Rabbi, true, but I preach that Judaism is "peoplehood first and religion second". As Ruth said: Amech Ami V'Elohayich Elohay (Your people is my people and you God is my God). Namely, once I become part of your people, accepting your G-d will be easy.

2022-06-09 07:46:23 @memarkelliottme @razingarizona Real estate is the least disputed commodity in the ME conflict. The peculiar issue has always been to reconcile a conflict in which one side says "Let's share" and the other: "Not even the size of a postage stamp".

2022-06-09 04:20:50 @rjherber @memarkelliottme This is because most Jews, even most Jewish organizations, do not insist on correcting the mistake. Some enjoy the ambivalence: pleading "tolerate us like any other religion" with their leftist friends, and demanding "two-states for two peoples" in international arena.

2022-06-09 04:10:40 I will be taking a few days off from Tweeting - need to go to Spain, soon, to pickup the BBVA award I told you about:https://t.co/rEFImBzKvsIf you happen to be at Bilbao, Spain, next Thursday, please drop by the presentation ceremony, 17:30pm at the Euscalduna Conf Center. 1/2

2022-06-09 02:53:49 @razingarizona @memarkelliottme No, NO. Don't get G-d involved. I can do a much better job of explaining. G-d's always assume that everybody knows what they are supposed to know. I'm more tolerant. Besides, most Israelis are secular.

2022-06-09 02:20:20 Glad you brought up this question, because I know that many people think Jews are a religion, as opposed to a "people", bonded by common history, attachment to one land, and common destiny. With this in mind, I hope you join the Jewish aspiration: Two state for two people,1/2 https://t.co/uopaiuCDw9

2022-06-09 01:54:45 100 Years Ago This Month: When Congress Embraced Zionism—Unanimously https://t.co/uKhFmeeKTV via @jewishjournal

2022-06-07 18:47:15 What a song! One would bet it was written yesterday, or at least two weeks ago, when @CNN came out with their Kangaroo verdict as to WHO killed Abu Akleh:"His enemies they all chant he's killing their journalists,He has nothing better to do, but to kill and to kill." https://t.co/VKHnjDQ6s5

2022-06-07 11:08:05 As a carrier of a prize given in his honor, I am posting my Alan Turing lecture of 2012: https://t.co/0s8fldZmjsThe works of other Turing laurates is described here:https://t.co/olirAhwXkQ. (Click on the photo) https://t.co/agx2cMkWAc

2022-06-07 04:52:36 I remember D-Day. No, I wasn't on a boat, I was only 8 yr old. But I remember my father rushing home with a newspaper screaming: HaPlisha Hechela (The Invasion Started), and everyone gathered around the newspaper which had a huge headline: Haplisha Hechela. All I knew: it's BIG. https://t.co/t854iCQUM0

2022-06-07 03:12:55 @ishapiro @GeorgetownLaw Hard to read the small print. Do you havea summary of the issue you had at @GeorgetownLaw?

2022-06-07 02:33:51 @AvivaKlompas @AyaIsleemEn I don't understand why everybody is condemning anti-Israel @AyalsleemEn for pretending this olive tree is in Palestine. History is a life-sustaining commodity, like water and oxygen

2022-06-07 01:29:32 RT @eliasbareinboim: As promised, link to my keynote on the interactive causal learning conf. last week, including my talk + nice Q &

2022-06-07 01:25:48 I am starting a new service on this educational channel: A weekly list (not endorsements) of newly published papers touching on causal inference. First installment: https://t.co/VT5XMP58NYI believe it should help readers obtain a panoramic view of where the field is going.1/2

2022-06-07 00:11:33 RT @emilykschrader: CHAMPION! Meet Monia, an Israeli Druze teenager from Julis who just won the World Kickboxing World Championship in Buda…

2022-06-06 16:51:57 @CureT1Diabetes Thanks, Ellen, I can't get enough of them. I'm in love.

2022-06-06 16:06:20 Those of us tuned to Palestinian social media often read serious articles on what ought to be done with the Jews once they overrun Israel. Who will live and who will die, who will be deported and who be allowed to stay, etc. Like in 1967, they are grossly underestimating 1/2 https://t.co/mB9er80YLk

2022-06-06 15:37:44 RT @michaeldickson: He was killed for supporting Israel: Robert F. Kennedy, murdered by a Palestinian terrorist 54 years ago today https://…

2022-06-06 15:19:28 @mel_blue @EinatWilf Thanks for the correction.

2022-06-06 11:31:32 @rina31420183 @NYT @DrMichaelOren Good to know. Let's see how they celebrate the day that changed the dynamics of the Middle East.

2022-06-06 10:19:01 explicit statement nowadays:"Two states for two people,equally legitimate and equally indigenous".As I've explained in several tweets before, a commitment to equal indigeneity status isof crucial importance, making one's commitment to peace truly hard to fake.2/2

2022-06-06 10:19:00 This article is a must for anyone seeking a concise, 3-page history of the Israel-Palestinian conflict:https://t.co/oXCxb4kTL9Note that @DrMichaelOren set the criterion for peace as: Stating that they favor “two states for two peoples”. I think Israelis will demand a more 1/2 https://t.co/TBoCX1Z32k

2022-06-06 07:44:58 I almost fell from my chair in disbelief: Can @NYT print a piece celebrating 1967? I was about to send a Kudo to @DrMichaelOren, only to discover that it was a 2017 article, not 2022. The content is as relevant today as it was in 2017 but, boy, the @NYT has changed. https://t.co/XFOb7S29zX

2022-06-06 07:19:05 @olivier_klein Wrong. These attitudes are as "ingrained" as mother milk and morning prayers. I have lived with them since 1936 and 1947, before refugees, before "occupation" and before restrictions. Please read R. Khalidi or Said. Why ignore what they are saying? It is ingrained and universal.

2022-06-06 06:57:03 @olivier_klein You can get any result you want from a pole, depending how you phrase the questions. What most Palestinians mean by "two-states" is a temporary arrangement for them to prepare for the next assault. You'd get opposite results asking: Does Israel have the RIGHT to exist. Zero.

2022-06-06 06:46:06 @WahbaRachel @Ostrov_A @bobjcarr There is one difference, of great significance. I don't know an antisemite who admits to being one. I don't know a Zionophobe who denies being one. They consider it a duty and an honor to deny Jews a homeland.

2022-06-06 05:08:06 @Ostrov_A @bobjcarr Save your questions, Professor @bobjcarr. No one will call you anti-Semitic again. We have a new title for those who make vile allegations of "apartheid" and attribute a few racist shouts to 40,000 peaceful dancers. We call them Zionophobes, something they can't deny.

2022-06-06 04:45:25 @olivier_klein There is no danger of antagonizing those who don't match this description because, and this is the cardinal mistake of most Western observers, analysts and peace makers, the number of exceptions is infinitesimally miniscule. Standing to be corrected, all evidence admissible.

2022-06-06 03:55:09 The West Bank is still under occupationbecause since 1936, Palestinians view "co-existence" as a defeat. https://t.co/OoyKY5oaq2

2022-06-06 03:49:56 I would put it even stronger, in present tense: “More than the Palestinians want a state for themselves, they want to destroy and deny a state to the Jewish people.” Moreover, they would rather give up all hope for the former to retain some hope of the latter. https://t.co/REMsgqATEt

2022-06-06 02:21:19 What have you done to me? I am listening to Isaac Y Nora again and again https://t.co/JCUpNsDbhbI can't quit, and I can't work. I don't want to work, I want to sing with them, more and more!VEINTE AÑOS, Caballo Viejo! Que Nadie Sepa Mi Sufrir! Amazing! https://t.co/12G8PPDvw2

2022-06-06 01:54:40 June 6th marks 74 years to the 6 day war. Some readers ask: "how did it start?" Many more ask: How come the West Bank is still under Israeli Rule? @EinatWilf has a simple answer, with which I fully agree, it's a necessary temporary control till the Arabs say: "Equally permanent" https://t.co/6sQ5dDtVCL

2022-06-06 01:42:31 @KhaledAbuToameh May their forecast of the future matchtheir knowledge of the past. Amen!

2022-06-06 01:31:18 RT @romanhistory1: Today 70AD Titus and his Roman legions breach the middle wall of Jerusalem. Titus, a renowned military commander and fut…

2022-06-06 01:27:01 RT @DavidHarrisAJC: #OTD in 1967, war erupted in Middle East after bloodcurdling calls for Israel's end.Until 1967, Egypt occupied Gaza&

2022-06-05 23:33:47 @artistexyz @RWJE_BA Gee, I love this song, the music, and this family. I love them so much, that I can't see how it is related to EBM, unless EBM no longer stands for Evidence Based Medicine . Is it?

2022-06-05 20:58:47 @DKedmey @sapinker @Gena_I_Gorlin @DavidDeutschOxf @richard_landes @MattiFriedman @NeilRogachevsky This "projection project" is of crucial importance in the fight against Israel delegitimization. Most BDS followers are the gullible and misinformed, who go for "'progressives' cannot all be wrong". Only a credible explanation of WHY they ARE wrong would sway back the gullible.

2022-06-05 15:41:16 Not bad at all! Reminds me of Omar Khayyam:Here, with a book of verse beneath the Bough,A Jug of Wine, A Loaf of Bread—and ThouBeside me singing in the Wilderness—Oh, Wilderness is a Paradise enow!#Bookofwhy https://t.co/Qh4UNGHbzp

2022-06-05 15:32:06 @databacked101 their sons in the war, nor faces of orphans of fallen Russian soldiers for fear of raising sympathy for Putin is not doing his/her duty as a journalist.2/2

2022-06-05 15:24:28 @databacked101 I love this comparison, because Putin's declaration that Ukraine is not a legitimate country matches Palestinian declaration that Israel is not legitimate. But forgetting this parallel, a journalist stationed in Russia and wouldn't show faces of Russian mothers who lost 1/2

2022-06-05 08:38:42 Today is the Jewish holiday of Shavuot, the day the Israelites received the Torah on Mt. Sinai. We also read the Book of Ruth https://t.co/FQE6B6U0e9 for a variety of reasons: It's a story about the harvest time, and it is a story about conversion to Judaism. But for me,1/2

2022-06-05 07:54:18 RT @Ostrov_A: The gorgeous iconic Rothschild Blvd in Tel Aviv, Israel! https://t.co/98023hYw5D

2022-06-05 07:42:10 @databacked101 No free lunch! Humanity not-shown is humanity denied.

2022-06-05 04:25:16 There is no law against mapping. The danger of mapping is operationalized by ideologists such as Rashida Tlaib and Ilhan Omar, whom BDS's lunatics view as authorities on who deserves Allah's wrath. https://t.co/qWWTxRMG5R

2022-06-05 03:33:32 @kareem_carr No one guessed.My two words are "curve fitting".But, since this would also make some statisticians cry, which @kareem would hate to do, I would modify it to read:"Theory-free curve fitting", thus sparing statisticians' feeling.

2022-06-05 03:21:17 I can't get over @NYT platitude: “Journalists cannot do their jobs if they are targeted with impunity by any side in a conflict.” Of course! But would they accept the corollary: “Nor can Journalists do their jobs if they consistently target with impunity one side in a conflict.” https://t.co/56n67vExQa

2022-06-05 02:08:54 @kareem_carr I can do it in two, does it count?Guess!

2022-06-05 01:54:57 @FJnyc @nytopinion @NYT There is nothing wrong in being "state run media" if it remains committed to the Hippocratic oath of the profession. The trouble with Al Jazeera is that has become the Mecca of journalistic oath-breaking.

2022-06-05 01:39:43 I am proposing this slogan as a prerequisite qualification for every reporter sent to cover the Middle East. https://t.co/rHNOtOrwXh

2022-06-04 22:29:21 I have humbly asked: In her 25 yrs of covering the Middle East, has Abu Aqleh once described the human face of Israeli society? How many times has she portrayed an Israeli child or a grandmother or any human feature of 8 million people trapped in that conflict? https://t.co/1GYHaxHoME

2022-06-04 21:41:16 I think it is only reasonable to ask my esteem colleagues at the @NYT to spell out a few litmus tests that would help naïve readers distinguish an "honest reporter" from a "dishonest reporter". I've formulated one such test, but would be curious to hear from @NYT. https://t.co/ZUu1A7JQky

2022-06-04 20:01:25 @ben_shore I am not familiar with the "scaling hypothesis". Curious.

2022-06-04 19:04:05 @amiralex @HaShemsWife @Doc_Muh I would add that, having given this phenomenon a name: "projection" does not absolve us from continuing an in-depth, psychopathological study of pseudo-progressives' animosity towards Israel."Projections" come in different shades and may emerge from other pathologies.

2022-06-04 18:53:33 @MerkinMuffley5 @SpclDelivery The reason such a term is not coined is that the world has not seen another case of a beleaguered nation enduring a 74 of war and finding it impossible to defend herself except by controlling the territory from which her enemies wish to operate, encouraged by Western observers.

2022-06-04 18:39:04 @nytopinion Putting aside serious questions about her death, permit me to ask on what basis you determine that Abu Akleh "served as a model of courageous, honest reporting." Do you mind spelling out a few litmus tests to distinguish "honest reporting" from "agenda-driven propaganda"? @NYT

2022-06-04 18:15:16 I concur. This was one of my best interviews. It is so rare to find an accomplished scientist who hasn't already formed a fixated opinion about causation, willing to explore its mysteries from fresh perspectives. Highly recommended. https://t.co/cgqeoboxZG

2022-06-04 07:27:53 UCLA Students Urge UC Regents to Take a Stronger Stance Against Antisemitism https://t.co/ZiBPGFUHWG via @jewishjournal

2022-06-04 06:17:28 @MerkinMuffley5 @SpclDelivery And I have no idea why people use the word "occupation" to label a "temporary control".It is "temporary" until the attacker agrees to end the war of 1948, and "control" to distinguish it from "annexation" which would have made Gaza and WB "part of Israel".

2022-06-04 05:45:06 @shalom123211 @stefanharmeling The question is whether there are graphs where Rule 3 applies and in which you can't remove the action in two step, Rule 2 followed by Rule 1.

2022-06-04 03:33:46 Have mercy on the poor guy. Did you give him a chance to repair the damage? Did you draft the statement you would like him to issue campus-wide. How about a list of items he canact upon next week? Most administrators are bewildered by the BDS Zoo. They never encountered fascists https://t.co/vvQ3pEZgEa

2022-06-03 21:32:49 @hyounpark Interesting. The people who are making AI-driven business decisions, what kind of literature do they read? Perhaps we can submit a business-minded version to the top magazine they read?

2022-06-03 19:27:20 @TobiasNavarro Welcome to the causal revolution.

2022-06-03 17:13:18 @MerkinMuffley5 Not needed? The fact the Israelis can fly to Dubai did not deter Diaa Hamarsheh from killing 5 people in my home town, Bnei Brak. IDF presence in Jenin (Diaa's home) is needed to detect and prevent future attacks. Until when? Until you advise Palestinians to accept co-existence.

2022-06-03 07:36:48 RT @AmbDaniDayan: We in @yadvashem mourn the passing of Belgian Righteous Among the Nations Andree Geulen (1921-2022). She saved innumerab…

2022-06-03 07:35:07 RT @LahavHarkov: It's done: Israel and Saudi Arabia agree, with US mediation, on security arrangements in Straits of Tiran, and Israeli air…

2022-06-03 06:25:13 @shalom123211 You are right about the red arrow.As to Rule 3, can you deduce it from 1 and 2?I remember someone else mentioned such a possibility, so I am going to check on my side too.

2022-06-03 06:03:25 Again, there is no need for #ADL to display its credentials (as "good" Jews) by the trivial slogan "one can reasonably object or criticize Israeli policies." We count on #ADL to inspire unapologetic, secured Jews. The image is libelous on its on de-merit. @JGreenblattADL https://t.co/ZcVuE5xG2E

2022-06-03 05:51:40 I don't see why one must be #antisemitic to be dangerous, repulsive and stupid. Suppose this image points exclusively to the Israeli-Palestinian conflict, and blood-libels never existed in Jewish history, shouldn't everyone be appalled by this image? Shouldn't #ADL condemn it? https://t.co/woYMNaY0JZ

2022-06-03 04:59:18 @MerkinMuffley5 I'm responding to the practicality issue (not to populist slogans). Permanent occupation does not solve a thing, but temporary control, awaiting your enemy to accept co-existence is a necessary measure for survival, and might just lead to peace, if Westerners focus on the hurdle.

2022-06-03 04:35:29 Brilliant! This is the most convincing explanation I've heard. It does not lean on "antisemitism", not even a latent one, and it explains why Asian-American departments are the first to slander Israel as the source of absolute evil. We need to study the anatomy of "projection". https://t.co/XJ6sMjBghd

2022-06-03 03:15:20 @MerkinMuffley5 We are getting to the root of Westerners delusion. They are saying "some", "some" and "some", unable to resign to the fact that NO Palestinian is willing to accept a permanent Israel, in any borders. Some "accept" in English, none in Arabic, and surely not near school children.

2022-06-03 02:50:36 Well put, with one addition: 4. Western Anti-Israel NGO's. Moreover, terrorist groups have ceased to exist

2022-06-03 02:30:09 @MerkinMuffley5 As to mental illness, I don't know you personally, but I honestly believe there is something profoundly wrong in the mentality of Israel de-legitimizers who do not pause even once to remind Palestinians how they can get freedom and dignity, if this is indeed their goal.

2022-06-03 02:10:42 @MerkinMuffley5 Let's analyze it a bit. Assume for the sake of argument that Palestinians are dead-sworn to eradicate Israel no matter what Israel does. What human rights should they enjoy while Israel is preventing them from fulfilling their dreams? Voting? Importing weapons? Come and go?

2022-06-03 00:19:32 @MerkinMuffley5 This is one way of putting it. But progressives are enlightened, right? So they shouldn't ignore an alternative theory, according to which the "occupation" is a temporary control of a territory from which millions of people declare their sworn intention to destroy a neighbor.

2022-06-03 00:09:28 Well said. Did I say it? https://t.co/DMqViMTRv7

2022-06-02 23:54:17 @FJnyc I think her sign is very very meaningful: A society that lacks history and lacks distinct culture would resort to any trick, even the age of its elderly, to fake a semblance of "culture".

2022-06-02 23:38:27 @MerkinMuffley5 The pathology resides not in condemning one policy or another but in obsessively negating the legitimacy a people's homeland and in willfully suppressing the possibility that at least some of those policies are forced upon Israel by its neighbors' actions and declared intentions.

2022-06-02 20:24:13 @MerkinMuffley5 If you have found a non-pathological explanation for this obsession, please share. Readers are seriously puzzled.

2022-06-02 20:19:34 For educators, instructors and bystanders who are waiting for gentle material for teaching causal inference in AI and ML curriculum,here is a paper titled "Causal Inference in AI Education", which has just been accepted for publication by JCI: https://t.co/WyxUnL8peI1/2

2022-06-02 19:44:58 Looking forward to you book. I hope it doesn't limit its coverage to the absurdities of modern Zionophobia, but includes a serious psychopathological study of this obsession. It defies every scientific explanation I could find, thus contributing to its longevity. https://t.co/8fnoFyUxg1

2022-06-02 18:03:22 @soboleffspaces It takes philosophers 1-2 generations to rid themselves of "problems" that older generations have concocted. This is what happened, for example, to "evidential decision theory."

2022-06-02 17:49:47 How I wish you were right @mishtal. How I wish Islamists were our only obstacle to peace. In truth: the overwhelming majority (99%) of our neighbors refuse to accept Jewish self-rule

2022-06-02 17:11:43 The first logical rationale for destroying a state: YOUTH.And they call it "culture". https://t.co/Ei8Ydg6RmY

2022-06-02 16:57:39 Scientifically speaking, progressives animosity towards Israel is the most challenging, and barely studied psychopathological puzzle of our time. https://t.co/xNPXRoSzB9

2022-06-02 16:32:55 @soboleffspaces Backtracking cf doesn't solve any problem. But some philosophers insist that the utterance: "Had the soldier shot, the captain would have given a signal" is legitimate, rather than an incorrect distortion of "if the soldier shot, we can conclude that the captain signaled."

2022-06-02 06:27:13 @GnomeSortingHat Trying

2022-06-02 00:52:02 @MarkusSchacher The article conveys useful information about CI implementations. But the technical portions could stand polishing.

