2022-03-17 16:31:07 @joftius Either way, thank you for getting them in early! 2022-03-17 16:30:36 @theduckiest But there isn't a magical cutoff between people who are the youngest year in Gen X and the oldest Millenials. 2022-03-17 16:16:55 @theduckiest Not to be provocative, but.... is 45 years old actually a different age bin from us? 2022-03-16 20:06:44 So, this is literally the only tweet I have ever been tempted to pay to promote. Please, if you are a FAccT reviewer, don't make me cry in 5 days. Do your reviews. https://t.co/VtWC6IOeiG 2022-03-16 19:56:35 If you know me, you know I love public defenders. So, here's a really cool opportunity to use your data skills to support one of the absolutely critical (and often under-appreciated) parts of the criminal justice system. https://t.co/iOv54zxgIL 2022-03-16 15:01:04 @binary_aaron found him! He really does blend in well 2022-03-16 14:54:49 @binary_aaron I don't see it! 2022-03-15 20:54:59 RT @sivavaid: We care way too much what rich people think. 2022-03-15 17:55:22 RT @jjvincent: look at this gorgeous thing a friend sent me in the post — a book of metric conversions from 1933 sold to traders and mercha… 2022-03-15 15:14:30 Day of rest? More like day of fillings and orthotics. Can I get a do-over? 2022-03-12 08:11:00 CAFIAC FIX 2022-01-24 20:08:32 With all of the attention COVID gets, I'd almost forgotten that colds are also a thing that really suck. 2022-01-22 02:01:09 @ubiquity75 Good try :) 2022-01-21 23:42:35 Kicking myself for agreeing to have the FAccT deadline on a Friday night. 2022-01-21 18:48:32 @mdekstrand Glad you caught that error! 2022-01-21 17:37:10 @L_badikho @ruchowdh Will do. Thanks for flagging! 2022-01-21 16:17:01 The #FAccT2022 paper deadline is coming up! Don't forget to submit your final versions before easychair turns into a pumpkin at 11:59 pm AoE today! 2022-01-21 15:55:33 @heysarah @Twitter Congratulations! 2022-01-20 16:23:24 @evijitghosh Fingers crossed that the COVID situation will be better and policies will have relaxed to reflect that by June! 2022-01-20 16:22:45 @laughsovich We don't talk about Bruno...... BUT! 2022-01-17 08:11:00 CAFIAC FIX 2022-01-12 16:22:34 RT @FAccTConference: Due to recent disruptions & 2022-01-12 16:08:51 @rajiinio Just to be clear, definitely not just me. But I'll take it! 2022-01-12 16:05:27 @struthious I'm not seeing it for some reason, but I'm glad it will be useful for people! 2022-01-12 15:52:58 You're welcome. :) 2022-01-11 08:11:00 CAFIAC FIX 2022-01-06 02:14:46 @JChrisBraun Great to hear. That’s reassuring. 2022-01-05 23:15:04 Clarification: I personally do not have it….. yet. Hoping to keep it that way. 2022-01-05 23:06:26 @brian_root I'd already resigned myself to the inevitability of this but it still sucks when it actually finally happens. 2022-01-05 22:55:29 Made it two years without anyone in my household getting COVID. And now, it's here. Wish us luck! 2022-01-04 17:02:20 @JubaZiani I guess it's a good way to not get a bunch of emails you don't want? 2022-01-03 20:21:28 @thosjleeper @kenhungkk It works! I should have googled this long ago. Thank you for doing that for me. 2022-01-03 20:11:37 @thosjleeper @kenhungkk Hold on, but what if I don't want it to open gmail either? What if I literally just want to be able to copy it? Is that allowed? 2022-01-03 18:32:58 @thosjleeper Huh, iiiiiiinteresting. So you're telling me I could do something about this instead of just complain on Twitter? :D 2022-01-03 17:45:40 People who put their email on their website with the link that opens a mail client rather than just writing out the darn email: whyyyyyyyyyyyy?? Please save me the extra right click --> 2021-12-30 19:51:44 @seanmylaw I mean, I do too. But also, I'll be like sitting there agonizing over a few sentences and thinking that realistically four people are ever going to read this. Why do I do this to myself?? 2021-12-30 19:34:18 @jmtroos let's never work together :) -- and I mean this not because it wouldn't be a joy to work with you but can you imagine rabbit holes we'd be spiraling down into? 2021-12-30 18:59:35 @Aaron_Horowitz Have you looked in the overleaf doc lately? So much tinkering is happening right now. So much. 2021-12-30 18:57:57 I feel bad for my co-authors because I am *the worst* when it comes to tinkering with and wordsmithing to absolute death all papers I'm involved in. 2021-12-30 18:44:39 RT @krvarshney: After an arduous journey over the last two plus years, I am happy to announce that I have written a 300 page book "Trustwor… 2021-12-29 19:12:00 RT @Aaron_Horowitz: Unoriginal take: so much CS ("AI") is just re-discovered stats and so much CJ is just really outdated stats. 2021-12-28 19:58:54 @struthious Oh yeah, forgot about the arm! 2021-12-28 19:47:02 Moderna covid booster review Terrible headaches, body aches, joint aches, chills, and mild nausea for 36 hours. Still better than rolling the dice with covid. 10/10 would do again. 2021-12-27 08:20:00 CAFIAC FIX 2021-12-21 02:47:47 @cklshorall It has a Molly as an accessory. That should be revealing about whether it’s you or not. Lol 2021-12-21 02:40:19 There is a “historical” American Girl doll from 1986. 2021-12-17 21:46:24 This program is incredible-- joint teaching in Data Science at U Chicago and community colleges. I benefitted so much from community college, and I know tons of other people do, too. https://t.co/ecL4G6lgLL 2021-12-17 17:35:47 RT @TheMichelleBao: summary: https://t.co/pexwmpvEWi this paper is a mixture of thoughts from @angelamczhou @samzottola @BrianBrubach @DrS… 2021-12-16 22:17:09 Wow, ok. So that was a lot of abstract submissions. https://t.co/1s8BLAShv5 2021-12-14 21:41:00 Reminder: if you're planning to submit to FAccT 2022, go ahead and submit that abstract now! Time's running out! 2021-12-14 01:39:31 @linzcpage Oooooih now that’s an idea! 2021-12-13 20:55:06 @MeganTStevenson I could see how this could be nice, too. My kid seems to act up and not listen to teachers a lot more if I'm present, so at least for right now, some separation would be a little better. 2021-12-13 19:00:38 Does this exist: a sports league where the kids and guardians both have practice simultaneously? E.g. kids play soccer and adults separately play soccer, or kids do gymnastics and adults do yoga. 2021-12-10 15:26:31 RT @jlkoepke: for the Friday twitter world: @TeamUpturn is hiring a senior-level Project Director to lead some our amazing economic justic… 2021-12-09 21:18:01 RT @binary_aaron: fuck you, jira https://t.co/GWIUsJd39L 2021-12-09 21:17:44 @binary_aaron I literally laughed out loud at that 2021-12-09 21:15:05 I don't like what it says about me that I kinda enjoy filling out paperwork, provided it doesn't take effort to figure out what to put in the fields. 2021-12-09 19:08:13 @zschutzy do I even want to ask?? What is this Meme ETF?? And will it make me want to invent a time machine to go back to the days before NFTs or whatever? 2021-12-08 22:13:39 @WellsLucasSanto I feel this. So much. 2021-12-08 22:12:33 @MilkaLichtblau Ooh, I wonder if it's that! I think this one is higher res than my old one... 2021-12-08 15:51:10 Found myself increasing the font size on all of my apps today. I think part of it is that I just got a new monitor and it ends up sitting a little farther away due to the stand, but.... is this middle age???? 2021-12-08 15:41:31 @alexhanna @ruchowdh @mims Maybe it has to do with when you adopted it? I didn't move there until I was 18. 