2022-06-01 21:49:22 June 1, 1941, is the day of the Farhud pogrom in Baghdad https://t.co/Jyn5XjO6wa. My late wife was 6 yr old when the looting and killing broke up in her street. This is how she described Jewish Baghdad in her own words: https://t.co/DnhD9fg9U8

2022-06-01 08:11:42 The violence comes from the pepper spray gun the "old lady" is holding. Watch the guy with the blue hat crunching after being sprayed. He was interviewed on Israeli TV, to explain how one could possibly be less than brotherly to an old Palestinian lady. https://t.co/wYrP0GrZ3S

2022-06-01 07:24:12 Our persistent clinging to good old probability theory is one of the main obstacles to the understanding of causality. Here is a recent Royal Society paper falling into this rabbit hole:https://t.co/XEuvetOctnI was happy nevertheless to find a reference to Papoulis (1965), 1/2

2022-06-01 07:04:43 In judging the wisdom of this UN media program, permit me to mention an alternative program: The Daniel Pearl International Journalism Institute https://t.co/bQ4Rcdk2zCwhich I believe stands up more faithfully with the Hippocratic oath of the journalistic profession.@Ostrov_A

2022-06-01 07:04:42 UN names Palestinian media program for slain Al Jazeera reporter https://t.co/oWDQNykwKA via @timesofisrael

2022-06-01 05:35:58 Best analysis by far of the death of Abu Akleh.Related in some fundamental way to the discussion we had about the importance of explanatory theories in Popperian epistemology. The idea that Israeli soldiers have everything to lose and nothing to gain 1/2https://t.co/oVVs8C5MqF

2022-06-01 02:48:43 If Jewish leadership failed us on Jewish Heritage Month, @newzionists should fill the gap, since Israel is the culmination of Jewish heritage. How about extending it by one week, till June 7, and make up for this sloppiness by celebrating seven pillars of Heritage, one each day. https://t.co/vW1ULfnxMa

2022-06-01 00:07:36 Hats off to @emrek and @amt_shrma for the development of DoWhy and its new PyWhy platform. Next step is for @MSFTResearch to help starving academia re-educate the thousands ML folks who are stuck on curve fitting and will soon be needed in the workforce. https://t.co/tecHBFdxJQ

2022-05-31 21:41:39 @memosisland @MSFTResearch Thanks for alerting. I just retweeted to cheerful readers.

2022-05-31 21:37:15 Gee, this looks like one of our slides, cheering us to know that at least one company, Microsoft, is on the right track. Makes me wonder what keeps the other ML companies in business ... I know, shortsighted funders.@erichorvitz

2022-05-31 21:37:14 DoWhy evolves to independent PyWhy model to help causal inference grow https://t.co/vrGSxihrGO

2022-05-31 11:59:44 Irwin Cotler, former Minister of Justice and Attorney General of Canada, is the most knowledgeable Human Right scholar I've met.Watch his video below, regarding Zionophobia and its subliminal process of spreading. https://t.co/sBALer25zj

2022-05-31 11:45:09 RT @naftalibennett: The Iranian regime won’t want you to watch this video until the end.But you should. https://t.co/5BdO6xRvf4

2022-05-31 11:38:48 And I am amazed that the author seems totally unaware of modern Causal Inference, a science that formalizes and operationalizes the interplay between mental theories and experimental data.He certainly has not read #Bookofwhy or modern critics of RCT https://t.co/B0TKO5MzlJ. https://t.co/xu6zfePUob

2022-05-31 11:04:07 This short article by my friend Lori Lowenthal explains why weaponized forms of ethnic studies must be fought individually, in every school district. How? By contacting your school board and insist it not be permitted within their district. More in https://t.co/fEy4EbkeXR

2022-05-31 11:04:06 Print Issue: Uncovering a Curriculum | May 27, 2022 https://t.co/jwis0T4Wcy via @jewishjournal

2022-05-31 08:19:59 If they agree to host the next AI-ML conference in Riyadh, I'd volunteer to organize a Causal-AI workshop, and submit a paper on Middle East Peace - Necessary and Sufficient Conditions. https://t.co/Rxapaf5d4a

2022-05-31 06:59:40 @TriciaRoth18 @deborahlipstadt Thanks, Tricia. I've just asked Alvin what affiliation exists between his Institute and the new Center. Aren't you surprised by the absence of "Israel" in the Center's charter? For me, it's a warning sign of reluctance to face issues, or reluctance to rethink stale paradigms.

2022-05-31 06:18:22 Not bad for a politician https://t.co/F8MBrrJhg1

2022-05-31 03:25:54 Can data reveal the flow of ideas? Here is an interesting paper "ON THE EVOLUTION OF A.I. AND MACHINE LEARNING: TOWARDS MEASURING AND UNDERSTANDING IMPACT,INFLUENCE, AND LEADERSHIP AT PREMIER A.I. CONFERENCES"https://t.co/BCs5INWT0HBut who influenced whom? &

2022-05-31 01:51:12 Seeing a new Jewish Center always cheers my heart, but seeing one that does not explicitly address the pressing issue of our time - Zionophobia - makes me worry: Isn't it another waste of Jewish philanthropic resource, or another license for University inaction? @deborahlipstadt https://t.co/rrV6l95OBi

2022-05-31 01:11:58 Socrates would say it more poetically: If my culture, language and holidays offend your religious sensibilities, you better switch religion. https://t.co/GjBElJCqpU

2022-05-30 22:40:33 Funds Cut From City Program for CUNY Law After Faculty Backs Israel Boycott https://t.co/EjGqL9jNpp

2022-05-30 20:59:40 Reminds me how our son, Daniel, got lost on this very beach when he was 6 - abruptly swallowed by the crowd. We found him 3 hours later, two miles away, drinking soda with the life guards. "Why were you worried" he asked. https://t.co/TaGMQqfcPU

2022-05-30 20:15:01 China's nervousness about what Israelis think of her is a good sign. Next they might be nervous about what their own citizens think, and that's democracy live. https://t.co/84tEX3WImk

2022-05-30 20:03:55 RT @havivrettiggur: For Ethiopian Jewry, Jerusalem Day is something immense and visceral.

2022-05-30 19:58:10 @CotlerWunsh Must be a brain child of @KenRoth. We can forgive them for the blood libels, but we can't forgive them for staining the words "Human Right"

2022-05-30 19:42:36 @cubic_logic We are light years away from each other. To know how to continue, please tell me if you've read #Bookofwhy. My view is diametrically opposite: What you want and what is possible CANNOT be dictated by statistical properties of the data. See many examples in https://t.co/LZes88schJ

2022-05-30 14:38:31 It was about a year ago that Stanford came up with the term "Foundation Models". Our discussions did not help me understand what it is, and I thought the term has died. Today I got this message:https://t.co/Eq3UdpYbx3So I'm wondering: Anyone knows what it is? @machinelearnflx

2022-05-30 14:23:11 @RDub2 @sadneurons @f2harrell @DrewLevy So the "spuriousness" was not in the kind of regression you aimed to estimate, but in the method you used to estimate it? I would not call the latter "spurious correlations" because what you got is not the correlation but some surrogateof it.

2022-05-30 10:34:43 thus depriving our students of the little protection they can derive from the label "anti-Semitism" and the general revulsion it evokes. History will judge these authors by the consequences of their action.@EinatWilf @GilTroy @havivrettiggur @BenMFreeman @GeraldNGOM @blakeflayton

2022-05-30 10:34:42 The authors of JDA make their living by showing how creative they can be in generating nuances and sub-nuances of definitions. What history will never forgive them for is doing their acrobatics on anti-Semitism, calling it "Declaration", and appropriating the name Jerusalem, 1/2 https://t.co/sTusYawk1Y

2022-05-30 08:35:50 @hosamindeed Joy is in the eyes of the beholder. These girls seem very joyful to me: https://t.co/kr0pLMlf52perhaps because I know their song. An Al-Jazeera watcher will see no joy in anything expressed by an Israeli, unless it can be used to vilify Israel. Yes, Zionophobia is a disease.

2022-05-30 08:05:24 @cubic_logic We need a description of what you want -- thus far missing from our discussion &

2022-05-30 07:51:13 @cubic_logic @gottfriedmath @sadneurons @f2harrell @DrewLevy "Fooled by attempting" requires some de-cryption. Do you mean "you can't estimate them" or, "even if you estimate them, you won't get what you want."? If the latter, we need a description of what you want -- thus far missing from our discussion &

2022-05-30 07:29:04 @stephensenn @AdanZBecerra1 @f2harrell @sadneurons @DrewLevy My serious point is that we should separate definitions from estimation. One can discuss whether Pearson's "spurious correlation shock" was justified without detailing how he estimated those correlations.

2022-05-30 07:18:30 The authors of the "Jerusalem Declaration of Antisemitism" will have to answer history for the consequences of their action. They have given hard anti-Semites a powerful new argument: "You see, even Jews do not agree on what antisemitism is." @deborahlipstadt @JGreenblattADL https://t.co/aYUEVyFI1L

2022-05-30 06:42:31 @nirhasson Why don't we see any clips of the dancing girls in the west. Can you share some?

2022-05-30 06:32:23 @bianca_aguglia Many papers, books and other resources can be found on my website https://t.co/vNOB7Oq4vIWorks by other teams are well referenced, but only if they are relevant, clear and correct.

2022-05-30 04:48:50 The question below lured me into re-reading the first 6 pages of https://t.co/LZes88schJ. Wow, I'm so enchanted by how clearly and beautifully it's written, that I can't resist but invite you all to enjoy it. Please compare to what we have been seeing lately in ML, Stat and Econ. https://t.co/1NHu0vEKfM

2022-05-30 04:15:22 @cubic_logic Disagree. What Pearson lacked in 1899 had nothing to do with the way correlation was estimated, but with the way it was interpreted.See Stigler "History of Statistics" or @Bookofwhy p.79. “To those who persist in lookingupon all correlations as cause and effect ... a shock.”

2022-05-30 03:58:30 @cubic_logic "Spurious relations that may come up by the tool itself" should not be called "correlations", but "wrongly measured correlations". "Spurious correlations" are still correlations, defined by the probability distribution governing the data, not by the tool chosen to estimate them.

2022-05-30 03:52:03 @cubic_logic @gottfriedmath @sadneurons @f2harrell @DrewLevy If you are trying to estimate correlations and you are estimating something else, you dont call it "spurious correlations". I would call it "wrongly estimated correlations".

2022-05-30 03:46:51 @EliSennesh @Eradicator_NYC The problem is not "being more aggressive" but knowing to what to ask that would be both effective and in compliance with academic free speech.They think they know, but end up praising administrators for objecting to boycott, instead of demanding an explicit mention of Zionism.

2022-05-30 03:35:45 That is what the Jewish establishment said after Munich Beer Hall Party (1923). It's wrong! Hate must be nipped in its bud, and the bud is @CUNYLaw. @deborahlipstadt @JGreenblattADL @SAFECUNY https://t.co/PA0zYeD91U

2022-05-30 03:19:58 Very good question. Ans. Causality textbooks all use statistics as a tool, enriched with causal relations and causal notation. See https://t.co/LZes88schJ for a good example. Causal notation is what Pearson lacked in 1899, and what modern stat textbooks still lack in 2022. https://t.co/qngqyK7VPn

2022-05-30 03:06:53 @UweSiebert9 @sadneurons @f2harrell @DrewLevy If "spurious correlations" may stand for anything that is useless, then we should not use the term "correlations". If I want to predict Y given X and I wrongly measure the correlation between Z and W, is that "spurious correlations" too? How about if I wrongly divide by 100?

2022-05-30 02:57:12 @cubic_logic @gottfriedmath @sadneurons @f2harrell @DrewLevy Is the problem with the pre-smoothing? Or with the correlations themselves?

2022-05-30 02:49:18 @stephensenn @AdanZBecerra1 @f2harrell @sadneurons @DrewLevy Causation was needed for defining "spurious correlations". We are now discussing @f2harrell's "reliable association" vs. ""cherry-picked correlations". Are the latter a species of "spurious"??

2022-05-30 02:33:49 @Wetassprior @KeurigBad That's the first reason I heard that makes some sense.

2022-05-30 02:23:19 CUNY Law Faculty Endorse Israel Boycott Resolution https://t.co/BDdGOJxHP1

2022-05-29 22:20:42 Amazingly joyful celebration of Jerusalem Day! With unexpectedly high number of participants and, despite Hamas threats, and contrary to Al-Jazeera "reporting", a successfully managed and peaceful celebration with no major problems. Hats off to Israeli police and Israeli public. https://t.co/B1UPsLp38j

2022-05-29 21:46:18 @JeremyBenAmi Teacher: Happy birthday, Jenny.Jenny: Thank you, teacherTeacher: Charlie, what are you hitting Jenny for? Charlie: I wish her dead, I must spoil her birthday.Jenny: No way! It's my birthday! @JeremyBenAmi: Jenny, Do you have no critique of what your birthday turned into?

2022-05-29 07:46:24 @DavidDeutschOxf Profoundly concise summary, agree. But I have yet to see an illegitimate use of power which is not justified as a prevention of its illegitimate use. And that includes Bin Laden justification of 9/11, the Arab armies attack in 1948, and Putin's invasion of Ukraine.

2022-05-29 07:20:51 @StillTr05207382 Good question and, indeed, the Model-based paradigm advocates "model-assisted" learning @slashML

2022-05-29 07:15:13 I am watching images from the roof of Al Aqsa Mosque, the third holiest place to our neighbors, on which a nervous warrior is doing his best to protect the holiness of the place. https://t.co/B2Yoj2T7Ns

2022-05-29 06:52:20 As usual, the most thoughtful and inspirational expression of Yom Yerushalayim comes from @YKleinHalevi. Note his last remark on why we all need to be thankful.She'Hecheyanu! https://t.co/Z9Qu4TgJBG

2022-05-29 05:24:19 It would be amusing to read how @AlJazeera reports on these images, whether the rocks were not thrown by the police towards the peaceful prayers, and whether Al Aqsa is still in imminent danger. https://t.co/FSRsRvuaIb

2022-05-29 05:11:02 CUNY's president will claim, like all the others, that he is against divestment. That does not exonerate him from his duty to state clearly and explicitly why Israeli and Zionist students and faculty are most welcome to @CUNYLaw. https://t.co/pNUFqM8pv6

2022-05-29 04:50:20 Tell me if it doesn't sound more punchy: "Those greedy, thieving, Jewish supremacist, apartheid-loving Nazi Zionists keep trying to say that I am a Zionophobe" https://t.co/p0JNQQezFM

2022-05-29 03:40:03 The #MachineLearning enterprise seems to be splitting. While some insist on going from "Data-Centric to Model-Centric" paradigms, https://t.co/KcmiOil9roothers recommend going backward, from model-centric to data-centric: https://t.co/u4g3OxK7RGCan #BeADeepLearner post this?

2022-05-29 03:19:15 @MaccormickIan @JessicaHullman @CzannerG @aaronylee @DrXiaoLiu I couldn't find much commonality between the #MachineLearning and #psychology papers considered. The former are data-centric. I am also wondering what kind of integration methods are available to those who wish to integrate predictive models &

2022-05-29 02:52:39 @martin_garcia_a Yes, precisely. I got to know it by being on the examining board.

2022-05-29 02:37:11 Another slip of tongue in my lecture yesterday was the date when Pearson discovered "spurious correlations". It was 1899, not 1999 as I wrongly uttered. Please forgive, and lookup @Bookofwhy how surprised and confused Pearson was. https://t.co/BhZQSgBXm9

2022-05-29 02:20:07 @erikbiz @brianstelter @CNN As I said, the team overseeing CNN editorials has changed, and is now dominated by Ex Al-Jazeerapseudo journalists. Actually, it needs not be "dominated" by this ilk, enough for it to be "seeded" with one or two loudmouths, for CNN to go down the slope of Al-Jazeera

2022-05-29 00:32:06 The best way to access the literature for each of the Seven Wisdoms of Causal Inference is to go to our website https://t.co/vNOB7Oq4vIand search for the key words in the titles. The references in each of these articles will lead you to relevant works done by other teams. https://t.co/D86F71DDr4

2022-05-29 00:20:22 80 years passed since the first time I visited Jerusalem with my mother. 50 years, since my son Daniel had his Bar-Mitzva at this place. But time does not seem to dim the special light that shines from this city upon her lovers, poets, and dreamers.Katonti Mi Tsair Bannayich. https://t.co/Dqz6jqn0XI

2022-05-28 16:01:27 This slide provides a taxonomy of the gifts that Causal Inference has given to science. Once a research problem is classified into one of these seven compartments, the relevant literature, algorithms and software packages can be harnessed to walk you towards a solution https://t.co/ofomux8SFX

2022-05-28 15:37:28 @erikphoel @philipcball @newscientist @renzocom I'm referring to PS, PN, and PNS which, although they involve probabilities, cannot be expressed by the language of probabilities alone (Rung-1), without counterfactuals, which are Rung-3. It is the counterfactual component that tells you where they emerge from.

2022-05-28 15:05:26 Thank you Ross. Here it is, my talk of yesterday already made into a video on youtube, accompanied by the slides, ready to be used for fun, profit and science."What is Causal Inference all About?" https://t.co/46Ju37SXFn

2022-05-28 07:51:49 No, No, @DerbyChrisW is not an antisemite - some antisemites are rational organisms. He is suffering from Zionophobia - a total collapse of one's moral compass. No treatment is known, though some research is in progress. Imagine UK under Corbyn. @mishtal https://t.co/yD7xVplbCY

2022-05-28 06:23:04 @erikphoel @philipcball @newscientist @renzocom I wish I could follow your ideas on "causal emergence", but I am stuck at your definitions of necessary and sufficient causation, which are cast in probability language - a mathematical impossibility. For example, PNS is not A/B. See https://t.co/rGHIc9kH0O

2022-05-28 05:48:02 @sadneurons @f2harrell @DrewLevy I wonder indeed how authors of modern statistics texts are defining "spurious correlations". Are they still trying? They obviously cannot do it w/o causation but, since causation is still taboo-ish, it is always amusing to see how they try.

2022-05-27 11:55:21 Paraphrasing Gillerman, it's a tiny country that ha contributed to the world more than all those who wish her dead. https://t.co/x9izjjGwl0

2022-05-27 11:49:12 I am shocked! When did Iraqis acquire Palestinian mentality?This is serious matter. If any of our readers is of Iraqi nationality, he/she might be hanged for tweeting on this account, since I have dual, Israeli-American nationality. Watch out!!! https://t.co/E0mOiApfo6

2022-05-27 07:55:00 Almost forgotten. A devastating fall into 1st place in # of Turing Prize winners per capita: https://t.co/VgGWJKhRGPPartially guilty of this fall, I dare make a personal appeal to our neighbors: Join us in falling, just say "equally indigenous" and things will fall into place. https://t.co/1MzOAX5bfO

2022-05-27 07:27:49 As a scientist, I must note another terrible fall, straight into 11th place in # of Nobel Prize winners per capitahttps://t.co/CbT0GJJt2qFalling and falling... https://t.co/Mz0psa18yg

2022-05-27 06:48:39 And for those who can't attend all the upcoming meetings on causal inference and graphical models, here is a competitions to consider:https://t.co/WgREpIFzgFI am not sure what the rules are, but if you have a powerful problem solver, it may be the right arena to demonstrate.

2022-05-27 06:35:09 I am in receipt of an announcement for another interesting meeting:First Workshop Causal Representation Learning at UAI 2022: https://t.co/WgREpIFzgF5 August 2022, Eindhoven, The Netherlands, hybridSubmission deadline: June 6, 2022, 23:59 AoE

2022-05-27 06:26:29 RT @analisereal: (1/2) We have an exciting line-up of virtual keynotes coming up for the Workshop in Interactive Causal Learning (https://t…

2022-05-27 04:30:42 @TheAcsMan @hearNtalk @ShimonLevit @CNN Disagree. Editorial standards in every news outlet were instituted to prevent glaring flukes like that from happening. The fact that it did happen indicates that basic standards were not adhered to, namely that it was not a "fluke" but systemic change in the overseeing management

2022-05-27 00:52:03 @TheAcsMan @hearNtalk @ShimonLevit @CNN CNN has a better track record, agree. That's why everybody is surprised by the amateurish way they reported on Shireen's death. My point: It's not sloppiness, but the result of a process where graduates of Al-Jazeera and Bir-Zeit have taken over CNN coverage of this conflict.

2022-05-27 00:43:46 Good news for inquiring readers. I've just been assured by the organizers that my talk tomorrow will be recorded and will eventually be made available on Youtube. How come so many readers want to know "Where data science is going" ? https://t.co/Un0uv736XL

2022-05-26 22:47:13 @Tiff__Annie_ @EBRheum @rachel_loeb @TIFF_annIE Why don't you take the lead and show them the way to say something meaningful, and expand on what I've started?