2021-12-08 15:30:24 @ruchowdh @mims After 12 years I do, however, feel entitled to my very strong opinion that it "y'all" and definitely not "ya'll", which I've seen many native Texans write. 2021-12-08 15:29:44 @ruchowdh @mims I lived in the south for, oh, 12 years? And I still feel like a fraud when I say "y'all". 2021-12-07 15:52:43 @KacchaBaniyan I'm honestly kind of more impressed by that. The foundational stuff is legit. 2021-12-07 15:52:24 RT @eliaseythorsson: @KLdivergence https://t.co/6lbA9zWx5L 2021-12-06 22:20:36 @lisalendway Haha! Maybe! But I have the benefit of several decades of other advances on my side to pave the way. 2021-12-06 21:39:53 Whoever decided to take the USB port off the new version of the kinesis keyboard.... I'd like to have a word. 2021-12-06 21:37:04 @npparikh to be clear, in no world would I have been thinking of methods in the context of eugenics though. 2021-12-06 21:36:27 @npparikh No, I'd been taught all of those before I had a chance to figure them out myself :) 2021-12-06 21:14:58 @jepusto Yeah, @GermsAndNumbers will remember the time I invented propensity score weighting in the same week he invented hash tables. At the time I was bummed, but in retrospect, I independently came up with something developed in the 80s. Not bad! 2021-12-06 20:43:43 @Abebab Congrats, Dr. Birhane! 2021-12-06 20:43:14 I love when I think I've figured out a really cool stats thing and then it turns out that some famous statistician figured out the same really cool stats thing like a decade (or two) before I was born. 2021-12-04 00:14:58 @laughsovich Moana is so good. 2021-12-03 23:44:18 Eventful week at the bird place, eh? 2021-12-02 21:51:41 @karen_ec_levy @wsisaac @themarkup I literally didn't even see that typo. 2021-12-02 21:51:29 @karen_ec_levy @wsisaac @themarkup haha shit, I mean @wsisaac and I! Jeeeeez what is wrong with me today. 2021-12-02 21:49:53 @ASankin @theduckiest @wsisaac @RiceUniversity WHAT! Jones '06. That is awesome! 2021-12-02 21:37:26 @theduckiest @ASankin @wsisaac @RiceUniversity are there Rice people here that I'm unaware of?? I mean, besides me. 2021-12-02 21:30:52 @ASankin @wsisaac This part of the article was definitely one of biggest revelations. Super important! 2021-12-02 19:48:45 Almost six years ago, @wsisaac wrote an article using real data that showed how PredPol could exacerbate racial disparities in policing. @themarkup's release today shows that that is *exactly* what happened. https://t.co/u9yNXKvYeW https://t.co/f7OZFTbs5h 2021-12-02 19:21:58 RT @nealpatwari: Wow, congrats to this team for an amazing look behind the curtain! It's ironic that 6 years ago @KLdivergence, @wsisaac… 2021-12-02 15:50:19 @ProfFerguson I'm omnipresent on twitter now :) 2021-12-02 15:02:51 @wsisaac @ProfFerguson Also checking in to say hi @KLdivergence 2021-12-01 15:33:07 @mmbronstein Ha, yes, it should have been "increasingly" or "recently" 2021-12-01 15:32:28 @ErickGalinkin automatic accept, no revisions. 2021-11-30 23:56:24 @rajiinio @mdekstrand's got you covered 2021-11-30 23:55:50 @mdekstrand I knew I could count on you! 2021-11-30 23:52:56 I dare you to write an opening sentence of a paper that does not use the word "increasingly". 2021-11-30 22:39:42 @PawarBI @evijitghosh If it's back, let me know. I'm now dying to try some good biryani. 2021-11-30 22:32:53 @BoseShamik @evijitghosh @ruchowdh If you think of the name, please let me know! 2021-11-30 22:16:38 @evijitghosh I guess my question is... what is *good* biryani, then? Apparently I need to try it! 2021-11-30 22:10:02 @evijitghosh I will pop in with a westerner who loves Indian food but doesn't eat biryani perspective. I only tried it once and it was meh. It was just rice with a few things in it and it is expensive relative to other things that aren't like 90% rice, which kinda feels like a rip off. 2021-11-30 17:36:58 #dofewerthings 2021-11-30 15:18:52 Algorithmic. Justice. Specialist. Apply! https://t.co/Vb4aiAJrZz 2021-11-30 15:16:29 RT @achould: Heinz is an AMAZING place to work! Apply! 2021-11-29 18:45:10 @Navdeep_Gill_ probably with tears. But you'll soon recover :) 2021-11-29 17:56:54 @teemu_roos @ruchowdh @hackylawyER @paraga https://t.co/jJWsFzsLmM 2021-11-29 17:40:33 @ruchowdh @teemu_roos @hackylawyER @paraga Oh gosh, it's my moment... 2021-11-29 16:42:06 @samclifford < 2021-11-29 14:55:43 Mom content : Charging into her first day of preschool. Yes, I cried. https://t.co/aXzqAW41vW 2021-11-25 01:11:32 RT @quicola: I founded our RML initiative and hired the team we know today, however META was first founded by @naz_erkan and @__lucab in 20… 2021-11-23 15:15:33 @jbakcoleman @ruchowdh @hypervisible Welp, need a subscription to read it. Saves me the trouble of feeling frustrated this morning :) 2021-11-19 21:49:54 @npparikh Omg yes! 2021-11-19 21:31:13 Why is checking something off of a digital check list so much less satisfying than doing it physically with a pen on paper? 2021-11-18 21:14:36 RT @metallicbranch: New pub out in JESP! The presence of more women in a field causes people to be more likely to label it a soft science (… 2021-11-17 21:20:37 RT @SFBART: If you haven't mixed us, a public transit system, up with a nonparametric Bayesian regression approach which uses dimensionally… 2021-11-17 21:20:00 @WellsLucasSanto @SFBART @rabarretom BART? Yes. Just used it in a project actually. 2021-11-17 20:49:30 RT @roxannefelig: No one prepares you for the absolute high that comes with being able to help someone with their R code 2021-11-17 00:44:16 RT @lukestein: My daughter—feeling interrupted at dinner—explained that when she’s done talking she’ll let us know she’s “ready for questio… 2021-11-16 15:35:00 RT @heymatthenry: In a WHOSTP/@EPICprivacy panel, @KLdivergence just said that, in models meant to predict flight risk, “the signal is low,… 2021-11-16 14:57:36 Wearing a blazer for the first time in years. https://t.co/5uto4Jf1WX 2021-11-16 14:00:32 @evijitghosh Yeah I normally would have taken the train, but I was with my mom and we both had heavy suitcases, and the elevator at my stop has been broken lately. 2021-11-16 13:54:54 @evijitghosh Was it about the same price, too? That's what I found. I was very pleasantly surprised. 2021-11-15 20:26:27 You know how when you get an Uber at the airport there's that awkward thing where you wait outside and you frantically scan the license plates of the cars going by making sure not to miss your Uber? Yeah, I totally skipped that part of it by just taking a cab. It was glorious. 2021-11-13 15:57:18 @zschutzy How does it feel being a celebrity? 2021-11-13 13:19:00 RT @RaftoFoundation: "We have to be right. We dedicate ourselves to providing the most rigorous assistance for human rights defenders as po… 2021-11-13 01:02:32 RT @vaurorapub: “I’ll take ‘hyperscrutenization of women’s vocal mannerisms’ for $400, Alex” 2021-11-13 00:48:42 @Navdeep_Gill_ And thank you so much for coming! 2021-11-13 00:47:57 @Navdeep_Gill_ THE PAPER WILL BE OUT WHEN I CAN GET MANUSCRIPT CENTRAL TO COOPERATE. It's forthcoming (maybe eventually?) in JRSS-A. Here's a link to the arxiv version: https://t.co/GZwHEBeWfv 2021-11-12 22:13:39 #FreedBritney 2021-11-12 21:46:12 @npparikh OK, if median number overall is zero, then I'll up the ante. Of those that had at least one citation, I bet the medium number of citations by papers that use (and not extend) the method is still zero. 2021-11-12 20:42:30 @Miles_Brundage @huggingface WAVING hand, not waving HAND 2021-11-12 20:39:23 @Miles_Brundage @huggingface happy Friday! 