2022-05-26 21:56:14 @Tiff__Annie_ @EBRheum @rachel_loeb @Tiff_Annie, I am not exactly sure what you find to be missing in the literature you are probing.

2022-05-26 20:18:20 @humanai4ron @HarrySurden And I was hoping the first chapter alone would reconfigure readers neural architecture. Challenge for the 2nd edition.

2022-05-26 13:30:55 @hearNtalk @TheAcsMan @ShimonLevit @CNN Good point. People often ask me: How can you judge a network, or a politician, or a colleague by what they think about Israel - a small corner in the global scene? My answer: It's small, but it's a corner I know very well, allowing me to tell where they get their knowledge from.

2022-05-26 08:39:30 In 2001, @CNN did not join the "Dancing Israelis" conspiracy theory. Why? It's editors and reporting staff were not so heavily dominated by graduates of the @AlJazeera school of journalism, where Qatari thinking is prerequisite for Qatari salaries. https://t.co/0ZFsFrKh8G

2022-05-26 07:49:05 Thanks for sharing. I see it for the first time, and Gee, had I known they would turn it into a movie, I would have acted more professor-like, as if I knew what's good for PhD students. Great movie. https://t.co/1hiKDOktBy

2022-05-26 03:10:34 @OJPAC @nyspolice @NewYorkStateAG @GovKathyHochul @JakeAAdler @JShabs95 Too early to tell if they are David Duke disciples or @AOC-inspired?

2022-05-26 02:14:13 I visited the place in 1941, which was a dark alley, hardly 3 meters wide. Enough to say a prayer, but not enough to grasp its historical significance. https://t.co/KmhOAecFg4

2022-05-26 00:47:31 Suissa, as usual, has it right conceptually. However, the word "antisemitism," by its history and its very nature, connotes "begging for protection" -- unbefitting to winners. Winners' fighting word is Zionophobia. See https://t.co/7jn0po33Op.@DavidSuissaJJ

2022-05-26 00:47:30 Fighting Jew-Hatred As Winners, Not Victims https://t.co/KXKinJiROQ via @jewishjournal

2022-05-25 21:53:35 RT @Efune: 1. Per CNN's own map, the Israeli forces were positioned in between Abu Akleh and the militants. If they were shooting at each o…

2022-05-25 10:38:03 @mishtal @Ostrov_A @noatishby They can do it! CNN used to have such teams before they started hiring graduates of @AJEnglish school of journalism = Muslim Brotherhood in secular cloaks, British accents and Qatari salaries.

2022-05-25 10:18:37 For CNN to restore its reputation as a journalistic news outlet, its teams must include at least one reporter who, at least once in his/her career, has unveiled, at least one human aspect of the 8 million human beings that make up Israeli society. @Ostrov_A @noatishby @mishtal https://t.co/Wp6A5Whb1S

2022-05-25 10:16:55 @Ostrov_A When Al-Jazeera "can prove" things, we must summon the world's greatest logicians, lest the logic of right and wrong be in danger of extinction.

2022-05-25 10:04:27 Ruth Wisse, the only General who understands the war on US campuses: "Don't let the war of words ever be fought about Israel's nature, let it be fought about why you can't accept Israel." If only @Hillel and @ADL would listen. https://t.co/bF2PbwnXcF

2022-05-25 06:03:12 The reliability of a "reporter", especially in the Middle East, is measured by the ratio of facts to accusations in the writings of that "reporter". Those interviewed by CNN barely reach a 1:9 ratio each. A combined report may obtain a higher level of plausibility, if those1/2 https://t.co/u32XnD4QaY

2022-05-24 20:07:13 I'm reminded that I'll be giving a talk on Friday on"What is Causal Inference and Where is Data Science Going"https://t.co/OusBu6medRIf you know the answer to one of the questions, the answer to the second is not far off.

2022-05-24 00:59:15 I bet if you ask an @AlJazeera reporter she won't know either. https://t.co/QTqJEcjQTr

2022-05-24 00:19:03 @firoozye I can tell you haven't read my son's stories from Iran, Pakistan, and North Africa. There is a different kind of journalism in the world, and @AlJazeera is its antonym.

2022-05-23 23:36:22 @wayneholmes @AJEnglish @ajimran For @AJEnglish to be taken seriously, they would have to raise the ratio of facts to accusations from 1:9 to 9:1. For a rational listener not to revolt the stench of accusations takes a real stomach. My admiration.

2022-05-23 21:21:03 @firoozye We have dealt with the incidents surrounding her tragic death before, e.g., https://t.co/3IY7ZTTRQ4A totally separate issue is whether she was a journalist or a propagandist, and I have offered a universal litmus test. Can we agree with the test?

2022-05-23 01:02:00 To readers requesting good books on the history of the Yishuv, I am recommending Anita Shapira "Israel: A history" (2012, Brandeis U.)https://t.co/U3vX4vq3N9Shapira is a first class, authoritative historian, who understands the pulse of history through its sources. @noatishby https://t.co/dGvRFntJB3

2022-05-22 22:53:11 Speaking about Fisher and Wright, I chose the latter as my hero. Whereas Fisher talks about "causes," it was Wright who realized that Science had been negligent in denying this beast a mathematical symbol, and gathered the courage to create one (eg., an arrow) when none existed. https://t.co/L5TqFBhHjF

2022-05-22 20:32:52 It's hard to summarize the week that was better than @noatishby, except perhaps to explicitly mention @AOC as an icon of the hate mob. Why her? Because our tax dollars pay for her podium. https://t.co/3IY7ZUbsHC

2022-05-22 20:03:43 There is a tiny difference. Nazi Germany did not say: "Some Jews are good Jews". The new haters are more sneaky, hence much more dangerous. They say or imply: "Nor all Jews are sub-humans, only Israel supporters." Even @AOC has learned the mantra: "My Jewish brothers and sisters" https://t.co/AKcpqeke1i

2022-05-22 18:30:08 @instrumenthull @chdausgaard @analisereal I couldn't understand the setup. Must wait for @analisereal to translate to structural language.Glad we still have translators among the living.

2022-05-22 18:05:37 Anyone knows of a serious study to assess whether the "ingrained habits of successful scientists" have become more "ingrained" in the age of internet and social media? https://t.co/XftDbeQXyD

2022-05-22 17:55:51 @LomaahhMore @oac Thanks for the correction.

2022-05-22 17:50:10 @chdausgaard I'm not familiar with SSIV. If it has resemblance to front-door, it should be expressible in graphs

2022-05-22 10:55:20 @LomaahhMore It's a mistake of many Westerners. They assume that Al-Jazeera venomous style turns readers off or, at least, will make readers question their reliability. It doesn't. Readers, even Westerners drink that poison like honey, and become like @OAC, true believers in Israel's evil.

2022-05-20 08:11:00 CAFIAC FIX

2022-10-29 08:42:41 @AliAbunimah, the founder of Electronic Intifada sides with Putin. Not surprising. The Russians and the Palestinians are the only two peoples who decided that their neighbors do not have the right to exist. Both fight for ONE-STATE solutions, and deem co-existence as injustice. https://t.co/5VM5PYi3kI

2022-10-29 07:25:38 The only explanation I have why @EinatWilf's video https://t.co/XSVDjqgHlL does not garner >

2022-10-29 01:28:35 @RaulMachadoG @soboleffspaces @AlfredoMorabia The paper is under a pay-wall but, if you think it oughtt to be part of the discussion, please describe Miettinen's contribution in modern language? I can't penetrate it on my own.

2022-10-29 00:28:54 @KordingLab @metrics52 I am just trying to understand your setup. If it is just drawing samples (y,x) from the P(y|do(x)) distribution, then obviously ML would take you from finite sample to the distribution. To comment, I need input-output - a terrible weakness I acquires in my Engineering schooling.

2022-10-29 00:18:18 @AlfredoMorabia @soboleffspaces @ken_rothman I believe I've read this article at the time, but it confirmed the perspective I had while writing Causality, in 2000. Do you think I've missed important conceptual milestones in my recent review: https://t.co/GIaEtSH0y3, ???

2022-10-28 23:44:13 @KordingLab @metrics52 Only one input and one output? How about effect of drugs on the weather, or education, or seatbelt usage? Is there a place for piece of knowledge: "drugs do not change a person's sex" ??

2022-10-28 23:36:43 @totteh Hilarious! If it were not serious. This means that readers on our channel would have to start faking it and talk like PO parishioners, eg, "treatment assignments" "target trials" "death to Rung-3" etc. I feel manipulated already.

2022-10-28 23:23:42 @DKedmey Please do nominate me to Twitter's "content moderation council" but, warning, my first recommendation: Zionophobic content is moderated same as "child pornography"

2022-10-28 22:51:38 @KordingLab @metrics52 I don't care either whether it's technically correct but, to anchor my thoughts, I need to know what "prior causal knowledge" is, how it is encoded, what data is taken, etc. tec.To anchor our thoughts, can we talk firing-squad?(I forgot my transistor circuits class).

2022-10-28 22:14:48 @KordingLab @metrics52 For me to "think", I need an input-output description, hopefully in SMC language. What is assumed, what kind of data is available, what is the output? Eg. Can we deduce the firing squad model by observing a zillion executions?

2022-10-28 15:46:11 @AlfredoMorabia @soboleffspaces At least no one was able to tell me what it is - perhaps you can? BTW, what was "the standard definition of confounding" in Epi prior to 1986 ? I hope it was not statistical, because there isn't any, see Sec. 2 of https://t.co/7lUwca9KP02/2

2022-10-28 15:33:28 @AlfredoMorabia @soboleffspaces Hi @AlfredoMorabia, Boris was referring to your upcoming talk at Harvard, celebrating Miettinen's paper of 1976. So, I asked: "Have I missed Miettinen's contributions when I wrote my historical account of causation in epidemiology?"So far, I do not see that I've missed any1/2

2022-10-28 15:09:51 American friends often ask me: What is it about Israel/Palestine that you, Israelis, think you know and that we don't. I cannot think of a better answer than this video clip of @EinatWilf https://t.co/XSVDjqgHlLPlease don't miss a word, even if you think you know and we don't.

2022-10-28 13:36:26 Dear @FranceskAlbs Those who deny Israel's right to Tel-Aviv, are hardly in position to lecture us how she can defend herself in Ramallah. https://t.co/VBKzJq9txN

2022-10-28 12:49:03 Check out this article: https://t.co/DLV49zoUZd

2022-10-28 09:47:35 @kanyewest has given bad name to antisemitism through his incoherent and blunt racism. Much more dangerous are the soft spoken "apartheid hyenas" at the UN, whom many people, mostly the gullible and uneducated, tend to parrot. https://t.co/w5p9GDhSdV

2022-10-28 09:20:53 @soboleffspaces @AlfredoMorabia My question to those who read Miettinen: Did he state conditions for the set Z to make the adjustment formula correct, or merely wrote down the formula by intuition. I haven't read him, assuming (perhaps prematurely) that, lacking notation, he did the latter only. Did I err?

2022-10-28 05:26:45 @soboleffspaces @AlfredoMorabia My question remains, if you have no notation for causal effects, no do(x), no Y_x, we are back in Yule's days

2022-10-28 03:18:49 Curious. How did Miettinen handle stratification without causal notation? I have credited the scientification of epidemiology to Greenland and Robins (1986), https://t.co/NMja3mjhtK,seeing that Miettinen tried to define confounding using probabilities. @AlfredoMorabia https://t.co/Yn5mCilXmR

2022-10-28 02:53:01 We need a stronger word, perhaps "Jew-cleansing" to describe "Jew washers" like Mehdi Hasan, who get terribly indignant when people point them to this glaring hypocrisy. https://t.co/TOx6NEdi1W

2022-10-28 02:34:02 @AleksanderMolak @rasbt However, I would not call it "implicit", because in my mind I have a clear arrow going between X=water and Y=tree-growth. I therefore conjecture that we extract causal knowledge from language by filling in the missing parts of a preprogrammed template (eg, DAG) = curiosity.2/2

2022-10-28 02:25:36 @AleksanderMolak @rasbt Most of my causal knowledge came from language. My father taught me why we need to water the orange trees, and how often. My Dr. told me to exercise or else. There is no reason therefore that LLM will not obtain causal knowledge from text sprinkled with causal utterances.1/2

2022-10-27 23:37:48 @itamarcaspi @DuduLagziel @giladmp I would also recommend the paper: "Understanding Simpson's Paradox" https://t.co/GpSKmVcKR7which goes beyond #Bookofwhy, and resolves the "paradox" from both decision making and cognitive perspectives.

2022-10-27 20:42:13 @xiaodai10838475 The word "biased" is a compliment to a Zionophobe.

2022-10-27 11:39:06 No Mehdi, it's got nothing to do with your being or not being an antisemite. No, you are a carrier of a much uglier affliction: Zionophobia - a moral deformity that cannot be blamed on alcoholic parents or abusive upbringing

2022-10-27 10:02:44 Glad this Tweet was re-liked, for it has become timely again in light or the new @ylecun - @GaryMarcus debate. I'm not sure though whether this definition of "Deep Understanding" would meet universal acceptance in "Deep Learning" land. https://t.co/Ik0Q4pi8Ie

2022-10-27 08:53:38 I don't know about you, but I get emotional watching high profile meetings such as this: https://t.co/w6v9QXUQuH. Just watching two great nations rising above the cynics and re-asserting their unique greatness makes my heart melt.

2022-10-27 08:21:29 @arieljalali I am not taking position here on whether hate mongering or child pornography should or shouldn't be allowed on Twitter and other social media channels. I am merely requesting that Mehdi's Tweets on anti-Semitism should be given equal treatment.

2022-10-27 07:03:05 @blakeflayton Once you baptize yourself with the holy waterof anti-Zionism, people might begin to take you seriously, even when you have nothing to say.

2022-10-24 05:14:40 @soboleffspaces @ObservStudies Your take-aways are fine for all practical purposes but, philosophically, @HarvardEpi would fight you tooth and nail when you remind them that the do-operator operates on a model of reality (SCM) while the g-formula emerges from a mysterious FFRCISTG =?? @EpiEllie #Bookofwhy

2022-10-24 04:45:19 @policytensor @bpetershome OK, peace on earth, but please warn me next time that your "regression" comes from you.

2022-10-24 04:43:07 @policytensor @bpetershome I remember doing tensor analysis in grad school and I bow to Einstein's theory of coordinate independence, but causal models do depend on choice of variables. Here I show how the direction of statistical time is reversed by choice of coordinates: https://t.co/Ty53coWZrI (11.2)

2022-10-24 04:30:19 @policytensor @bpetershome I was misled by the term, sorry. This is what happens to you after fighting 3 decades against the regression/structural confusion in the econometric (and statistical) literature. (eg, https://t.co/CEpVusa5eB) I would using suggest the term "linear SCM", to avoid confusions.

2022-10-24 04:24:00 @calimagna @noahholl @FrankRHutter I don't see why we need to guess about the gain that SCM gives them when we have given them a simple tool to diagnose and pinpoint the source of their gain: Run a toy problem, and see.

2022-10-24 04:14:55 @policytensor @bpetershome I follows your arguments, but not the language. In my mother tongue, regression is a statistical notion having nothing to do with the causal structure. Similarly, I do not see where "tensors" come in. Could be y weakness.

2022-10-23 19:27:27 @calimagna @noahholl @FrankRHutter This is what we are trying to understand on a simple 4-variable toy problem. Take a correct SCM and a fake one (of same size) that is equally compatible with the observed distribution. Would the former lead to better predictions?

2022-10-23 18:50:44 @noahholl @FrankRHutter @calimagna Here comes the beauty of toy-problems

2022-10-23 12:41:53 @Ostrov_A @StandWithUs @SAFECUNY SHSHSHSHSH!!!! Don't tell the @nytimes!!!

2022-10-23 12:05:47 A new film is out, "Why Ukraine", by my friend and mentor Bernard-Henri Levy @BHL. Don't miss! Oct. 27. At the UN. https://t.co/Qj5FBJ0Nsc

2022-10-23 10:35:58 @StandWithUs @SAFECUNY The looks of it warns me - it was probably stolen from Palestinians

2022-10-23 07:40:19 @mhd01 @AsraNomani @BloodBrief @fcpsnews Thanks. I hope I haven't used it innocently, as an endearing adjective, on anyone.

2022-10-23 07:32:39 Great to get cultural content on Twitter once in a while - stay tuned. https://t.co/14JyB7FVIS

2022-10-23 07:02:06 RT @elderofziyon: It might be time to start working on my book of cartoons. https://t.co/wspX6qAtXy

2022-10-23 06:59:09 Recalling personal experience, Reza Aslan is not a person I would trust. Not surprised he is now working for the Iranian regime. https://t.co/3o9G6gRTLu

2022-10-23 06:52:08 Should be in the tool-kit of every instructor. https://t.co/2ubgENwJ1x

2022-10-23 06:49:11 @AsraNomani @BloodBrief @fcpsnews I'm 62 years in America and can't figure out what the R-word is. Please educate a fellow immigrant.

2022-10-23 05:17:45 @erikbiz @SAFECUNY @AdamSandler @DavidSuissaJJ Agree, honest and simple. Which makes you think: Shat does it take for celebrities like @BarbraStreisand or @SaraSilverman to be honest and simple and come out of their silent-Zionist closets.

2022-10-23 02:21:53 @FJnyc @jjz1600 @arnoldroth The Mongol raiders, so I am told, complained bitterly when the Great China Wall was erected: Apartheid! Apartheid! They shouted. "If this ain't apartheid what is?" Western analysts echoed, and echoed, and echoed.

2022-10-23 01:31:47 @skdh I've noticed a sense of comradeship in your tweets. They will not replace us.

2022-10-22 22:48:52 @mvanhout71 @wv012 "Human anatomy" in healthcare translates into "intrinsic properties of target population" in social science, invariant across backgrounds and aspirations of participants.

2022-10-22 21:33:21 @FrankRHutter Speaking basic principles of "rung 1.5", why should causal information help us improve predictions? Once we understand the principle, we should be able to quantify how sensitive prediction quality is to causal misspecifications.@calimagna

2022-10-22 21:13:46 @mvanhout71 @wv012 My Dutch is a bit rusty, but I hope you touched on these considerations: https://t.co/A2MguL54zE

2022-10-22 20:22:10 @wv012 Off hand, not getting into the details of the habits, my answer would be YES. Why? Because associations are less stable than causal relations

2022-10-22 10:05:01 Jewishness is based on "Amech Ami, Ve'Elohaich Elohay" ("Your people are my people and your God is my God" - Ruth to Naomi). Meaning: "Peoplehood first - Religion second". https://t.co/2P8y0tfKVF

2022-10-22 09:55:39 @AsraNomani @BOSCityCouncil @Quillette What an Orwellian inversion. #MahsaAmini must be turning in her grave to see the hijab, which to her (and to majority of Iranian women) meant a symbol of women subjugation celebrated as a symbol of free expression in a US city called Boston, home to MIT and Harvard.

2022-10-22 09:40:31 @HenMazzig There is one logical flaw in your argument, @HenMazzig, the word "share" does not exist in Palestinian vocabulary. When it comes to Israel, "share" is replaced by "stolen", and "I choose" is now "I'm a victim".

2022-10-22 09:32:07 @Eve_Barlow @ArielElyseGold Inspired by @Eve_Barlow https://t.co/w5kr5QfDFN

2022-10-22 08:51:08 @AleksanderMolak @rasbt I am trying to understand what you mean by "implicit heuristic world models". If this is something that is learned by the agent, then it must be stored someplace, in some code. That code then is explicit. What makes it different from "explicit world model"???

2022-10-22 08:27:52 I was about to swear this is a joke ridiculing our "apartheid parrots". It isn't! Seeing @ArielElyseGold, I reckon these parrots still exist, seriously squeaking and chirping: "Me too, Me too, Dying to be original, dying to be provocative, Me apartheid parrot too, Me too!" https://t.co/V8QfAhBqGD

2022-10-22 07:34:13 RT @richard_landes: https://t.co/LWy0ssOfOL

2022-10-22 07:31:11 @CitalanEdward @WealthInc247 It was my first exposure to the philosophy of science, dissolving my guilt for not understanding speculative philosophers like Kant and Hegel, and leading me to Carnap, Hampel , Goodman and Popper.

2022-10-22 04:53:28 @tbednall @WealthInc247 It changed my life too, forcing me to re-examine long-held assumptions from a totally new perspective.

2022-10-22 04:50:12 @WealthInc247 Hans Reichenbach "The Rise of Scientific Philosophy".