2021-11-12 20:33:18 So I just now realized that @huggingface should be pronounced with the emphasis on the hugging, like hugging is the adjective modifying face. Not with the emphasis on the face like the popular 90s colloquialism “sucking face”. 2021-11-12 20:27:32 Oh and gravy. The magical mixture of flavors and spices that is the sauce in Indian food bears absolutely no resemblance to what I think of as gravy. 2021-11-12 19:29:00 @lizbmarquis do you know if there someone I could pay to get this manuscript uploaded to manuscript central? I mean, seriously. The amount of time I've spent on this is ridiculous. 2021-11-12 19:25:26 manuscript central might actually be the death of me. 2021-11-12 18:45:51 @Amherst73 Yes. Clearly the tail is long. 2021-11-12 18:32:34 @SMukherjee89 how many methodology papers have you written? 2021-11-12 18:31:21 Ahh of course there is a typo. Where's that edit button when you need it? Ignore the second "to tell anyone"... 2021-11-12 18:12:44 So honored to be giving the Blackwell Seminar at UW today. See you there? https://t.co/rSH0fGVBC7 2021-11-12 18:12:09 @IgorBrigadir @SemanticScholar I mean for me, not you :) 2021-11-12 18:12:01 @IgorBrigadir @SemanticScholar maybe a side project 2021-11-12 18:07:26 @IgorBrigadir @SemanticScholar So what was the answer? 2021-11-12 17:59:55 Has anyone done on a study on the median number of citations a stats or ML "methods paper" gets from papers that actually use the method it introduces? I'm specifically excluding other methods papers that build on the original method. I'd be shocked if it's > 2021-11-12 17:57:33 @alexhanna OK, well I can't argue with that. 2021-11-12 17:55:07 @alexhanna no 2021-11-12 17:43:49 @npparikh lol, maybe something like "delicious Indian style cheese". But I'm not a marketer. I await people with more expertise in this area to come up with some great descriptions. 2021-11-12 17:39:34 @npparikh I'd probably just go with cheese tbh. 2021-11-12 17:37:52 @sameersoi Totally! What Americans typically call cottage cheese is infinitely inferior to paneer. Honestly it would be pretty gross subbed into any of the dishes I've had paneer in. 2021-11-12 17:32:43 Far be it from me to tell anyone how to tell anyone how to run their business, but paneer is possibly the most delicious cheese ever and I think more Americans would be into it if it wasn't always described as cottage cheese on menus. 2021-11-11 18:57:01 RT @hannah_kerner: Today I presented "Troubling Trends in Machine Learning Scholarship" (@zacharylipton, Jacob Steinhardt) to my group and… 2021-11-11 17:15:28 So happy for all the parents of kids finally getting vaccinated! And also slightly jealous. Can’t wait for the under-5s to be eligible! 2021-11-10 19:46:33 @rcalo Omg Ryan calo I both love and hate you 2021-11-10 19:39:31 @juliasilge @ruchowdh Yes. Though people typically give things like racial identity which can be…complicated. Hence the identity crisis :) 2021-11-10 19:27:18 @juliasilge See @ruchowdh’s reply for an example 2021-11-10 18:52:05 Shout out to all the people who go into a full blown identity crisis when asked to introduce themselves on a video call with a physical description. 