2022-10-22 03:59:00 In Rubin's framework, ignorability conditions are assumed not because they are deemed true, but because they are needed to support a claim

2022-10-22 03:58:59 We used to have Rubin's disciples among our readers (eg, Yale economists), perhaps they can help us interpret the analogy. I assume he refers to the fact that SCMs appear to be redundant because they provide, not only ignorability conditions, but also their justifications. 1/3 https://t.co/sAbb8m9yVV

2022-10-21 17:33:21 @penguinvondoom @GaryMarcus @hajivat @nirsd Forgot what "cool RPG premise" stands for.

2022-10-21 10:16:29 @PHuenermund Agree with the insights. I will soon post a cleaned up copy, with unmolested references.

2022-10-21 10:14:02 @AvivaKlompas And when I point out that Zionophobes are suffering from a moral deformity, friends tell me "don't get down to their level".

2022-10-21 09:49:01 Today, all Palestinian factions have declared a "Day of Rage" in memory of Super-Martyr Uday Tamimi (22), who murdered Noa Lazar 11 days ago.I am used to Days of "Rage", but I can't get used to Noa's eyes and smile, which are haunting me for 11 days:https://t.co/qZEpzX6Kz3 https://t.co/LS1fvK4IaU

2022-10-21 09:27:17 @jeremy1174 I thought you understood what Zionism means: Coexistence as equally indigenous peoples. The disrespectful word "entity" tells me that you don't.

2022-10-21 09:14:51 @jeremy1174 Nothing would please me more than seeing our Palestinian neighbors greet this Ethiopian boy as a comrade and close kin. That is precisely what my grandfather expected from our cousins.

2022-10-21 08:56:41 The only thing missing from the video is whether this man listens to Al-Jazeera or Mehdi Hasan on@MSNBC https://t.co/U4lV6tKKi8

2022-10-21 08:48:55 Family lore has it that, when my grandfather arrived at Jaffa port, in 1924, he likewise fell down and kissed the ground. This makes me a comrade and close kin of this Ethiopian boy, transcending all age and background differences. https://t.co/pEOhTPTyWB

2022-10-21 08:41:39 @RaulMachadoG @soboleffspaces @ObservStudies The criterion: "50% chance that the claim is true," sounds statistical until we realize that the "claim" itself is causal, and renders statistics totally helpless when it comes to legal responsibility.

2022-10-21 08:27:56 @hajivat @GaryMarcus @nirsd The #Bookofwhy was translated to 14 languages, showing in chapter and verse how the belief that all information is in the data fails. Yet this has not prevented data-centric enthusiasts from erecting taller and taller cathedrals to a handcuffed God.

2022-10-21 01:51:00 @AviMayer No need to get upset, it's just a matter of semantics. When I first heard (in 2000) that the land my grandfather bought in Bnai Brak (1924) was "stolen", I realized that "stolen" in Arabic means "cultivated". Thus, Hadera swamps, Tel-Aviv, Falafel, Humus, Debka, are all "stolen".

2022-10-21 00:41:17 Great question. I did some thinking about it.See Section 2.8 on "Statistical Time" here: https://t.co/xfevdJms8r. Especially last paragraph. Also, Section 11.2 in https://t.co/Ty53coWZrI, on "Reversing Statistical Time". https://t.co/4A4d0QBrRe

2022-10-20 21:24:55 @AI_AM001 @isaacdecastrog To the best of my knowledge no one denies Iran the right to exist. Enjoy.

2022-10-20 18:52:12 @soboleffspaces @ObservStudies Anyone who rejects the First Law would naturally see "causes of effect" to be a "cocktail conversation", including legal responsibility and scientific explanations (moon and tides). How else can you define "but for"?. I bet Heckman would think likewise, rejecting the First Law.

2022-10-20 11:57:20 @AleksanderMolak @rasbt That's why I prefer toy problems

2022-10-20 11:15:39 Journalism Award Stripped from Palestinian Journalist Over Antisemitic Facebook Posts https://t.co/d95kdYYiIM

2022-10-20 10:29:22 @isaacdecastrog Some "writers" think that their arguments become convincing if they sprinkle foul adjectives around, marinated in the cesspool of Al-Jazeera. Surprisingly, some of their readers love that stench.

2022-10-20 10:18:41 My take-aways from Rubin's and Heckman's interviews @observStudies match precisely what I anticipated, as expressed in my interview, even though it was completed 2 years ago, and we were not permitted to see each other's texts. https://t.co/N8vmMp6Mxo

2022-10-20 08:08:24 @Claire_V0ltaire We have also seen Jewish scholars expressing their irritation, some of them inadequately:https://t.co/HkeihG1qBd@michaeldickson @Ostrov_A @EinatWilf @CotlerWunsh @SSI_Movement @BarbraStreisand @SSIUCLA @vardi @CAMERAonCampus @bariweiss @JGreenblattADL @deborahlipstadt

2022-10-20 07:10:48 @AleksanderMolak @rasbt If it's a "trick" then I dare to bet that it would fail on the 3 toy examples in Fig. 1 of https://t.co/dEPwcuLbPS What gives me the audacity to bet w/o even understanding the trick? Simple. The 3 examples generate the same data, while the correct answers are different. Smart!

2022-10-20 06:12:10 @HenMazzig Would you object if I were to replace "antisemitism" with "anti-Zionism", and say out laud: Anti-Zionism isn’t okay, period.You see, Hitler gave a bad name to antisemitism, so people (including antisemites) would rush to join your statement

2022-10-20 05:18:36 I beg to respectfully disagree with Waxman's 2nd interpretation of Trump's statement, the "Inversion of Dual Loyalty", according to which saying that American Jews SHOULD care about Israel amounts to making "antisemitic assumptions". Americans are mature enough to understand 1/3 https://t.co/IkUGrBlvkL

2022-10-20 02:57:36 @rasbt I havn't heard of the "hard, practical truth which consists of retraining one or more layers on a different dataset." Is this "truth" or "trick"? Do we have a theory to guarantee the outcome? Better yet, can we try it on the 3 toy problems inFig. 1 of https://t.co/dEPwcuLbPS ??

2022-10-19 22:06:45 @KirkDBorne I was not familiar with Barber's book, fairly impressive!And I truly love the quote from https://t.co/YmnsZPM3tL"Pearl believes that current AI is mostly just glorified curve fitting, and that adding causal intelligence to current AI will be necessary ...".Amen!

2022-10-19 21:25:58 @AmandaKMontoya @b_snef @rlmcelreath So, UCLA Psyc dpt has a future after all - best news I got from campus in months. Hoping our paths cross.

2022-10-19 18:56:20 @rasbt To the best of my understanding, "transfer learning" is not an "approach" but a goal, a desideratum. Also to my understanding, causal inference is the only approach that achieves this goal (via selection diagrams,) see https://t.co/KBGKcmJxM8, https://t.co/E1ap1N4ZCh

2022-10-30 01:33:32 @soboleffspaces @RaulMachadoG @AlfredoMorabia I am willing to modify my summary sentence. Would you be happy with: "Here, for the first time, Miettinen is introducing the concept of 'blocking', and Epi has not been the same since

2022-10-29 23:37:09 @KordingLab @metrics52 I don't see where the generalization comes in. Is it that we have SEVERAL problems, P_1, P_2...? or is it that we have both P(y|do(x)) and P(y|x) ??Do we need more than one DAG to explain what your method is doing? Recall, one DAG can represent many problems and many realities.

2022-10-29 22:20:43 @RaulMachadoG @soboleffspaces @AlfredoMorabia Thanks, @raulMachadoG and others for sending me links to 1970's papers. But I was waiting for a summary sentence: e.g., "Here, for the first time, Miettinen is introducing the concept of xyz to epidemiology, and Epi has not been the same since

2022-10-29 18:44:21 @a_sarentorbic @bsgallagher @GaryMarcus Such characterization makes me suspicious that you haven't read #Bookofwhy.

2022-10-29 17:48:09 @GaryMarcus A cause-effect question. Do we appreciate Kant because we are here, or should we read him to get further ahead of where we are?

2022-10-29 17:30:09 @anndvision The word for this is "logic". Formal logic does not tell us what the world looks like, it merely tells us that, if you make premises 1 and 2 about the world, and observe data D, then conclusion C is inevitable. In our case the, premises are causal, rather than propositional.

2022-10-29 11:17:16 @AnnGreenberg @noatishby But that has not prevented campus climate from becoming what it is, and has not equipped students with the weapons to tell their bullies: "You sound like a Zionophobic racist! Don't tell me you are one!" Bullies smell weakness.

2022-10-29 10:28:31 Something is missing. These ignorant students who are regurgitating populist slogans at @noatishby do not even begin to suspect that they are engaged in less than super-righteous activity, let alone a racist movement with genocidal goals. The word "Zionophobia" is badly missing. https://t.co/JxldS1coQu

2022-10-29 10:13:30 RT @RabbiWolpe: In the ‘box score’ dept… https://t.co/bSffUZ7WdM

2022-10-29 09:55:22 OTD. On October 29, 1969, 53 years ago, the internet was born at UCLA. https://t.co/ylPiShf4G8I was hired only 3 weeks earlier, and I didn't quite understand what the people on the floor below me were celebrating. Now I understand.

2022-10-29 09:21:01 @AlfredoMorabia @soboleffspaces So, can I assume that my account of causation in epidemiology, narrated here: https://t.co/GIaEtSH0y3, is not missing a major conceptual breakthrough, although I skipped many prominent names. (I may still have a chance to add, if convinced otherwise.)

2022-10-29 08:42:41 @AliAbunimah, the founder of Electronic Intifada sides with Putin. Not surprising. The Russians and the Palestinians are the only two peoples who decided that their neighbors do not have the right to exist. Both fight for ONE-STATE solutions, and deem co-existence as injustice. https://t.co/5VM5PYi3kI

2022-10-29 07:25:38 The only explanation I have why @EinatWilf's video https://t.co/XSVDjqgHlL does not garner >

2022-10-29 01:28:35 @RaulMachadoG @soboleffspaces @AlfredoMorabia The paper is under a pay-wall but, if you think it oughtt to be part of the discussion, please describe Miettinen's contribution in modern language? I can't penetrate it on my own.

2022-10-29 00:28:54 @KordingLab @metrics52 I am just trying to understand your setup. If it is just drawing samples (y,x) from the P(y|do(x)) distribution, then obviously ML would take you from finite sample to the distribution. To comment, I need input-output - a terrible weakness I acquires in my Engineering schooling.

2022-10-29 00:18:18 @AlfredoMorabia @soboleffspaces @ken_rothman I believe I've read this article at the time, but it confirmed the perspective I had while writing Causality, in 2000. Do you think I've missed important conceptual milestones in my recent review: https://t.co/GIaEtSH0y3, ???

2022-10-28 23:44:13 @KordingLab @metrics52 Only one input and one output? How about effect of drugs on the weather, or education, or seatbelt usage? Is there a place for piece of knowledge: "drugs do not change a person's sex" ??

2022-10-28 23:36:43 @totteh Hilarious! If it were not serious. This means that readers on our channel would have to start faking it and talk like PO parishioners, eg, "treatment assignments" "target trials" "death to Rung-3" etc. I feel manipulated already.

2022-10-28 23:23:42 @DKedmey Please do nominate me to Twitter's "content moderation council" but, warning, my first recommendation: Zionophobic content is moderated same as "child pornography"

2022-10-28 22:51:38 @KordingLab @metrics52 I don't care either whether it's technically correct but, to anchor my thoughts, I need to know what "prior causal knowledge" is, how it is encoded, what data is taken, etc. tec.To anchor our thoughts, can we talk firing-squad?(I forgot my transistor circuits class).

2022-10-28 22:14:48 @KordingLab @metrics52 For me to "think", I need an input-output description, hopefully in SMC language. What is assumed, what kind of data is available, what is the output? Eg. Can we deduce the firing squad model by observing a zillion executions?

2022-10-28 15:46:11 @AlfredoMorabia @soboleffspaces At least no one was able to tell me what it is - perhaps you can? BTW, what was "the standard definition of confounding" in Epi prior to 1986 ? I hope it was not statistical, because there isn't any, see Sec. 2 of https://t.co/7lUwca9KP02/2

2022-10-28 15:33:28 @AlfredoMorabia @soboleffspaces Hi @AlfredoMorabia, Boris was referring to your upcoming talk at Harvard, celebrating Miettinen's paper of 1976. So, I asked: "Have I missed Miettinen's contributions when I wrote my historical account of causation in epidemiology?"So far, I do not see that I've missed any1/2

2022-10-28 15:09:51 American friends often ask me: What is it about Israel/Palestine that you, Israelis, think you know and that we don't. I cannot think of a better answer than this video clip of @EinatWilf https://t.co/XSVDjqgHlLPlease don't miss a word, even if you think you know and we don't.

2022-10-28 13:36:26 Dear @FranceskAlbs Those who deny Israel's right to Tel-Aviv, are hardly in position to lecture us how she can defend herself in Ramallah. https://t.co/VBKzJq9txN

2022-10-28 12:49:03 Check out this article: https://t.co/DLV49zoUZd

2022-10-28 09:47:35 @kanyewest has given bad name to antisemitism through his incoherent and blunt racism. Much more dangerous are the soft spoken "apartheid hyenas" at the UN, whom many people, mostly the gullible and uneducated, tend to parrot. https://t.co/w5p9GDhSdV

2022-10-28 09:20:53 @soboleffspaces @AlfredoMorabia My question to those who read Miettinen: Did he state conditions for the set Z to make the adjustment formula correct, or merely wrote down the formula by intuition. I haven't read him, assuming (perhaps prematurely) that, lacking notation, he did the latter only. Did I err?

2022-10-28 05:26:45 @soboleffspaces @AlfredoMorabia My question remains, if you have no notation for causal effects, no do(x), no Y_x, we are back in Yule's days

2022-10-28 03:18:49 Curious. How did Miettinen handle stratification without causal notation? I have credited the scientification of epidemiology to Greenland and Robins (1986), https://t.co/NMja3mjhtK,seeing that Miettinen tried to define confounding using probabilities. @AlfredoMorabia https://t.co/Yn5mCilXmR

2022-10-28 02:53:01 We need a stronger word, perhaps "Jew-cleansing" to describe "Jew washers" like Mehdi Hasan, who get terribly indignant when people point them to this glaring hypocrisy. https://t.co/TOx6NEdi1W

2022-10-28 02:34:02 @AleksanderMolak @rasbt However, I would not call it "implicit", because in my mind I have a clear arrow going between X=water and Y=tree-growth. I therefore conjecture that we extract causal knowledge from language by filling in the missing parts of a preprogrammed template (eg, DAG) = curiosity.2/2

2022-10-28 02:25:36 @AleksanderMolak @rasbt Most of my causal knowledge came from language. My father taught me why we need to water the orange trees, and how often. My Dr. told me to exercise or else. There is no reason therefore that LLM will not obtain causal knowledge from text sprinkled with causal utterances.1/2

2022-10-27 23:37:48 @itamarcaspi @DuduLagziel @giladmp I would also recommend the paper: "Understanding Simpson's Paradox" https://t.co/GpSKmVcKR7which goes beyond #Bookofwhy, and resolves the "paradox" from both decision making and cognitive perspectives.

2022-10-27 20:42:13 @xiaodai10838475 The word "biased" is a compliment to a Zionophobe.

2022-10-27 11:39:06 No Mehdi, it's got nothing to do with your being or not being an antisemite. No, you are a carrier of a much uglier affliction: Zionophobia - a moral deformity that cannot be blamed on alcoholic parents or abusive upbringing

2022-10-27 10:02:44 Glad this Tweet was re-liked, for it has become timely again in light or the new @ylecun - @GaryMarcus debate. I'm not sure though whether this definition of "Deep Understanding" would meet universal acceptance in "Deep Learning" land. https://t.co/Ik0Q4pi8Ie

2022-10-27 08:53:38 I don't know about you, but I get emotional watching high profile meetings such as this: https://t.co/w6v9QXUQuH. Just watching two great nations rising above the cynics and re-asserting their unique greatness makes my heart melt.

2022-10-27 08:21:29 @arieljalali I am not taking position here on whether hate mongering or child pornography should or shouldn't be allowed on Twitter and other social media channels. I am merely requesting that Mehdi's Tweets on anti-Semitism should be given equal treatment.

2022-10-27 07:03:05 @blakeflayton Once you baptize yourself with the holy waterof anti-Zionism, people might begin to take you seriously, even when you have nothing to say.

2022-10-24 05:14:40 @soboleffspaces @ObservStudies Your take-aways are fine for all practical purposes but, philosophically, @HarvardEpi would fight you tooth and nail when you remind them that the do-operator operates on a model of reality (SCM) while the g-formula emerges from a mysterious FFRCISTG =?? @EpiEllie #Bookofwhy

2022-10-24 04:45:19 @policytensor @bpetershome OK, peace on earth, but please warn me next time that your "regression" comes from you.

2022-10-24 04:43:07 @policytensor @bpetershome I remember doing tensor analysis in grad school and I bow to Einstein's theory of coordinate independence, but causal models do depend on choice of variables. Here I show how the direction of statistical time is reversed by choice of coordinates: https://t.co/Ty53coWZrI (11.2)

2022-10-24 04:30:19 @policytensor @bpetershome I was misled by the term, sorry. This is what happens to you after fighting 3 decades against the regression/structural confusion in the econometric (and statistical) literature. (eg, https://t.co/CEpVusa5eB) I would using suggest the term "linear SCM", to avoid confusions.

2022-10-24 04:24:00 @calimagna @noahholl @FrankRHutter I don't see why we need to guess about the gain that SCM gives them when we have given them a simple tool to diagnose and pinpoint the source of their gain: Run a toy problem, and see.

2022-10-24 04:14:55 @policytensor @bpetershome I follows your arguments, but not the language. In my mother tongue, regression is a statistical notion having nothing to do with the causal structure. Similarly, I do not see where "tensors" come in. Could be y weakness.

2022-10-23 19:27:27 @calimagna @noahholl @FrankRHutter This is what we are trying to understand on a simple 4-variable toy problem. Take a correct SCM and a fake one (of same size) that is equally compatible with the observed distribution. Would the former lead to better predictions?

2022-10-23 18:50:44 @noahholl @FrankRHutter @calimagna Here comes the beauty of toy-problems

2022-10-23 12:41:53 @Ostrov_A @StandWithUs @SAFECUNY SHSHSHSHSH!!!! Don't tell the @nytimes!!!

2022-10-23 12:05:47 A new film is out, "Why Ukraine", by my friend and mentor Bernard-Henri Levy @BHL. Don't miss! Oct. 27. At the UN. https://t.co/Qj5FBJ0Nsc

2022-10-23 10:35:58 @StandWithUs @SAFECUNY The looks of it warns me - it was probably stolen from Palestinians

2022-10-23 07:40:19 @mhd01 @AsraNomani @BloodBrief @fcpsnews Thanks. I hope I haven't used it innocently, as an endearing adjective, on anyone.

2022-10-23 07:32:39 Great to get cultural content on Twitter once in a while - stay tuned. https://t.co/14JyB7FVIS

2022-10-23 07:02:06 RT @elderofziyon: It might be time to start working on my book of cartoons. https://t.co/wspX6qAtXy

2022-10-23 06:59:09 Recalling personal experience, Reza Aslan is not a person I would trust. Not surprised he is now working for the Iranian regime. https://t.co/3o9G6gRTLu

2022-10-23 06:52:08 Should be in the tool-kit of every instructor. https://t.co/2ubgENwJ1x

2022-10-23 06:49:11 @AsraNomani @BloodBrief @fcpsnews I'm 62 years in America and can't figure out what the R-word is. Please educate a fellow immigrant.

2022-10-23 05:17:45 @erikbiz @SAFECUNY @AdamSandler @DavidSuissaJJ Agree, honest and simple. Which makes you think: Shat does it take for celebrities like @BarbraStreisand or @SaraSilverman to be honest and simple and come out of their silent-Zionist closets.

2022-10-23 02:21:53 @FJnyc @jjz1600 @arnoldroth The Mongol raiders, so I am told, complained bitterly when the Great China Wall was erected: Apartheid! Apartheid! They shouted. "If this ain't apartheid what is?" Western analysts echoed, and echoed, and echoed.

2022-10-23 01:31:47 @skdh I've noticed a sense of comradeship in your tweets. They will not replace us.

2022-10-22 22:48:52 @mvanhout71 @wv012 "Human anatomy" in healthcare translates into "intrinsic properties of target population" in social science, invariant across backgrounds and aspirations of participants.