2021-11-10 12:30:46 RT @wsisaac: Amazing Fellowship Opportunity I'm so excited to finally talk about this program!! I strongly believe the Just Tech Fello… 2021-11-09 17:12:26 @stuartbuck1 2021-11-09 17:04:26 @stuartbuck1 You work at a foundation. (Or… I guess not quite anymore??) Of course you do :) 2021-11-08 11:58:59 RT @heymatthenry: @KLdivergence Mourning is so apt for this. We did a similar move this year, and I think (hope?) it was the right thing fo… 2021-11-07 22:41:26 @Canadian_JACD Oh it’s so hard. I hope things are getting better for you 2021-11-07 19:24:38 @heymatthenry Same. I love my new job but still find myself mourning that other life and the academic career my husband worked so hard for. 2021-11-07 18:31:12 I definitely had some good times here. It wasn’t all bad. I had nice, smart colleagues and a beautiful house, but I think the overarching feeling I remember in coming back is isolation and loneliness. 2021-11-07 18:31:11 Back in Philly one last time to move out of my house & 2021-11-06 23:20:00 CAFIAC FIX 2021-11-01 19:20:00 CAFIAC FIX 2021-11-01 17:30:00 CAFIAC FIX 2021-08-19 18:39:15 @nealpatwari Or, maybe, "who has the power to audit, and who has the power to edit" :) 2021-08-19 17:56:53 @binary_aaron I'm still kind of shocked this article didn't get more attention at the time. It's just so unethical. 2021-08-19 17:50:57 @binary_aaron A similar thing has been documented as happening with this lie detection software. They changed the hyperparameters when they don't like the answer and voila, the model tells police what they want it to tell them. https://t.co/cz8axbA0Bd 2021-08-19 16:02:42 @lawrennd @ChengSoonOng I think that's always the lesson. Observational data can be misleading. :) I'd love to see a similar analysis disaggregated by time since vax and age. 2021-08-19 15:30:22 @ColinDoyleLaw I love this! And I hope T Rex had some good questions. 2021-08-19 14:59:54 Shotspotter employees change model outputs at the request of police. Makes it pretty clear that you can just pay a company to have "neutral tech" to provide whatever "evidence" you want. Shady, shady stuff. https://t.co/rPKS8GsFRO 2021-08-19 14:36:15 @lawrennd @ChengSoonOng I agree that this isn't the traditional use of Simpson's paradox because of the additional layer of applying Bayes's rule. But I think it's revealing the same underlying phenomenon so I'm good with it. 2021-08-18 15:14:55 RT @jennifernvictor: God, I love it when other disciplines discover social science. https://t.co/xyPMK4agi3 2021-08-18 13:35:59 @LisaDeBruine @NoahHaber @Corey_Yanofsky I mean anything’s possible but the only ones who benefit here are the authors so I’m inclined to suspect them first and foremost. 2021-08-18 13:34:57 @LisaDeBruine @NoahHaber @Corey_Yanofsky If that were the only thing it would be much easier to just simulate noise all around rather than noise but with a big effect size for the research Q. It would also hide their fabrication better because the work would get much less attention. 2021-08-18 13:32:29 @LisaDeBruine @NoahHaber @Corey_Yanofsky What incentive would they have to do this?? The company is anonymous so it’s not like they get publicity. I guess you’ve identified a single employee covering up that they didn’t complete their work…. But why fabricate something that matched the authors’ research hypothesis? 2021-08-17 22:56:39 I would read the shit out of a journal that literally just had articles describing examples of Simpson’s paradox arising in the wild. https://t.co/2pkqznLIGb 2021-08-17 22:04:44 @gshotwell I was thinking about doing a tweet thread on how to fabricate data, considering I used to do it professionally :) (But, like, ethically....) 2021-08-17 21:57:39 @gshotwell Yeah, though getting too many duplicates could be a red flag so you might want to add some noise of some kind anyway. But, yeah, there are so many better ways. And the uniform is almost insultingly bad. 2021-08-17 21:44:14 If you're going to fabricate data, at least have the decency to not use a UNIFORM distribution! At least use a normal, or, like a mixture of a normal and something heavy-tailed. Sheesh! 