2022-10-22 21:33:21 @FrankRHutter Speaking basic principles of "rung 1.5", why should causal information help us improve predictions? Once we understand the principle, we should be able to quantify how sensitive prediction quality is to causal misspecifications.@calimagna

2022-10-22 21:13:46 @mvanhout71 @wv012 My Dutch is a bit rusty, but I hope you touched on these considerations: https://t.co/A2MguL54zE

2022-10-22 20:22:10 @wv012 Off hand, not getting into the details of the habits, my answer would be YES. Why? Because associations are less stable than causal relations

2022-10-22 10:05:01 Jewishness is based on "Amech Ami, Ve'Elohaich Elohay" ("Your people are my people and your God is my God" - Ruth to Naomi). Meaning: "Peoplehood first - Religion second". https://t.co/2P8y0tfKVF

2022-10-22 09:55:39 @AsraNomani @BOSCityCouncil @Quillette What an Orwellian inversion. #MahsaAmini must be turning in her grave to see the hijab, which to her (and to majority of Iranian women) meant a symbol of women subjugation celebrated as a symbol of free expression in a US city called Boston, home to MIT and Harvard.

2022-10-22 09:40:31 @HenMazzig There is one logical flaw in your argument, @HenMazzig, the word "share" does not exist in Palestinian vocabulary. When it comes to Israel, "share" is replaced by "stolen", and "I choose" is now "I'm a victim".

2022-10-22 09:32:07 @Eve_Barlow @ArielElyseGold Inspired by @Eve_Barlow https://t.co/w5kr5QfDFN

2022-10-22 08:51:08 @AleksanderMolak @rasbt I am trying to understand what you mean by "implicit heuristic world models". If this is something that is learned by the agent, then it must be stored someplace, in some code. That code then is explicit. What makes it different from "explicit world model"???

2022-10-22 08:27:52 I was about to swear this is a joke ridiculing our "apartheid parrots". It isn't! Seeing @ArielElyseGold, I reckon these parrots still exist, seriously squeaking and chirping: "Me too, Me too, Dying to be original, dying to be provocative, Me apartheid parrot too, Me too!" https://t.co/V8QfAhBqGD

2022-10-22 07:34:13 RT @richard_landes: https://t.co/LWy0ssOfOL

2022-10-22 07:31:11 @CitalanEdward @WealthInc247 It was my first exposure to the philosophy of science, dissolving my guilt for not understanding speculative philosophers like Kant and Hegel, and leading me to Carnap, Hampel , Goodman and Popper.

2022-10-22 04:53:28 @tbednall @WealthInc247 It changed my life too, forcing me to re-examine long-held assumptions from a totally new perspective.

2022-10-22 04:50:12 @WealthInc247 Hans Reichenbach "The Rise of Scientific Philosophy".

2022-10-22 03:59:00 In Rubin's framework, ignorability conditions are assumed not because they are deemed true, but because they are needed to support a claim

2022-10-22 03:58:59 We used to have Rubin's disciples among our readers (eg, Yale economists), perhaps they can help us interpret the analogy. I assume he refers to the fact that SCMs appear to be redundant because they provide, not only ignorability conditions, but also their justifications. 1/3 https://t.co/sAbb8m9yVV

2022-10-21 17:33:21 @penguinvondoom @GaryMarcus @hajivat @nirsd Forgot what "cool RPG premise" stands for.

2022-10-21 10:16:29 @PHuenermund Agree with the insights. I will soon post a cleaned up copy, with unmolested references.

2022-10-21 10:14:02 @AvivaKlompas And when I point out that Zionophobes are suffering from a moral deformity, friends tell me "don't get down to their level".

2022-10-21 09:49:01 Today, all Palestinian factions have declared a "Day of Rage" in memory of Super-Martyr Uday Tamimi (22), who murdered Noa Lazar 11 days ago.I am used to Days of "Rage", but I can't get used to Noa's eyes and smile, which are haunting me for 11 days:https://t.co/qZEpzX6Kz3 https://t.co/LS1fvK4IaU

2022-10-21 09:27:17 @jeremy1174 I thought you understood what Zionism means: Coexistence as equally indigenous peoples. The disrespectful word "entity" tells me that you don't.

2022-10-21 09:14:51 @jeremy1174 Nothing would please me more than seeing our Palestinian neighbors greet this Ethiopian boy as a comrade and close kin. That is precisely what my grandfather expected from our cousins.

2022-10-21 08:56:41 The only thing missing from the video is whether this man listens to Al-Jazeera or Mehdi Hasan on@MSNBC https://t.co/U4lV6tKKi8

2022-10-21 08:48:55 Family lore has it that, when my grandfather arrived at Jaffa port, in 1924, he likewise fell down and kissed the ground. This makes me a comrade and close kin of this Ethiopian boy, transcending all age and background differences. https://t.co/pEOhTPTyWB

2022-10-21 08:41:39 @RaulMachadoG @soboleffspaces @ObservStudies The criterion: "50% chance that the claim is true," sounds statistical until we realize that the "claim" itself is causal, and renders statistics totally helpless when it comes to legal responsibility.

2022-10-21 08:27:56 @hajivat @GaryMarcus @nirsd The #Bookofwhy was translated to 14 languages, showing in chapter and verse how the belief that all information is in the data fails. Yet this has not prevented data-centric enthusiasts from erecting taller and taller cathedrals to a handcuffed God.

2022-10-21 01:51:00 @AviMayer No need to get upset, it's just a matter of semantics. When I first heard (in 2000) that the land my grandfather bought in Bnai Brak (1924) was "stolen", I realized that "stolen" in Arabic means "cultivated". Thus, Hadera swamps, Tel-Aviv, Falafel, Humus, Debka, are all "stolen".

2022-10-21 00:41:17 Great question. I did some thinking about it.See Section 2.8 on "Statistical Time" here: https://t.co/xfevdJms8r. Especially last paragraph. Also, Section 11.2 in https://t.co/Ty53coWZrI, on "Reversing Statistical Time". https://t.co/4A4d0QBrRe

2022-10-20 21:24:55 @AI_AM001 @isaacdecastrog To the best of my knowledge no one denies Iran the right to exist. Enjoy.

2022-10-20 18:52:12 @soboleffspaces @ObservStudies Anyone who rejects the First Law would naturally see "causes of effect" to be a "cocktail conversation", including legal responsibility and scientific explanations (moon and tides). How else can you define "but for"?. I bet Heckman would think likewise, rejecting the First Law.

2022-10-20 11:57:20 @AleksanderMolak @rasbt That's why I prefer toy problems

2022-10-20 11:15:39 Journalism Award Stripped from Palestinian Journalist Over Antisemitic Facebook Posts https://t.co/d95kdYYiIM

2022-10-20 10:29:22 @isaacdecastrog Some "writers" think that their arguments become convincing if they sprinkle foul adjectives around, marinated in the cesspool of Al-Jazeera. Surprisingly, some of their readers love that stench.

2022-10-20 10:18:41 My take-aways from Rubin's and Heckman's interviews @observStudies match precisely what I anticipated, as expressed in my interview, even though it was completed 2 years ago, and we were not permitted to see each other's texts. https://t.co/N8vmMp6Mxo

2022-10-20 08:08:24 @Claire_V0ltaire We have also seen Jewish scholars expressing their irritation, some of them inadequately:https://t.co/HkeihG1qBd@michaeldickson @Ostrov_A @EinatWilf @CotlerWunsh @SSI_Movement @BarbraStreisand @SSIUCLA @vardi @CAMERAonCampus @bariweiss @JGreenblattADL @deborahlipstadt

2022-10-20 07:10:48 @AleksanderMolak @rasbt If it's a "trick" then I dare to bet that it would fail on the 3 toy examples in Fig. 1 of https://t.co/dEPwcuLbPS What gives me the audacity to bet w/o even understanding the trick? Simple. The 3 examples generate the same data, while the correct answers are different. Smart!

2022-10-20 06:12:10 @HenMazzig Would you object if I were to replace "antisemitism" with "anti-Zionism", and say out laud: Anti-Zionism isn’t okay, period.You see, Hitler gave a bad name to antisemitism, so people (including antisemites) would rush to join your statement

2022-10-20 05:18:36 I beg to respectfully disagree with Waxman's 2nd interpretation of Trump's statement, the "Inversion of Dual Loyalty", according to which saying that American Jews SHOULD care about Israel amounts to making "antisemitic assumptions". Americans are mature enough to understand 1/3 https://t.co/IkUGrBlvkL

2022-10-20 02:57:36 @rasbt I havn't heard of the "hard, practical truth which consists of retraining one or more layers on a different dataset." Is this "truth" or "trick"? Do we have a theory to guarantee the outcome? Better yet, can we try it on the 3 toy problems inFig. 1 of https://t.co/dEPwcuLbPS ??

2022-10-19 22:06:45 @KirkDBorne I was not familiar with Barber's book, fairly impressive!And I truly love the quote from https://t.co/YmnsZPM3tL"Pearl believes that current AI is mostly just glorified curve fitting, and that adding causal intelligence to current AI will be necessary ...".Amen!

2022-10-19 21:25:58 @AmandaKMontoya @b_snef @rlmcelreath So, UCLA Psyc dpt has a future after all - best news I got from campus in months. Hoping our paths cross.

2022-10-19 18:56:20 @rasbt To the best of my understanding, "transfer learning" is not an "approach" but a goal, a desideratum. Also to my understanding, causal inference is the only approach that achieves this goal (via selection diagrams,) see https://t.co/KBGKcmJxM8, https://t.co/E1ap1N4ZCh

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-11-18 23:48:12 @jeremy1174 @EinatWilf You can evaluate the mentality of Al Jazeera through your words and beliefs, according to which blood-thirsty IDF is shooting children for sport. And they have succeeded to convince even people of your intelligence that this is journalism.

2022-11-18 17:05:18 RT @AdrianRaftery1: Congratulations, Carlos. Great to see your whole dissertation available. (For some reason unsent at the time, but even…

2022-11-18 17:03:05 Police said “four shots were fired from a loaded weapon” into the exterior of the home next to the city’s Old Synagogue but that no one was injured. https://t.co/1gIjSV10jc

2022-11-18 16:55:18 @jeremy1174 @EinatWilf Often false and sometimes merely distorted. You would never see an Israeli child on Al-Jazeera -- where have all the mothers gone?

2022-11-18 05:19:36 @soboleffspaces Amazing, isn't it? Now imagine my child-like glee upon discovering it in 1999 (See Theorem 4 here: https://t.co/QQa4sdcJMT). I felt like I solved Abraham's problem "What if there were 50 righteous people in the city?"

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-11-20 03:23:16 We are in receipt of a link to an illuminating symposia of Anthroponeny: https://t.co/3pah2sxdvn About -1:52 into the video, we can find Terry Sejnowski's talk which asks whether ML, Neural Nets, and Large Language Models can understand "intelligence", "explainability", 1/3

2022-11-20 03:22:25 @jeremy1174 @EinatWilf If you do not Al Jazeera, where did you get this maligned understanding of the "Zionist plan." My understanding of the plan is "Co-existence" with only ONE condition: "Existence". This entails no "buckling under" ecept dismantling the current Palestinian plan of "Sole-existence".

2022-11-19 23:55:46 Honored to join and support Moment Magazine's gala https://t.co/WhSupVLYcS, noting that Moment's Daniel Pearl Investigative Journalism Initiative https://t.co/VtqCxSs4Zv is launching new projects. dpiji story "An Inconvenient Genocide" won two first place journalistic awards.

2022-11-19 23:15:31 This 4:1 ratio is at least reversed in all previous reports by campus watchers (i.e., #ADL, AJC, Hillel, etc). The ratio approaches ZERO in ALL reports issued by University administrations or EDI offices (including UCLA). Hail to statistics! One thing it's good for.

2022-11-19 23:15:30 A groundbreaking and well documented report on campus assaults on Jewish identity: https://t.co/yM6szdGRGs. How do I know it hits the right issues? Pure statistics: The word "anti-Zionism" appears 4 times more often than the word "antisemitism"

2022-11-19 22:40:15 For our German readers: A "must read" for everyone who is really interested in a solution to to the Israeli-Palestinian conflict. https://t.co/HUyYhZrEFY

2022-11-19 21:25:47 Well put! Plus note the strange correlation between those who oppose IHRA and those who deny Israel's right to exist. By endorsing the former, Stewart emboldens the latter. https://t.co/al9lxyyUzG

2022-11-19 21:11:12 @haig Show me one "peace agreement" in the world that was not signed out of some practical "incentives" to the sound of warnings: "don't be naïve!". On Nov. 19, 1977, 45 years ago today, President Sadat landed in Israel to the sound of similar warnings. Yet it's worked, &

2022-11-19 20:57:00 Will President Abbas ever rise to Sadat's height and address the Israeli Knesset with a message of peace? In 2008 I was so naïve as to write the speech for him, https://t.co/iTXOtbDXf7 a speech that would convince Israelis that he means what he says. https://t.co/Msl5sMilvC

2022-11-19 20:39:15 There is nothing like waking up Saturday morning to news about PEACE - the first Shi'ite Muslim country to have an Embassy in Israel. You feel like singing something to all the doomsday prophets, but the song is yet to be written. https://t.co/18bMIdqQBl

2022-11-19 10:58:51 @ShreeParadkar Your piece implies that @TorontoStar's readers are too feeble minded to distinguish “Denying the Jewish people their right to self-determination" from criticizing policies of Israeli government. Invariably, those who oppose IRHA are also guilty of the former. Not a coincidence!

2022-11-19 02:59:09 @PHuenermund Who left?

2022-11-18 23:48:12 @jeremy1174 @EinatWilf You can evaluate the mentality of Al Jazeera through your words and beliefs, according to which blood-thirsty IDF is shooting children for sport. And they have succeeded to convince even people of your intelligence that this is journalism.

2022-11-18 17:05:18 RT @AdrianRaftery1: Congratulations, Carlos. Great to see your whole dissertation available. (For some reason unsent at the time, but even…

2022-11-18 17:03:05 Police said “four shots were fired from a loaded weapon” into the exterior of the home next to the city’s Old Synagogue but that no one was injured. https://t.co/1gIjSV10jc

2022-11-18 16:55:18 @jeremy1174 @EinatWilf Often false and sometimes merely distorted. You would never see an Israeli child on Al-Jazeera -- where have all the mothers gone?

2022-11-18 05:19:36 @soboleffspaces Amazing, isn't it? Now imagine my child-like glee upon discovering it in 1999 (See Theorem 4 here: https://t.co/QQa4sdcJMT). I felt like I solved Abraham's problem "What if there were 50 righteous people in the city?"

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-11-21 01:51:41 Mehdi Hasan's logic eclipses only his mentality. He just discovered a couple of Qatari mosques that do NOT teach hate. Even Al Jazeera program "Sharia On Line" is no loner hosted by Al Qaradawi -- Lay off! Qatar is no longer the hate capital of the world, there are others! https://t.co/CYOjNyhOiU

2022-11-20 11:10:32 A clash of two "rights" does not make one of them "wrong". See how the notion of "equally indigenous" is defined in the post-colonial era: https://t.co/9XOCOs9hGT. https://t.co/ResC3LzbzQ

2022-11-20 09:33:04 We now received a link to the original @CWRUPresKaler statement of Nov. 9 https://t.co/q3HvNqFbKI and, as expected, President Kaler is being attacked by @CAIRNational and other Zionophobes. He deserves a broad community support at erik.kaler@case.edu https://t.co/vwI6T2FTg0

2022-11-20 08:49:39 @jeremy1174 @EinatWilf You just defined the essence of the conflict in no ambiguous terms: "The Palestinians are happy to coexist with Israelis - just not in Palestine, which they consider solely theirs, all of it, to the last inch, to the last man!" Coexistence Palestinian style and their tragedy.

2022-11-20 08:36:00 Jewish members our community.” What sets this statement apart from so many others is: (1) It does not propose a "philosophical study of antisemitism," and (2) It explicitly takes "a vote against Israel" to be a moral depravity, incompatible with community values. Hats off!! 2/2

2022-11-20 08:35:59 @CWRUPresKaler, President of Case Reserve, belongs to a handful of academic leaders who know the effective language with which to curb BDS hate. "A vote for this [BDS] resolution," he wrote on Nov. 11, "is clearly a vote against Israel and an aggression toward the 1/2

2022-11-20 07:45:58 Readers asking to see formal treatments of the Firing Squad and Oxygen-Match stories, can find them here https://t.co/yUnVKc07BE (Figs. 7, 9) and here https://t.co/yd1Z4QGARx. The reason I take them as litmus tests of "understanding" is: (1) Every 10-year old would vouch 1/2 https://t.co/60vgIYpZUN

2022-11-20 03:23:16 We are in receipt of a link to an illuminating symposia of Anthroponeny: https://t.co/3pah2sxdvn About -1:52 into the video, we can find Terry Sejnowski's talk which asks whether ML, Neural Nets, and Large Language Models can understand "intelligence", "explainability", 1/3

2022-11-20 03:22:25 @jeremy1174 @EinatWilf If you do not Al Jazeera, where did you get this maligned understanding of the "Zionist plan." My understanding of the plan is "Co-existence" with only ONE condition: "Existence". This entails no "buckling under" ecept dismantling the current Palestinian plan of "Sole-existence".

2022-11-19 23:55:46 Honored to join and support Moment Magazine's gala https://t.co/WhSupVLYcS, noting that Moment's Daniel Pearl Investigative Journalism Initiative https://t.co/VtqCxSs4Zv is launching new projects. dpiji story "An Inconvenient Genocide" won two first place journalistic awards.

2022-11-19 23:15:31 This 4:1 ratio is at least reversed in all previous reports by campus watchers (i.e., #ADL, AJC, Hillel, etc). The ratio approaches ZERO in ALL reports issued by University administrations or EDI offices (including UCLA). Hail to statistics! One thing it's good for.

2022-11-19 23:15:30 A groundbreaking and well documented report on campus assaults on Jewish identity: https://t.co/yM6szdGRGs. How do I know it hits the right issues? Pure statistics: The word "anti-Zionism" appears 4 times more often than the word "antisemitism"

2022-11-19 22:40:15 For our German readers: A "must read" for everyone who is really interested in a solution to to the Israeli-Palestinian conflict. https://t.co/HUyYhZrEFY

2022-11-19 21:25:47 Well put! Plus note the strange correlation between those who oppose IHRA and those who deny Israel's right to exist. By endorsing the former, Stewart emboldens the latter. https://t.co/al9lxyyUzG

2022-11-19 21:11:12 @haig Show me one "peace agreement" in the world that was not signed out of some practical "incentives" to the sound of warnings: "don't be naïve!". On Nov. 19, 1977, 45 years ago today, President Sadat landed in Israel to the sound of similar warnings. Yet it's worked, &

2022-11-19 20:57:00 Will President Abbas ever rise to Sadat's height and address the Israeli Knesset with a message of peace? In 2008 I was so naïve as to write the speech for him, https://t.co/iTXOtbDXf7 a speech that would convince Israelis that he means what he says. https://t.co/Msl5sMilvC

2022-11-19 20:39:15 There is nothing like waking up Saturday morning to news about PEACE - the first Shi'ite Muslim country to have an Embassy in Israel. You feel like singing something to all the doomsday prophets, but the song is yet to be written. https://t.co/18bMIdqQBl

2022-11-19 10:58:51 @ShreeParadkar Your piece implies that @TorontoStar's readers are too feeble minded to distinguish “Denying the Jewish people their right to self-determination" from criticizing policies of Israeli government. Invariably, those who oppose IRHA are also guilty of the former. Not a coincidence!

2022-11-19 02:59:09 @PHuenermund Who left?

2022-11-18 23:48:12 @jeremy1174 @EinatWilf You can evaluate the mentality of Al Jazeera through your words and beliefs, according to which blood-thirsty IDF is shooting children for sport. And they have succeeded to convince even people of your intelligence that this is journalism.

2022-11-18 17:05:18 RT @AdrianRaftery1: Congratulations, Carlos. Great to see your whole dissertation available. (For some reason unsent at the time, but even…

2022-11-18 17:03:05 Police said “four shots were fired from a loaded weapon” into the exterior of the home next to the city’s Old Synagogue but that no one was injured. https://t.co/1gIjSV10jc

2022-11-18 16:55:18 @jeremy1174 @EinatWilf Often false and sometimes merely distorted. You would never see an Israeli child on Al-Jazeera -- where have all the mothers gone?

2022-11-18 05:19:36 @soboleffspaces Amazing, isn't it? Now imagine my child-like glee upon discovering it in 1999 (See Theorem 4 here: https://t.co/QQa4sdcJMT). I felt like I solved Abraham's problem "What if there were 50 righteous people in the city?"

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-11-22 05:25:33 This Title IV Complaint against @BerkeleyLaw's bylaws, aside from protecting Jewish students from discrimination, will have another impact on campus life, perhaps more profound, to render anti-Zionism incompatible with societal values. https://t.co/lxWgTq2O2k

2022-11-22 03:27:38 Scary to realize they already have more than 7 followers. https://t.co/WnyzM7kjbU

2022-11-22 03:24:21 Glad they placed your "La Guerra del Retorno" next to "Palestina," to maintain some commensensical balance. https://t.co/iwoG3G09qw

2022-11-22 02:51:30 Would the "real Jews" step up and replace them! https://t.co/Vj79pGEpPU

2022-11-22 02:22:13 @DavidLobron Yes, all you need to do is ask Google Scholar to send you an alarm for specific citations. I use Causality (2009) and @Bookofwhy (2018).

2022-11-22 01:59:14 If in LA next Monday, join me in commemorating the 75th anniversary of the historical UN Vote of November 29, 1947. The event will take place November 28 (not 29), 11:30 am, at the Los Angeles Museum of Tolerance. See flyer: https://t.co/uBV8XjrLhc. The program will include 1/2

2022-11-21 21:32:16 We are ready again to scan this week's yield of Causal Inference papers:https://t.co/RGtPVsERkq I can spot already a few titles which should invite our curiosity, and I will come back to them during the week, as soon as deadlines and alligators are pacified.

2022-11-21 09:51:58 The more I read about the LLM controversy https://t.co/1YYe6xnAl2 the happier I am for doing research on problems that have true answers. Like why we say that the fire started by striking a match and not by the Oxygen in the room. It's such a pleasure getting machines to agree. https://t.co/PWg2IUoK1z

2022-11-21 01:51:41 Mehdi Hasan's logic eclipses only his mentality. He just discovered a couple of Qatari mosques that do NOT teach hate. Even Al Jazeera program "Sharia On Line" is no loner hosted by Al Qaradawi -- Lay off! Qatar is no longer the hate capital of the world, there are others! https://t.co/CYOjNyhOiU

2022-11-20 11:10:32 A clash of two "rights" does not make one of them "wrong". See how the notion of "equally indigenous" is defined in the post-colonial era: https://t.co/9XOCOs9hGT. https://t.co/ResC3LzbzQ

2022-11-20 09:33:04 We now received a link to the original @CWRUPresKaler statement of Nov. 9 https://t.co/q3HvNqFbKI and, as expected, President Kaler is being attacked by @CAIRNational and other Zionophobes. He deserves a broad community support at erik.kaler@case.edu https://t.co/vwI6T2FTg0

2022-11-20 08:49:39 @jeremy1174 @EinatWilf You just defined the essence of the conflict in no ambiguous terms: "The Palestinians are happy to coexist with Israelis - just not in Palestine, which they consider solely theirs, all of it, to the last inch, to the last man!" Coexistence Palestinian style and their tragedy.

2022-11-20 08:36:00 Jewish members our community.” What sets this statement apart from so many others is: (1) It does not propose a "philosophical study of antisemitism," and (2) It explicitly takes "a vote against Israel" to be a moral depravity, incompatible with community values. Hats off!! 2/2

2022-11-20 08:35:59 @CWRUPresKaler, President of Case Reserve, belongs to a handful of academic leaders who know the effective language with which to curb BDS hate. "A vote for this [BDS] resolution," he wrote on Nov. 11, "is clearly a vote against Israel and an aggression toward the 1/2

2022-11-20 07:45:58 Readers asking to see formal treatments of the Firing Squad and Oxygen-Match stories, can find them here https://t.co/yUnVKc07BE (Figs. 7, 9) and here https://t.co/yd1Z4QGARx. The reason I take them as litmus tests of "understanding" is: (1) Every 10-year old would vouch 1/2 https://t.co/60vgIYpZUN

2022-11-20 03:23:16 We are in receipt of a link to an illuminating symposia of Anthroponeny: https://t.co/3pah2sxdvn About -1:52 into the video, we can find Terry Sejnowski's talk which asks whether ML, Neural Nets, and Large Language Models can understand "intelligence", "explainability", 1/3

2022-11-20 03:22:25 @jeremy1174 @EinatWilf If you do not Al Jazeera, where did you get this maligned understanding of the "Zionist plan." My understanding of the plan is "Co-existence" with only ONE condition: "Existence". This entails no "buckling under" ecept dismantling the current Palestinian plan of "Sole-existence".

2022-11-19 23:55:46 Honored to join and support Moment Magazine's gala https://t.co/WhSupVLYcS, noting that Moment's Daniel Pearl Investigative Journalism Initiative https://t.co/VtqCxSs4Zv is launching new projects. dpiji story "An Inconvenient Genocide" won two first place journalistic awards.

2022-11-19 23:15:31 This 4:1 ratio is at least reversed in all previous reports by campus watchers (i.e., #ADL, AJC, Hillel, etc). The ratio approaches ZERO in ALL reports issued by University administrations or EDI offices (including UCLA). Hail to statistics! One thing it's good for.

2022-11-19 23:15:30 A groundbreaking and well documented report on campus assaults on Jewish identity: https://t.co/yM6szdGRGs. How do I know it hits the right issues? Pure statistics: The word "anti-Zionism" appears 4 times more often than the word "antisemitism"

2022-11-19 22:40:15 For our German readers: A "must read" for everyone who is really interested in a solution to to the Israeli-Palestinian conflict. https://t.co/HUyYhZrEFY

2022-11-19 21:25:47 Well put! Plus note the strange correlation between those who oppose IHRA and those who deny Israel's right to exist. By endorsing the former, Stewart emboldens the latter. https://t.co/al9lxyyUzG

2022-11-19 21:11:12 @haig Show me one "peace agreement" in the world that was not signed out of some practical "incentives" to the sound of warnings: "don't be naïve!". On Nov. 19, 1977, 45 years ago today, President Sadat landed in Israel to the sound of similar warnings. Yet it's worked, &

2022-11-19 20:57:00 Will President Abbas ever rise to Sadat's height and address the Israeli Knesset with a message of peace? In 2008 I was so naïve as to write the speech for him, https://t.co/iTXOtbDXf7 a speech that would convince Israelis that he means what he says. https://t.co/Msl5sMilvC

2022-11-19 20:39:15 There is nothing like waking up Saturday morning to news about PEACE - the first Shi'ite Muslim country to have an Embassy in Israel. You feel like singing something to all the doomsday prophets, but the song is yet to be written. https://t.co/18bMIdqQBl

2022-11-19 10:58:51 @ShreeParadkar Your piece implies that @TorontoStar's readers are too feeble minded to distinguish “Denying the Jewish people their right to self-determination" from criticizing policies of Israeli government. Invariably, those who oppose IRHA are also guilty of the former. Not a coincidence!

2022-11-19 02:59:09 @PHuenermund Who left?

2022-11-18 23:48:12 @jeremy1174 @EinatWilf You can evaluate the mentality of Al Jazeera through your words and beliefs, according to which blood-thirsty IDF is shooting children for sport. And they have succeeded to convince even people of your intelligence that this is journalism.

2022-11-18 17:05:18 RT @AdrianRaftery1: Congratulations, Carlos. Great to see your whole dissertation available. (For some reason unsent at the time, but even…

2022-11-18 17:03:05 Police said “four shots were fired from a loaded weapon” into the exterior of the home next to the city’s Old Synagogue but that no one was injured. https://t.co/1gIjSV10jc

2022-11-18 16:55:18 @jeremy1174 @EinatWilf Often false and sometimes merely distorted. You would never see an Israeli child on Al-Jazeera -- where have all the mothers gone?

2022-11-18 05:19:36 @soboleffspaces Amazing, isn't it? Now imagine my child-like glee upon discovering it in 1999 (See Theorem 4 here: https://t.co/QQa4sdcJMT). I felt like I solved Abraham's problem "What if there were 50 righteous people in the city?"

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-11-22 22:30:27 @hail2pitt01 These guys truly believe that Mehdi Hasan and other Al-Jazeera graduates should succeed in giving them an air-tight civilized cover. After all, some Western organizations have begun calling Al Jazeera cronies "journalists".

2022-11-22 21:37:04 Qatar must be having second thoughts on the wisdom of hosting the FIFA games, allowing the whole world discover what a rotten sewage Al-Jazeera has been trying to conceal all these years under its "news outlet" title and its British accented anchors. I've seen Doha 2012

2022-11-22 07:45:56 Friends ask: Why is it hard for Westerners to understand what's going on in the Israeli-Palestinian conflict? My answer: Because for Westerners, "understanding" entails SOME LOGIC while, in this conflict, some players defies ALL LOGIC. https://t.co/xwNKUVzAIx

2022-11-22 05:25:33 This Title IV Complaint against @BerkeleyLaw's bylaws, aside from protecting Jewish students from discrimination, will have another impact on campus life, perhaps more profound, to render anti-Zionism incompatible with societal values. https://t.co/lxWgTq2O2k

2022-11-22 03:27:38 Scary to realize they already have more than 7 followers. https://t.co/WnyzM7kjbU

2022-11-22 03:24:21 Glad they placed your "La Guerra del Retorno" next to "Palestina," to maintain some commensensical balance. https://t.co/iwoG3G09qw

2022-11-22 02:51:30 Would the "real Jews" step up and replace them! https://t.co/Vj79pGEpPU

2022-11-22 02:22:13 @DavidLobron Yes, all you need to do is ask Google Scholar to send you an alarm for specific citations. I use Causality (2009) and @Bookofwhy (2018).

2022-11-22 01:59:14 If in LA next Monday, join me in commemorating the 75th anniversary of the historical UN Vote of November 29, 1947. The event will take place November 28 (not 29), 11:30 am, at the Los Angeles Museum of Tolerance. See flyer: https://t.co/uBV8XjrLhc. The program will include 1/2

2022-11-21 21:32:16 We are ready again to scan this week's yield of Causal Inference papers:https://t.co/RGtPVsERkq I can spot already a few titles which should invite our curiosity, and I will come back to them during the week, as soon as deadlines and alligators are pacified.

2022-11-21 09:51:58 The more I read about the LLM controversy https://t.co/1YYe6xnAl2 the happier I am for doing research on problems that have true answers. Like why we say that the fire started by striking a match and not by the Oxygen in the room. It's such a pleasure getting machines to agree. https://t.co/PWg2IUoK1z

2022-11-21 01:51:41 Mehdi Hasan's logic eclipses only his mentality. He just discovered a couple of Qatari mosques that do NOT teach hate. Even Al Jazeera program "Sharia On Line" is no loner hosted by Al Qaradawi -- Lay off! Qatar is no longer the hate capital of the world, there are others! https://t.co/CYOjNyhOiU

2022-11-20 11:10:32 A clash of two "rights" does not make one of them "wrong". See how the notion of "equally indigenous" is defined in the post-colonial era: https://t.co/9XOCOs9hGT. https://t.co/ResC3LzbzQ

2022-11-20 09:33:04 We now received a link to the original @CWRUPresKaler statement of Nov. 9 https://t.co/q3HvNqFbKI and, as expected, President Kaler is being attacked by @CAIRNational and other Zionophobes. He deserves a broad community support at erik.kaler@case.edu https://t.co/vwI6T2FTg0

2022-11-20 08:49:39 @jeremy1174 @EinatWilf You just defined the essence of the conflict in no ambiguous terms: "The Palestinians are happy to coexist with Israelis - just not in Palestine, which they consider solely theirs, all of it, to the last inch, to the last man!" Coexistence Palestinian style and their tragedy.

2022-11-20 08:36:00 Jewish members our community.” What sets this statement apart from so many others is: (1) It does not propose a "philosophical study of antisemitism," and (2) It explicitly takes "a vote against Israel" to be a moral depravity, incompatible with community values. Hats off!! 2/2

2022-11-20 08:35:59 @CWRUPresKaler, President of Case Reserve, belongs to a handful of academic leaders who know the effective language with which to curb BDS hate. "A vote for this [BDS] resolution," he wrote on Nov. 11, "is clearly a vote against Israel and an aggression toward the 1/2

2022-11-20 07:45:58 Readers asking to see formal treatments of the Firing Squad and Oxygen-Match stories, can find them here https://t.co/yUnVKc07BE (Figs. 7, 9) and here https://t.co/yd1Z4QGARx. The reason I take them as litmus tests of "understanding" is: (1) Every 10-year old would vouch 1/2 https://t.co/60vgIYpZUN

2022-11-20 03:23:16 We are in receipt of a link to an illuminating symposia of Anthroponeny: https://t.co/3pah2sxdvn About -1:52 into the video, we can find Terry Sejnowski's talk which asks whether ML, Neural Nets, and Large Language Models can understand "intelligence", "explainability", 1/3

2022-11-20 03:22:25 @jeremy1174 @EinatWilf If you do not Al Jazeera, where did you get this maligned understanding of the "Zionist plan." My understanding of the plan is "Co-existence" with only ONE condition: "Existence". This entails no "buckling under" ecept dismantling the current Palestinian plan of "Sole-existence".

2022-11-19 23:55:46 Honored to join and support Moment Magazine's gala https://t.co/WhSupVLYcS, noting that Moment's Daniel Pearl Investigative Journalism Initiative https://t.co/VtqCxSs4Zv is launching new projects. dpiji story "An Inconvenient Genocide" won two first place journalistic awards.

2022-11-19 23:15:31 This 4:1 ratio is at least reversed in all previous reports by campus watchers (i.e., #ADL, AJC, Hillel, etc). The ratio approaches ZERO in ALL reports issued by University administrations or EDI offices (including UCLA). Hail to statistics! One thing it's good for.

2022-11-19 23:15:30 A groundbreaking and well documented report on campus assaults on Jewish identity: https://t.co/yM6szdGRGs. How do I know it hits the right issues? Pure statistics: The word "anti-Zionism" appears 4 times more often than the word "antisemitism"

2022-11-19 22:40:15 For our German readers: A "must read" for everyone who is really interested in a solution to to the Israeli-Palestinian conflict. https://t.co/HUyYhZrEFY

2022-11-19 21:25:47 Well put! Plus note the strange correlation between those who oppose IHRA and those who deny Israel's right to exist. By endorsing the former, Stewart emboldens the latter. https://t.co/al9lxyyUzG

2022-11-19 21:11:12 @haig Show me one "peace agreement" in the world that was not signed out of some practical "incentives" to the sound of warnings: "don't be naïve!". On Nov. 19, 1977, 45 years ago today, President Sadat landed in Israel to the sound of similar warnings. Yet it's worked, &

2022-11-19 20:57:00 Will President Abbas ever rise to Sadat's height and address the Israeli Knesset with a message of peace? In 2008 I was so naïve as to write the speech for him, https://t.co/iTXOtbDXf7 a speech that would convince Israelis that he means what he says. https://t.co/Msl5sMilvC

2022-11-19 20:39:15 There is nothing like waking up Saturday morning to news about PEACE - the first Shi'ite Muslim country to have an Embassy in Israel. You feel like singing something to all the doomsday prophets, but the song is yet to be written. https://t.co/18bMIdqQBl

2022-11-19 10:58:51 @ShreeParadkar Your piece implies that @TorontoStar's readers are too feeble minded to distinguish “Denying the Jewish people their right to self-determination" from criticizing policies of Israeli government. Invariably, those who oppose IRHA are also guilty of the former. Not a coincidence!

2022-11-19 02:59:09 @PHuenermund Who left?

2022-11-18 23:48:12 @jeremy1174 @EinatWilf You can evaluate the mentality of Al Jazeera through your words and beliefs, according to which blood-thirsty IDF is shooting children for sport. And they have succeeded to convince even people of your intelligence that this is journalism.

2022-11-18 17:05:18 RT @AdrianRaftery1: Congratulations, Carlos. Great to see your whole dissertation available. (For some reason unsent at the time, but even…

2022-11-18 17:03:05 Police said “four shots were fired from a loaded weapon” into the exterior of the home next to the city’s Old Synagogue but that no one was injured. https://t.co/1gIjSV10jc

2022-11-18 16:55:18 @jeremy1174 @EinatWilf Often false and sometimes merely distorted. You would never see an Israeli child on Al-Jazeera -- where have all the mothers gone?

2022-11-18 05:19:36 @soboleffspaces Amazing, isn't it? Now imagine my child-like glee upon discovering it in 1999 (See Theorem 4 here: https://t.co/QQa4sdcJMT). I felt like I solved Abraham's problem "What if there were 50 righteous people in the city?"

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-11-29 08:03:31 We have received a new weekly list of papers on Causal Inference: https://t.co/CYoabDOmUM From last week's list, I've read and liked "One counterfactual does not make an explanation" https://t.co/whZftDKW91 1/2

2022-11-29 07:33:40 @agrchaitanya Try this primer: https://t.co/eKGIHSXQkH It is based on teaching causal inference at the high school level.

2022-11-29 07:13:55 @BorealisAI I tried to appreciate the result. Alas, the problem is not defined in an input-output language, and the relationships to the theory of fusion are skipped. Saying: "There also exists a large body of work on...causal graphs [41, 42, 40]" is saying: "lets obscure the problem" in PO.

2022-11-29 05:30:14 It was an honor to participate in this commemorative event today. And to all friends and readers who could not make it, here is a link to the video shown (10 min): https://t.co/ToOyfvNsyX which proves Weizmann's point: “No state is served to a nation on a silver platter.” (1947) https://t.co/abJdpYSgb3

2022-11-28 06:20:25 On the counterfactual interpretation of "Giving Thanks", in the context of tomorrow's commemorative event https://t.co/uBV8Xj9C34 Does it signal doubts in one's right to exist? https://t.co/KQEOqLzKpI

2022-11-27 20:03:30 As a student of cause and effect (#Bookofwhy), and a lifetime observer of the conflict I could not agree more with the analysis of @EinatWilf. The evidence is overwhelming and the logic is unassailable. https://t.co/MeXUP1iyfC

2022-11-27 12:53:46 @r_liebman @Ostrov_A The reason for the rejection is clear, and was repeated by many Arab leaders, even recently by Palestinian negotiators. It boils down to: Jews do not belong here, period. See https://t.co/9XOCOs9hGT

2022-11-27 10:17:11 @AleksanderMolak Your tweet re-kindles my belief in the usefulness of Twitter. Some day Econometrica will open its pages to open discussions of these foundational issues https://t.co/sZPpL7Ifmg, and perhaps others: https://t.co/etASSEk4Or .

2022-11-27 01:53:49 Apropos, I find it hard to understand how economists can reject a framework that embraces, illuminates and generalizes the one they prefer. Teenage rebellion? #Econometric @causalin @PHuenermund @VC31415 https://t.co/giqSKa32hK

2022-11-27 01:26:24 @r_liebman @Ostrov_A An interesting legal question: Do we have signed documents attesting to Jews-yes, Arabs-no? The former is well documented in the tears and dances of Nov. 29, 1947, but how do we document a NO? Is a 5 armies attack a legal document? How about the "From the river.." chanted today?

2022-11-27 00:08:45 Glad the book is back on the shelf, last time I was at Ben Gurion airport they didn't have it. Frankly, it's the best "do it yourself" instruction manual for Kanye West and others who are trying to be the "true Jews" https://t.co/DBZTjljxCE

2022-11-26 23:51:04 I'll be delighted to comment on Nick's review of #Bookofwhy as soon as I climb the paywall. In the meantime please re-read my comments on Imbens https://t.co/sZPpL7Ifmg, and my monolog/debate with Heckman and Rubin: https://t.co/GIaEtSH0y3 https://t.co/h5BT1EPMhj

2022-11-26 21:40:00 @DBSpro539 @BedouinHarith @WSB_OfficialMod @CulturallHub @elderofziyon @amnestyusa These genetic studies are interesting. However, historical attachment plays no lesser role in defining indigeneity to a land.

2022-11-26 21:35:37 Please note the the event takes place on Monday, November 28 - not 29 The flyer is correct: https://t.co/bei6HQVmSX. https://t.co/dwMSKbeluk

2022-11-26 12:18:37 @PHuenermund I remember this one! I had a good copy editor (My son Danny) which made the whole difference.

2022-11-26 09:19:28 @prem_k In 1-2 years, we are going to have platforms that allow a physician to ask any question (in Nat. Lang.), insert any data she has, and get the answer in the form of annotated figures or formulas. We just need to transfer 0.0001% of the resources now going to data-centric ML.