2021-08-17 21:34:54 THE PLOT THICKENS https://t.co/X9a3vXQh5o 2021-08-17 15:47:54 @binary_aaron Yes, the font thing was my favorite part! 2021-08-17 14:57:54 My favorite genre of blog post: bullet-proof evidence of simulated data masquerading as "real" data. https://t.co/wXSkAHFiAr 2021-08-11 14:37:18 RT @MineDogucu: Where is my Berry Consultants pen this year? Did I miss it in the mail or something? #JSM2021 2021-08-11 14:37:09 @StatStas @AmeliaMN @MineDogucu @bradytwest Haha, you are repping HARD for AAPOR :) 2021-08-10 18:09:18 @CarriquiryAl @DevinCow I have literally never once woken up and thought to myself "hmmm I wonder what Alicia Carriquiry has to say on this topic." With that in mind, please take this as an invitation to mute me, so you don't feel the need to respond to my tweets anymore. 2021-08-10 17:34:38 Woke up to a reply from Devin Nunes's Cow @DevinCow then spoke in a #JSM2021 session where my and @5harad's talks complimented each other perfectly. Who slipped me the felix felicis? 2021-08-10 16:44:20 @sherrirose @BhramarBioStat should we maybe... put that thing we wrote up on arxiv? 2021-08-10 15:00:28 RT @weatherbuzzword: Thanks @KLdivergence for #JSM2021 talk today on analyzing election fraud claims. Also see https://t.co/jyUPjvPHan Some… 2021-08-10 14:26:54 @StatStas @jbakcoleman @KabirCreates @AAPOR @bradytwest @jbakcoleman, all you? 2021-08-10 14:15:34 @sherrirose @jbakcoleman @KabirCreates Collaborate with me and you might just get your wish! 2021-08-10 14:15:07 @jbakcoleman add it two your Twitter bio :) 2021-08-10 02:26:57 @bzaharatos 2021-08-10 02:23:28 @bzaharatos Hopefully by then you’ve saved enough money to take some time off and figure out something to do that isn’t bullshit while still being able to afford college for your kid…. So that she can try way too hard only to come to the realization herself that it’s all bullshit. 2021-08-10 02:04:51 @doomsuckle I’ll take one midlife crisis to go please! 2021-08-10 00:29:09 Talking at #JSM2021 tomorrow about a project I did with two complete randos I met on Twitter (@jbakcoleman and @KabirCreates) when we connected over trying to push back against crappy analyses of election data from the recent US presidential election. https://t.co/9udpsZT5Q4 2021-08-10 00:09:28 So proud of my colleagues at META for putting this together and huge congrats to @hiddenmarkov, the big winner! https://t.co/RRcJpRvLGB 2021-08-09 23:33:04 @binary_aaron I think the prospect of going back into lockdown over winter and re-masking up (have yet to find a mask that doesn't fog my glasses) is getting me down lately. 2021-08-09 22:53:06 @zschutzy just heat it up 2021-08-09 22:50:26 @zschutzy that does not look like soup 2021-08-09 22:44:57 @LMiratrix idk, man, I think it's probably a better work/life balance. 2021-08-09 20:11:17 @dnlmc Totally. There are many ways one could spend their time on this planet that are bullshit. 2021-08-09 18:34:06 RT @willplayforfood: @KLdivergence Yes! There are so many artificial dichotomies in our dialog about work (hussle vs slacking, mission vs… 2021-08-09 18:01:41 @willplayforfood LOL, I think 35 year old me is confused by 36 year old me. Hopefully I'm keeping it interesting for 37 year old me. 2021-08-09 17:52:48 @forestgregg Congrats! 2021-08-09 17:36:35 @Aaron_Horowitz You are always prescient, Aaron. 2021-08-09 17:17:59 No matter how far you've made it in academia, there's no shame in deciding it's all bullshit and just.... doing something else. 2021-08-06 23:07:57 @ruchowdh I know! I meant how did I fail to show up 2021-08-06 20:15:25 @ruchowdh Whaaaat?! How did I miss that??? 2021-08-06 20:13:31 @lzamparo Honestly, that seems inappropriate to me. Bad and thoughtless reviews are a problem but so is bullying and public shaming like that. Not cool. 2021-08-06 20:07:52 @ruchowdh tempted to change my bio to "puritan activist larping as tech savvy victorian moralist" 2021-08-06 20:06:38 @lzamparo I mean, I commend you for that! It's certainly a service to the profession to be so thorough and a kindness to the authors dig out the good in a piece of work that doesn't make its contributions clear. 