2022-11-26 09:11:40 Kielce, Poland, July 4th, 1946 https://t.co/naoS0xpoCM My aunt was there, to comfort the survivors. President Truman heard of this massacre and ordered his generals not to impede with the movement of Jewish refugees on their way to Italian ports. Destination: UK Blockaded Israel. https://t.co/dcNYfcrrAD

2022-11-26 08:51:26 My salute of the week. From the ashes of despair to a life of hope and meaning. https://t.co/WC37vgPbal

2022-11-26 08:27:01 RT @IsraelNitzan: Be their voice! #IranProtests

2022-11-26 06:07:46 The story about my science teacher, Dr. Feuchtwanger, https://t.co/2w0vo7zWxR is not a legend, but honest truth. He also taught Daniel Kahneman who was 1yr above me in high school. For seventy years I knew that he was right and that some logic must show him right. What 1/3 https://t.co/XfxeRf2gdo

2022-11-26 05:28:55 @RaulMachadoG I agree with your conclusion: "all roads lead to causal inference", but I could not follow your arguements being "acronym-illiterate."

2022-11-26 04:38:03 @AvivaKlompas What do you mean "repulsive"? This guys is threatening Jews with another Holocaust, unless they behave... His proof: it happened before, when Jews didn't behave. I've never heard this kind of open AS in America before. The Ayatollah are angels in comparison. Let's think Israel.

2022-11-26 01:02:58 If you are asking about the #Bookofwhy, the answer is definitely Yes. I've already commended the narrator for doing a better job than I thought possible. It sounds like an Agatha Christie thriller. https://t.co/1H5lFqwnOZ

2022-11-26 00:44:17 @maths_gee Curious. Where/how did you get this inner belief?

2022-11-26 00:29:19 Readers of #Bookofwhy asked why I used the smallpox story in Fig. 1.7, not Covid-19. Ans. I have a soft spot for smallpox, b/c it was my first hint of the dark side of causation (see https://t.co/2w0vo7QZzR). 75 years later, we found the solution here https://t.co/rGHIc9kH0O

2022-11-26 00:11:26 @FreeingAlanis Herzl was a secular Jew. When he spoke about "the Jewish fold" he meant the fold of Jewish peoplehood, not religion.

2022-11-25 22:14:11 @DoctorNazarian @NEWSMAX I would begin here: https://t.co/iWqdjOde7b Then go here: https://t.co/fSudyriXeY and then listen to why many decision makers take the antisemitic card as a license for inaction https://t.co/qHtORmprjm

2022-11-25 21:43:37 Recall what Herzl said in the First Zionist Congress (1897): "Zionism is a homecoming to the Jewish fold even before it becomes a homecoming to the Jewish land." See https://t.co/ICqKCBgrOh @newzionists https://t.co/DwEWo6bAfa

2022-11-25 21:34:16 @DoctorNazarian @NEWSMAX Don't forget to talk about Zionophobia, without which people will start yawning at you.

2022-11-25 21:29:39 @AnasBf1 @EinatWilf There is still a fundamental difference between the two examples. Muslims are quick to disassociate themselves from terrorism, while most Jews are proud of their Zionism and their care for Israel (Not counting mutants like @PeterBeinart and Neturei Kartha.) So, anti-Z is anti-J.

2022-11-25 19:03:12 @AnasBf1 @EinatWilf Good argument, and this is one of the reasons I do not feel comfortable calibrating the evils of anti-Zionism by the harm it does to Jews worldwide. It should be judged by its own moral depravities and genocidal consequences.

2022-11-25 09:31:47 @naivebaesian It's a good book. Chapter 4, on counterfactuals, is relevant to our discussion on "thanksgiving".See corrected version: https://t.co/Sf5yw3khNt

2022-11-25 09:25:33 It's refreshing to see so many "professors" competing to be the "true Jews." Thinking: Perhaps we do have something valuable to offer mankind? And perhaps doing it under false identity helps us do it well. https://t.co/99rc5fPOQC

2022-11-25 08:58:38 Can the savage murder of a young (18) man galvanize a united front against savagery? https://t.co/OXFeOSLqnp

2022-11-25 08:48:24 @elderofziyon @amnestyusa managed to distort the face of "Human rights" beyond recognition. I still welcome their recognizing the indigenous rights of my grandfather and his indigenous journey back to his indigenous homeland, in 1924. I hope this makes me indigenous too. Didn't I know this?

2022-11-25 08:24:38 The LA Times, again, did not get it right (purposely, I surmise). The complaint is NOT against "refusing to invite Zionist speakers" but against writing it into their Bylaws. A restaurant can refuse service to any disagreeable customer, but it can't hang a sign "no blacks served" https://t.co/hcpGT6pwIg

2022-11-25 08:16:40 RT @EinatWilf: “I can't think of a single instance where anti-Zionism turned out well for Jews or for anti-Zionists". Listen here as to why…

2022-11-25 05:28:38 @ColocateDisrupt To us, AI researchers, the Substance of the Thing lies in our ability to emulate the Thing on a digital machine. So don't be angry.

2022-11-25 05:25:32 @Sara_james_2022 Please do not leave us to fight this battle alone!

2022-11-25 04:36:45 @ColocateDisrupt Far easier, indeed. I used to prey to that Being every morning. But to those of us who are also interested in the cognitive processes involved in "thanking" and the computational resources needed for such processes, the theory of counterfactuals offers some insights.

2022-11-25 01:18:09 Speaking counterfactually about "Thanksgiving", we should not forget the "Jewish thanksgiving day" https://t.co/9AqC9fqo9d, and the incredible thanksgiving project that the city of Netanya has launched: https://t.co/TZXgTxfjIA just to offer its thanks to the Goddess of history. https://t.co/RV4gz5DOWD

2022-11-25 01:03:49 The logical absurdity of anti-coexistence philosophers demands a humorous exposure: I choke you to death. You scream: 'I can't breath!' I say: You see, I told you, co-existence doesn't work, now you must cease to exist! Causality 101. @harari_yuval https://t.co/ZU0lfsKD0z

2022-11-25 00:42:10 @elderofziyon Jamie Stern-Weiner's rhetoric is typical of Zionophobic racism. Who in the world would dare to deny that "It isn't antisemitic to oppose Jewish thirst for Christian blood." Great logic! Jamie.

2022-11-25 00:24:39 @YKampf @EinatWilf @YUVAL_YUVAL As a student of causality, I would first examine whether the perceived weakening of Israel's democracy is not a temporary byproduct of rising Palestinian eliminationism. Or, as the poet said: "Give us one day of normalcy, and we'll show what democracy looks like." @harari_yuval

2022-11-24 23:30:50 UC Berkeley student groups' refusal to invite Zionist speakers draws civil rights complaint https://t.co/5FqXDLU1dj

2022-11-24 23:10:06 Correcting my error: It's @harari_yuval that is profoundly wrong, not @yuval_yuval which does not exist. And I join @EinatWilf in urging @harari_yuval to analyze the prevailing attitude in the Palestinian society towards the dream we once called "co-existence". https://t.co/GLi8JMi486

2022-11-24 22:49:51 The jackal, so the legend goes, is the first to howl when its victim dies. https://t.co/IKNco9wQ0h

2022-11-24 22:41:31 @EinatWilf @Yuval_yuval is somewhat right when we superficially consider naked facts on the ground

2022-11-24 22:21:38 The counterfactual "Thanksgiving" is both an affirmation that "things could have been worse" and an attitude to "view the ordinary as a miracle." Do we want our robots to thank us for the ordinary? Too bothersome. But we certainly won't like them to take things for granted. https://t.co/Q9k9WqBY8f

2022-11-24 07:09:04 @LisaStadelbauer Truly offensive, @LisaStadebauer. Who wrote your "all forms of terror" tweet? The Canadian Society for the Banalization of Evil?

2022-11-24 07:01:11 RT @LtColRichard: Following the efforts of the security forces and in coordination with the Palestinian security forces and authorities, th…

2022-11-24 06:50:33 Extremely delighted to see a contest on SCM implementation in #python. Can we have more details, e.g., what the research question was, or the book's Table of Content . https://t.co/QuykwRc52l

2022-11-24 06:42:01 Why we filed a civil rights claim against UC Berkeley Law School | Opinion https://t.co/IYquYreI3m

2022-11-24 05:50:19 RT @michaeldickson: This is the teenager Palestinian jihadi terrorists murdered today. Aryeh Schupak was just 16. Totally heartbreaking. Th…

2022-11-24 05:49:32 Mother Theresa has struggled all her life to deserve the level of Sainthood and purity that Palestinians have achieved with ordinary bombs, stuffed with nails, painted "Justice in Palestine." https://t.co/4z7Itpu0nv

2022-11-23 23:01:20 It seems that Ambassador @DTzantchevdid did not get my Tweet. "Only people @DTzantchev follows or mentioned can reply." Can someone convey my message to him (below)? Else he would never know "what the Palestinians really want". Thanks https://t.co/HRoTSMAAmo

2022-11-23 10:47:22 He was only 16, and left us to ask the hard questions: When has our mind accepted the banalization of evil as a feasible process? Can we tell his parents that human mind is wired that way? HaShem Yinkom Damo, https://t.co/lgAzvsRBKu

2022-11-22 22:30:27 @hail2pitt01 These guys truly believe that Mehdi Hasan and other Al-Jazeera graduates should succeed in giving them an air-tight civilized cover. After all, some Western organizations have begun calling Al Jazeera cronies "journalists".

2022-11-22 21:37:04 Qatar must be having second thoughts on the wisdom of hosting the FIFA games, allowing the whole world discover what a rotten sewage Al-Jazeera has been trying to conceal all these years under its "news outlet" title and its British accented anchors. I've seen Doha 2012

2022-11-22 07:45:56 Friends ask: Why is it hard for Westerners to understand what's going on in the Israeli-Palestinian conflict? My answer: Because for Westerners, "understanding" entails SOME LOGIC while, in this conflict, some players defies ALL LOGIC. https://t.co/xwNKUVzAIx

2022-11-22 05:25:33 This Title IV Complaint against @BerkeleyLaw's bylaws, aside from protecting Jewish students from discrimination, will have another impact on campus life, perhaps more profound, to render anti-Zionism incompatible with societal values. https://t.co/lxWgTq2O2k

2022-11-22 03:27:38 Scary to realize they already have more than 7 followers. https://t.co/WnyzM7kjbU

2022-11-22 03:24:21 Glad they placed your "La Guerra del Retorno" next to "Palestina," to maintain some commensensical balance. https://t.co/iwoG3G09qw

2022-11-22 02:51:30 Would the "real Jews" step up and replace them! https://t.co/Vj79pGEpPU

2022-11-22 02:22:13 @DavidLobron Yes, all you need to do is ask Google Scholar to send you an alarm for specific citations. I use Causality (2009) and @Bookofwhy (2018).

2022-11-22 01:59:14 If in LA next Monday, join me in commemorating the 75th anniversary of the historical UN Vote of November 29, 1947. The event will take place November 28 (not 29), 11:30 am, at the Los Angeles Museum of Tolerance. See flyer: https://t.co/uBV8XjrLhc. The program will include 1/2

2022-11-21 21:32:16 We are ready again to scan this week's yield of Causal Inference papers:https://t.co/RGtPVsERkq I can spot already a few titles which should invite our curiosity, and I will come back to them during the week, as soon as deadlines and alligators are pacified.

2022-11-21 09:51:58 The more I read about the LLM controversy https://t.co/1YYe6xnAl2 the happier I am for doing research on problems that have true answers. Like why we say that the fire started by striking a match and not by the Oxygen in the room. It's such a pleasure getting machines to agree. https://t.co/PWg2IUoK1z

2022-11-21 01:51:41 Mehdi Hasan's logic eclipses only his mentality. He just discovered a couple of Qatari mosques that do NOT teach hate. Even Al Jazeera program "Sharia On Line" is no loner hosted by Al Qaradawi -- Lay off! Qatar is no longer the hate capital of the world, there are others! https://t.co/CYOjNyhOiU

2022-11-20 11:10:32 A clash of two "rights" does not make one of them "wrong". See how the notion of "equally indigenous" is defined in the post-colonial era: https://t.co/9XOCOs9hGT. https://t.co/ResC3LzbzQ

2022-11-20 09:33:04 We now received a link to the original @CWRUPresKaler statement of Nov. 9 https://t.co/q3HvNqFbKI and, as expected, President Kaler is being attacked by @CAIRNational and other Zionophobes. He deserves a broad community support at erik.kaler@case.edu https://t.co/vwI6T2FTg0

2022-11-20 08:49:39 @jeremy1174 @EinatWilf You just defined the essence of the conflict in no ambiguous terms: "The Palestinians are happy to coexist with Israelis - just not in Palestine, which they consider solely theirs, all of it, to the last inch, to the last man!" Coexistence Palestinian style and their tragedy.

2022-11-20 08:36:00 Jewish members our community.” What sets this statement apart from so many others is: (1) It does not propose a "philosophical study of antisemitism," and (2) It explicitly takes "a vote against Israel" to be a moral depravity, incompatible with community values. Hats off!! 2/2

2022-11-20 08:35:59 @CWRUPresKaler, President of Case Reserve, belongs to a handful of academic leaders who know the effective language with which to curb BDS hate. "A vote for this [BDS] resolution," he wrote on Nov. 11, "is clearly a vote against Israel and an aggression toward the 1/2

2022-11-20 07:45:58 Readers asking to see formal treatments of the Firing Squad and Oxygen-Match stories, can find them here https://t.co/yUnVKc07BE (Figs. 7, 9) and here https://t.co/yd1Z4QGARx. The reason I take them as litmus tests of "understanding" is: (1) Every 10-year old would vouch 1/2 https://t.co/60vgIYpZUN

2022-11-20 03:23:16 We are in receipt of a link to an illuminating symposia of Anthroponeny: https://t.co/3pah2sxdvn About -1:52 into the video, we can find Terry Sejnowski's talk which asks whether ML, Neural Nets, and Large Language Models can understand "intelligence", "explainability", 1/3

2022-11-20 03:22:25 @jeremy1174 @EinatWilf If you do not Al Jazeera, where did you get this maligned understanding of the "Zionist plan." My understanding of the plan is "Co-existence" with only ONE condition: "Existence". This entails no "buckling under" ecept dismantling the current Palestinian plan of "Sole-existence".

2022-11-19 23:55:46 Honored to join and support Moment Magazine's gala https://t.co/WhSupVLYcS, noting that Moment's Daniel Pearl Investigative Journalism Initiative https://t.co/VtqCxSs4Zv is launching new projects. dpiji story "An Inconvenient Genocide" won two first place journalistic awards.

2022-11-19 23:15:31 This 4:1 ratio is at least reversed in all previous reports by campus watchers (i.e., #ADL, AJC, Hillel, etc). The ratio approaches ZERO in ALL reports issued by University administrations or EDI offices (including UCLA). Hail to statistics! One thing it's good for.

2022-11-19 23:15:30 A groundbreaking and well documented report on campus assaults on Jewish identity: https://t.co/yM6szdGRGs. How do I know it hits the right issues? Pure statistics: The word "anti-Zionism" appears 4 times more often than the word "antisemitism"

2022-11-19 22:40:15 For our German readers: A "must read" for everyone who is really interested in a solution to to the Israeli-Palestinian conflict. https://t.co/HUyYhZrEFY

2022-11-19 21:25:47 Well put! Plus note the strange correlation between those who oppose IHRA and those who deny Israel's right to exist. By endorsing the former, Stewart emboldens the latter. https://t.co/al9lxyyUzG

2022-11-19 21:11:12 @haig Show me one "peace agreement" in the world that was not signed out of some practical "incentives" to the sound of warnings: "don't be naïve!". On Nov. 19, 1977, 45 years ago today, President Sadat landed in Israel to the sound of similar warnings. Yet it's worked, &

2022-11-19 20:57:00 Will President Abbas ever rise to Sadat's height and address the Israeli Knesset with a message of peace? In 2008 I was so naïve as to write the speech for him, https://t.co/iTXOtbDXf7 a speech that would convince Israelis that he means what he says. https://t.co/Msl5sMilvC

2022-11-19 20:39:15 There is nothing like waking up Saturday morning to news about PEACE - the first Shi'ite Muslim country to have an Embassy in Israel. You feel like singing something to all the doomsday prophets, but the song is yet to be written. https://t.co/18bMIdqQBl

2022-11-19 10:58:51 @ShreeParadkar Your piece implies that @TorontoStar's readers are too feeble minded to distinguish “Denying the Jewish people their right to self-determination" from criticizing policies of Israeli government. Invariably, those who oppose IRHA are also guilty of the former. Not a coincidence!

2022-11-19 02:59:09 @PHuenermund Who left?

2022-11-18 23:48:12 @jeremy1174 @EinatWilf You can evaluate the mentality of Al Jazeera through your words and beliefs, according to which blood-thirsty IDF is shooting children for sport. And they have succeeded to convince even people of your intelligence that this is journalism.

2022-11-18 17:05:18 RT @AdrianRaftery1: Congratulations, Carlos. Great to see your whole dissertation available. (For some reason unsent at the time, but even…

2022-11-18 17:03:05 Police said “four shots were fired from a loaded weapon” into the exterior of the home next to the city’s Old Synagogue but that no one was injured. https://t.co/1gIjSV10jc

2022-11-18 16:55:18 @jeremy1174 @EinatWilf Often false and sometimes merely distorted. You would never see an Israeli child on Al-Jazeera -- where have all the mothers gone?

2022-11-18 05:19:36 @soboleffspaces Amazing, isn't it? Now imagine my child-like glee upon discovering it in 1999 (See Theorem 4 here: https://t.co/QQa4sdcJMT). I felt like I solved Abraham's problem "What if there were 50 righteous people in the city?"

2022-11-18 00:03:18 @soboleffspaces Moreover, under monotonicity PN is identifiable, and so are PNS and PS.

2022-11-18 00:01:26 @jeremy1174 No. If you sift long enough through Al-Jazeera's texts, you are likely to find a fact or two that are untainted by hate and deceit, but who has the time. This is what I wrote about Al-Jazeera's tactics in 2008 https://t.co/q6DeznMy8s - they haven't changed since. @EinatWilf

2022-11-17 23:48:44 @Mahamoudmusa1 I have "The War of Return" with me on Kindle. Can you point me to ONE fact or ONE figure that you believe is fictional? Warning: I know my facts and figures, and I am also an eyewitness to some.

2022-11-17 23:44:16 @gorby_gorbachev @foodsevelt @besttrousers @HenryPorters @causalinf I am missing the mathematician's point. Is it worth un-missing?

2022-11-17 23:05:52 I believe your formula is valid only under monotonicity. How do I know? We know that PN is rung-3, while the formula you gave is rung-2. This is only possible under some Rung-3 assumptions. Monotonicity is such an assumption. https://t.co/5BjEBtp2JL

2022-11-17 21:50:50 How do we get Jon Stewart to read "The War of Return"? Ans. Don't call him "antisemite", call him an "innocent Al-Jazeera reader." https://t.co/jURpjmisvM

2022-11-17 14:33:53 @lucasbergkamp Invariably so. Non-causal problems are hardly perceived as "problems".

2022-11-17 07:23:45 @AvivaKlompas Worse yet

2022-11-17 06:20:17 An interesting court case affecting academia and beyond. I think the 6 CUNY professors have a point: If a given Union is spending its membership dues to malign symbols of identity of its members, the impacted members should have the option to obtain another union representation.

2022-11-17 06:20:16 STAN GREER: A Major Court Case Could Decide Just How Much Power Union Bosses Have Over Employees. Here’s How https://t.co/GIOZyy3leR via @dailycaller

2022-11-17 05:29:16 The word "parsimonious" was given to me by a leader in the SEM industry, when I asked: If SEM is but a statistical model of the data, why not estimate just any parametric model of the distribution? He answered: "SEM is both parsimonious and meaningful". The fact that both 1/2 https://t.co/dq9rHKJuZ3

2022-11-17 00:30:18 This handbook (and my chapter in it) reminds me of my decades-long struggles with SEM practitioners concerning the meaning of SEM. Watch how it started: https://t.co/Cz9SVFshD1 I suspect such nasty examiners are still ruling SEM education. https://t.co/7Z5QIBz1rm

2022-11-16 20:31:10 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn We don't even need to start with the overly complex infant's mind, because it's very easy to learn from data (ameba +evolution). I have examined this Tabula Rasa (model-free) paradigm here https://t.co/dxqdoqvpOm, and I am not too enthusiastic about its promises.

2022-11-16 20:14:43 We are in receipt of a free copy of "Handbook of Structural Equation Modeling" (2nd Ed) https://t.co/NvzOX67zbo I have a chapter (3) there, which is trying to convince SEM researchers that, contrary to tradition, SEM's are not parsimonious statistical models but causal models.

2022-11-16 20:05:27 @KordingLab @pmddomingos @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn How I love this answer!

2022-11-16 06:28:39 Reminds me how the media condemned "The Cycle of Violence" when my son was murdered. @TonyBlinken https://t.co/iMlmUlIlAg

2022-11-16 04:39:19 We are informed of a new workshop in the philosophy of statistics: https://t.co/BPdDIvc2DG. Searching the abstracts, I've found one reference to Causation, by Jon Williamson, who makes an astounding discovery: "Causal inference is not statistical inference". Needless to say 1/2

2022-11-16 03:28:02 @Claire_V0ltaire When Muhammad says "According to many reports" she is drawing on Yasser Arafat's technique of saying "Everyone knows that.." (e.g., to Clinton in Camp David: "Everyone knows the Jews never had a Temple in Jerusalem"). It works when your lie doesn't sound too credible.

2022-11-16 03:08:24 Jewish-power theories may have hidden virtues. Recently unearthed records of interviews with British officials reveal that it was fears of "international Jewish power" that jolted the British Government to issue the Balfour Declaration

2022-11-15 11:44:28 1947 U.N. Partition Resolution Memorial to be dedicated Nov. 29 https://t.co/upT4WrIfGJ via @JNS_org

2022-11-15 11:13:36 I am sure readers are eager to view this week's harvest of new causal Inference papers. Here they are : https://t.co/lOyu9ltGeN. I haven't had a chance to search for breakthroughs, perhaps our readers can alert us of any.

2022-11-15 10:10:22 An honest NYT account by @blakeflayton on being a progressive Zionist in 2022. The reason Americans fall for the lies they are being fed is that they are rational

2022-11-15 06:08:39 Congratulations to Naftali Weinberger for receiving the Philosophy of Science and Race Prize, and for introducing Causal Analysis to the analysis of Racial Discrimination. An author tagged @DAGophile should get an extra prize just for the courage. https://t.co/jhGCGwCqQX

2022-11-15 03:23:35 @pmddomingos @KordingLab @RasulElon @ylecun The answer is, almost no benefit. Causal theories to date have been viewed as orthogonal to inductive biases

2022-11-14 21:46:43 @pmddomingos @RasulElon @ylecun @KordingLab Agree. Now, causal theories have developed tools for dealing with this incompleteness (ie, missing actions). I am not familiar with such tools deployed in the "inductive bias theories." Are there?

2022-11-14 21:37:14 @pmddomingos @KordingLab @RasulElon @ylecun Causal theories are a type of inductive bias only in the very weak sense that both are needed on top of the data. But to say the XYZ is a type of ZYX is only helpful if XYZ can benefit from the tools developed for ZYX. Here the two diverge, with almost zero intersection.

2022-11-14 19:52:45 Speaking of "infinite data" one must distinguish "infinite sample on a given subset of variables" from "All subsets of variables". Both are unavailable yet representable. The former by a distribution function, the latter by a causal model in which assumptions are made about 1/2 https://t.co/KVWUgCc94G

2022-11-14 13:44:10 @RabbiWolpe That's why I much prefer the more powerful and less misconstrued accusation: Zionophobia. I haven't found many followers among Jewish leaders who either use the antisemitic card, or warn against overusing it.

2022-11-14 12:35:21 @KordingLab @RasulElon @ylecun @pmddomingos I sense progress coming from some corners of "Inductive-Bias Land", a recognition that experiments (or prior knowledge) on top of data are needed to cross the Rung-1 to Rung-2 barrier. Would there be a similar recognition coming regarding the barrier between Rung-2 to Rung-3 ?

2022-11-14 07:17:01 Readers wondering why I prefer toy examples to abstract discussions would be convinced, I hope, by this discussion on what "inductive bias" is, and whether an "inductively biased" machine can properly answer counterfactual questions in the firing-squad story of #Bookofwhy p.42. https://t.co/iBFmzlmNSP

2022-11-14 07:03:47 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn You can see now why I prefer toy examples to fancy words. Students of causal inference know that ANY DATA can equally fit another firing-squad story, in which Captain listens to Marksman-1 and Court listens to Captain's order. Here, counterfactuals get different truth values.

2022-11-14 06:38:57 Excellent story for the next edition of #Bookofwhy. https://t.co/raMKUq8prZ

2022-11-14 06:34:13 @RasulElon @ylecun The Ladder of Causation tells us that this beautiful property of convergence does not hold in Rung-2 and Rung-3 of the Ladder. The two models: (1) Symptom-->

2022-11-14 03:14:31 @ChristianBSmar1 What holds back many ML folks is this mantra of "Causation is just a species of 'inductive bias'", so, there is no need to worry, it would pop out automatically from more of the same, more data, and more talk about power of inductive bias to turn lead into gold.

2022-11-14 01:15:30 @jwbelmon What's #Fediverse or #Mastodon? I hope they are not buried in "inductive bias".

2022-11-14 01:13:11 @pmddomingos @KordingLab @scac1041 @PhilDawid @GaryMarcus @StephenPiment @ylecun @AlexTensor @gopnik @stephensenn #Bookofwhy p. 42 "The data will consist of two kinds of events: either all five variables are true, or all of them are false... There is no way that this kind of data can help predict the results of persuading marksman A not to shoot." I may come back to this thread in 2-3 days.

2022-12-07 14:48:27 @PHuenermund Why only "potentially suffer"? I've learned that, in observational studies, every method whatsoever requires untested assumptions to deal with unobservables. Is this not a golden rule in economics 101?

2022-12-07 14:19:03 @chroma_infinity I am not familiar. Can it be described in Tweetable space, using input-output language?

2022-12-07 12:27:44 @PHuenermund Do economists still believe propensity score matching helps reduce bias due to unobservables?

2022-12-07 10:14:30 @ya3ili @FranceskAlbs @EpshtainItay @UN What reputation? @FranceskAlbs has any?

2022-12-07 10:07:35 @AvivaKlompas Is this Jesus speaking behind the mask? He doesn't sound Jesus-like to me.

2022-12-07 09:53:43 RT @RozRothstein: Please retweet: Barcelona City Council: Your desire to cancel your sister city partnership with Tel Aviv is divisive. We…

2022-12-07 09:47:29 @_ambodi @DataSciFact Gelman has never given me a sign that he has an interest in causal identification of in selection bias. In fact, I do not know anyone from Rubin's school who has had such interests. Do you?

2022-12-07 09:11:24 @_ambodi @DataSciFact Thanks for posting my interview. But, What can we learn from Gelman's post?

2022-12-07 08:35:07 @DataSciFact What brings @DataSciFact to resurrect an ancient quote of Gelman about selection bias? A phenomenon that is now well understood, both statistically and causally

2022-12-07 08:05:44 @soboleffspaces in the process of academic stagnation. I might write about it some day, when I no longer have urgent questions to ask about the science of cause and effect. 2/2

2022-12-08 00:42:48 @Sam_J_Street This diagram may be helpful for the novice, but it is deceptive, in that it skips an important component in Bayesian analysis: Subjective knowledge. That's why I find the Billiard Table example (from Bayes' original paper) to be a better introduction to the issues involved.

2022-12-08 00:34:51 RT @yudapearl: @PHuenermund I have the feeling that economists are not aware of one result in https://t.co/jxWdLvbO1P: Given two arbitrary…

2022-12-08 00:13:41 @PHuenermund I have the feeling that economists are not aware of one result in https://t.co/jxWdLvbO1P: Given two arbitrary subsets Z and W, when would adjustment for Z be (asymptotically) the same as adjustment for W, namely, E_z P(y|x,z)=E_wP(y|x,w) We derived criteria for this equality.

2022-12-07 23:16:22 @_ambodi @DataSciFact So many people write that "causal identification is useful" that it's becoming hard to tell who is serious about it and who is just echoing hearsay.

2022-12-07 23:12:43 @GaryMarcus @johnsontoddr4 @TonyD993 Where is the text taken from? Statistics (1962) or Deep Learning (2022) ?

2022-12-07 14:48:27 @PHuenermund Why only "potentially suffer"? I've learned that, in observational studies, every method whatsoever requires untested assumptions to deal with unobservables. Is this not a golden rule in economics 101?

2022-12-07 14:19:03 @chroma_infinity I am not familiar. Can it be described in Tweetable space, using input-output language?

2022-12-07 12:27:44 @PHuenermund Do economists still believe propensity score matching helps reduce bias due to unobservables?

2022-12-07 10:14:30 @ya3ili @FranceskAlbs @EpshtainItay @UN What reputation? @FranceskAlbs has any?

2022-12-07 10:07:35 @AvivaKlompas Is this Jesus speaking behind the mask? He doesn't sound Jesus-like to me.

2022-12-07 09:53:43 RT @RozRothstein: Please retweet: Barcelona City Council: Your desire to cancel your sister city partnership with Tel Aviv is divisive. We…

2022-12-07 09:47:29 @_ambodi @DataSciFact Gelman has never given me a sign that he has an interest in causal identification of in selection bias. In fact, I do not know anyone from Rubin's school who has had such interests. Do you?

2022-12-07 09:11:24 @_ambodi @DataSciFact Thanks for posting my interview. But, What can we learn from Gelman's post?

2022-12-07 08:35:07 @DataSciFact What brings @DataSciFact to resurrect an ancient quote of Gelman about selection bias? A phenomenon that is now well understood, both statistically and causally

2022-12-07 08:05:44 @soboleffspaces in the process of academic stagnation. I might write about it some day, when I no longer have urgent questions to ask about the science of cause and effect. 2/2

2022-12-08 10:52:46 @Claire_V0ltaire @TankoZionism He is dead wrong. The Ethno-genetic claim of Jewish endogeneity did not start in 2022. It was recognized already by Balfour, in 1917 , see https://t.co/9XOCOsqkIT

2022-12-08 10:34:44 RT @VanJones68: When Black people and Jewish people get divided, democracy falters. When Black people and Jewish people come together, demo…

2022-12-08 10:20:40 Linda Sarsour, our fake feminist (see https://t.co/2nCKQxaZHX) is turning a fake Jew-lover. It doesn't occur to her that today's most marginalized community are Zionist Jews, yes, the same good Jews whom she adores and whom she tried to exclude from the feminist movement. https://t.co/elHYelh52H

2022-12-08 09:03:49 @richard_landes GERSHON BASKIN does not bring even one shred of evidence that Palestinians are now ready to be partners for peace. My litmus test has been, and still is, for Abbas to say the words "Equally Indigenous" (in Arabic, of course).

2022-12-08 08:36:29 As much as we hate to invoke the name "Hitler" in new contexts, the similarity between Qatar's Soccer Games and the Berlin's Olympic Games of 1936 did not escape us as we watched the former. This article provides a comprehensive account of the history of deception using sport.

2022-12-08 08:36:28 What Qatar Learned from Hitler’s Olympics https://t.co/GGgF21N4jD via @jewishjournal

2022-12-08 07:37:24 "Image and Reality" - The greatest counterfactual reversal of our era. https://t.co/WiaD5AEPVS

2022-12-08 07:27:34 Why has "drawing assumptions" become so threatening to scientists lately? Because when assumptions are CAUSAL, they are comprehensible. No one would criticize you for assuming Dirichlet priors

2022-12-08 07:12:00 @AvivaKlompas He doesn't even sound Jewish!

2022-12-08 07:04:53 @AleksanderMolak @PHuenermund This is an excellent paper. I can tell just by seeing its awareness of the Ladder of Causation. I do not agree however with the assumption that our works on Fusion https://t.co/E1ap1N4ZCh requires data on ALL variables

2022-12-08 06:07:08 Moving from the utopian to the practical, Ben Gurion actully followed Herzl steps and wrote (Nov. 1930): "In the same way that we want the Jewish people to be master of its own affairs,... so, too, we must seek for the Arabs." He would have been branded a "Traitor" by Ben-Gvir.

2022-12-08 06:07:07 Herzl Warned Us https://t.co/3vwilOEdsF via @jewishjournal

2022-12-08 05:23:38 I honestly do not understand @FranceskAlbs problem. If she is really concerned about her reputation as a fair-minded human-rights observer, all she has to do is to tell us whether Tel Aviv is stolen land or what kind of land it is! https://t.co/4SxLfWALVj

2022-12-08 00:42:48 @Sam_J_Street This diagram may be helpful for the novice, but it is deceptive, in that it skips an important component in Bayesian analysis: Subjective knowledge. That's why I find the Billiard Table example (from Bayes' original paper) to be a better introduction to the issues involved.

2022-12-08 00:34:51 RT @yudapearl: @PHuenermund I have the feeling that economists are not aware of one result in https://t.co/jxWdLvbO1P: Given two arbitrary…

2022-12-08 00:13:41 @PHuenermund I have the feeling that economists are not aware of one result in https://t.co/jxWdLvbO1P: Given two arbitrary subsets Z and W, when would adjustment for Z be (asymptotically) the same as adjustment for W, namely, E_z P(y|x,z)=E_wP(y|x,w) We derived criteria for this equality.

2022-12-07 23:16:22 @_ambodi @DataSciFact So many people write that "causal identification is useful" that it's becoming hard to tell who is serious about it and who is just echoing hearsay.

2022-12-07 23:12:43 @GaryMarcus @johnsontoddr4 @TonyD993 Where is the text taken from? Statistics (1962) or Deep Learning (2022) ?

2022-12-07 14:48:27 @PHuenermund Why only "potentially suffer"? I've learned that, in observational studies, every method whatsoever requires untested assumptions to deal with unobservables. Is this not a golden rule in economics 101?

2022-12-07 14:19:03 @chroma_infinity I am not familiar. Can it be described in Tweetable space, using input-output language?

2022-12-07 12:27:44 @PHuenermund Do economists still believe propensity score matching helps reduce bias due to unobservables?

2022-12-07 10:14:30 @ya3ili @FranceskAlbs @EpshtainItay @UN What reputation? @FranceskAlbs has any?

2022-12-07 10:07:35 @AvivaKlompas Is this Jesus speaking behind the mask? He doesn't sound Jesus-like to me.

2022-12-07 09:53:43 RT @RozRothstein: Please retweet: Barcelona City Council: Your desire to cancel your sister city partnership with Tel Aviv is divisive. We…

2022-12-07 09:47:29 @_ambodi @DataSciFact Gelman has never given me a sign that he has an interest in causal identification of in selection bias. In fact, I do not know anyone from Rubin's school who has had such interests. Do you?

2022-12-07 09:11:24 @_ambodi @DataSciFact Thanks for posting my interview. But, What can we learn from Gelman's post?

2022-12-07 08:35:07 @DataSciFact What brings @DataSciFact to resurrect an ancient quote of Gelman about selection bias? A phenomenon that is now well understood, both statistically and causally

2022-12-07 08:05:44 @soboleffspaces in the process of academic stagnation. I might write about it some day, when I no longer have urgent questions to ask about the science of cause and effect. 2/2

2022-12-08 10:52:46 @Claire_V0ltaire @TankoZionism He is dead wrong. The Ethno-genetic claim of Jewish endogeneity did not start in 2022. It was recognized already by Balfour, in 1917 , see https://t.co/9XOCOsqkIT

2022-12-08 10:34:44 RT @VanJones68: When Black people and Jewish people get divided, democracy falters. When Black people and Jewish people come together, demo…

2022-12-08 10:20:40 Linda Sarsour, our fake feminist (see https://t.co/2nCKQxaZHX) is turning a fake Jew-lover. It doesn't occur to her that today's most marginalized community are Zionist Jews, yes, the same good Jews whom she adores and whom she tried to exclude from the feminist movement. https://t.co/elHYelh52H

2022-12-08 09:03:49 @richard_landes GERSHON BASKIN does not bring even one shred of evidence that Palestinians are now ready to be partners for peace. My litmus test has been, and still is, for Abbas to say the words "Equally Indigenous" (in Arabic, of course).

2022-12-08 08:36:29 As much as we hate to invoke the name "Hitler" in new contexts, the similarity between Qatar's Soccer Games and the Berlin's Olympic Games of 1936 did not escape us as we watched the former. This article provides a comprehensive account of the history of deception using sport.

2022-12-08 08:36:28 What Qatar Learned from Hitler’s Olympics https://t.co/GGgF21N4jD via @jewishjournal

2022-12-08 07:37:24 "Image and Reality" - The greatest counterfactual reversal of our era. https://t.co/WiaD5AEPVS

2022-12-08 07:27:34 Why has "drawing assumptions" become so threatening to scientists lately? Because when assumptions are CAUSAL, they are comprehensible. No one would criticize you for assuming Dirichlet priors

2022-12-08 07:12:00 @AvivaKlompas He doesn't even sound Jewish!

2022-12-08 07:04:53 @AleksanderMolak @PHuenermund This is an excellent paper. I can tell just by seeing its awareness of the Ladder of Causation. I do not agree however with the assumption that our works on Fusion https://t.co/E1ap1N4ZCh requires data on ALL variables

2022-12-08 06:07:08 Moving from the utopian to the practical, Ben Gurion actully followed Herzl steps and wrote (Nov. 1930): "In the same way that we want the Jewish people to be master of its own affairs,... so, too, we must seek for the Arabs." He would have been branded a "Traitor" by Ben-Gvir.

2022-12-08 06:07:07 Herzl Warned Us https://t.co/3vwilOEdsF via @jewishjournal

2022-12-08 05:23:38 I honestly do not understand @FranceskAlbs problem. If she is really concerned about her reputation as a fair-minded human-rights observer, all she has to do is to tell us whether Tel Aviv is stolen land or what kind of land it is! https://t.co/4SxLfWALVj

2022-12-08 00:42:48 @Sam_J_Street This diagram may be helpful for the novice, but it is deceptive, in that it skips an important component in Bayesian analysis: Subjective knowledge. That's why I find the Billiard Table example (from Bayes' original paper) to be a better introduction to the issues involved.

2022-12-08 00:34:51 RT @yudapearl: @PHuenermund I have the feeling that economists are not aware of one result in https://t.co/jxWdLvbO1P: Given two arbitrary…

2022-12-08 00:13:41 @PHuenermund I have the feeling that economists are not aware of one result in https://t.co/jxWdLvbO1P: Given two arbitrary subsets Z and W, when would adjustment for Z be (asymptotically) the same as adjustment for W, namely, E_z P(y|x,z)=E_wP(y|x,w) We derived criteria for this equality.

2022-12-07 23:16:22 @_ambodi @DataSciFact So many people write that "causal identification is useful" that it's becoming hard to tell who is serious about it and who is just echoing hearsay.

2022-12-07 23:12:43 @GaryMarcus @johnsontoddr4 @TonyD993 Where is the text taken from? Statistics (1962) or Deep Learning (2022) ?

2022-12-07 14:48:27 @PHuenermund Why only "potentially suffer"? I've learned that, in observational studies, every method whatsoever requires untested assumptions to deal with unobservables. Is this not a golden rule in economics 101?

2022-12-07 14:19:03 @chroma_infinity I am not familiar. Can it be described in Tweetable space, using input-output language?

2022-12-07 12:27:44 @PHuenermund Do economists still believe propensity score matching helps reduce bias due to unobservables?

2022-12-07 10:14:30 @ya3ili @FranceskAlbs @EpshtainItay @UN What reputation? @FranceskAlbs has any?

2022-12-07 10:07:35 @AvivaKlompas Is this Jesus speaking behind the mask? He doesn't sound Jesus-like to me.

2022-12-07 09:53:43 RT @RozRothstein: Please retweet: Barcelona City Council: Your desire to cancel your sister city partnership with Tel Aviv is divisive. We…

2022-12-07 09:47:29 @_ambodi @DataSciFact Gelman has never given me a sign that he has an interest in causal identification of in selection bias. In fact, I do not know anyone from Rubin's school who has had such interests. Do you?

2022-12-07 09:11:24 @_ambodi @DataSciFact Thanks for posting my interview. But, What can we learn from Gelman's post?

2022-12-07 08:35:07 @DataSciFact What brings @DataSciFact to resurrect an ancient quote of Gelman about selection bias? A phenomenon that is now well understood, both statistically and causally

2022-12-07 08:05:44 @soboleffspaces in the process of academic stagnation. I might write about it some day, when I no longer have urgent questions to ask about the science of cause and effect. 2/2

2022-12-09 06:18:23 It's great they decided to listen to students! Next, I hope, they will listen to faculty! https://t.co/JQpOphlQQ9

2022-12-09 06:06:06 Upset by neighbors who vow to exist, International pressure they pledge to resist, Would Mr. Putin please join to assist? https://t.co/VBbE5iqCsV

2022-12-09 05:13:40 I lost tract