2021-08-06 19:55:28 @GordonMcGregor Ummm, excuse me? 800 papers? You sure you're not off by 2 orders of magnitude? :) 2021-08-06 19:22:36 @thejonullman Oh, relative to what I think people *should* be spending... yeah, I think 5 hours is too long. There are obviously exceptions, but I think a couple hours is generally totally sufficient. Sometimes even less, especially if it's in the left tail. 2021-08-06 19:05:41 @thejonullman I think people are spending less than 5+ hours on each review. 2021-08-06 17:50:03 @zschutzy 2021-08-06 15:46:43 @teemu_roos Yeah, i've seen some truly outlandish numbers in there as well, but I think 17 is even higher. 2021-08-06 15:28:38 @code_star slackers! *Only* 3-5 hours on a 9 page document. 2021-08-06 15:20:00 So, for real, are folks really spending 5 hours + on each NeurIPS review they write? I find that.... kinda hard to believe. 2021-08-06 15:17:17 Nightmare fuel. https://t.co/zKunTU6fiu 2021-08-05 19:13:29 @thegargiulian Incredible. Someone else has always thought of it first. :) 2021-08-05 19:09:17 Bonus if you can make some of them colorblind friendly! 2021-08-05 19:02:56 You find the most random stuff on Twitter. Someone please make an R package with fish-inspired color palettes. https://t.co/BSqxDNxVfI 2021-08-05 15:44:01 It's back! Attended (and spoke on a panel at) my first NeurIPS workshop last year and it was this one. Not only was it super interesting (and had a catchy name), but I felt super welcomed to the community. A great first NeurIPS experience. < 2021-08-04 03:16:57 @Ira_Sass @KendraSerra @ToddTheLinguist Honestly you’re not selling it very well with this concussion talk 2021-08-03 21:27:52 @Ira_Sass @KendraSerra @ToddTheLinguist Hold on a second there! Your profile says EX roller derby player.... 2021-08-03 20:01:07 @Ira_Sass @KendraSerra @ToddTheLinguist ahahahahahahahahahahaha you apparently have not seen my roller blading progress... 2021-08-03 19:25:36 @KendraSerra @ToddTheLinguist I would KILL for a nickname that good. 2021-08-03 15:37:18 RT @MartinJBCoulter: For the past few months, @HughLangley and I have been investigating bullying allegations against Deepmind cofounder Mu… 2021-08-02 19:26:41 @wsisaac @DeepMind So happy for you, friend! You deserve it! 2021-08-02 17:21:26 @SMukherjee89 Ha! Tell your friend to go for it! Just maybe give me a cut when they become the next big thing :) 2021-08-02 16:17:08 @Just_Nag_Em Can I still get them? 2021-08-02 16:16:34 @alexhanna Not exactly? I feel like tap dancing shoes have a different sound (higher maybe?) than heels. 2021-08-02 16:10:29 Big idea: Comfy flat shoes that make that high heel click-clack sound when you walk. 2021-07-30 16:12:18 So much of what I do when I write is cut and paste small thoughts and well-worded phrases into different places until a coherent narrative starts to emerge. How on earth did people do this before word processors? 2021-07-30 14:59:00 RT @ruchowdh: We are launching the first ever algorithmic bias bounty tomorrow at 9am PT! Learn more about it and hear from our judges at o… 2021-07-30 14:42:08 @a_ramach Another one of mine is very similar: "How about this one?" 2021-07-30 14:38:06 @zacharylipton why not? I'm extremely cheugy. 2021-07-30 13:09:53 @TheMichelleBao @Aaron_Horowitz Ok I didn’t have sound on before and now that I heard it with sound I TOTALLY get it. Also I did #carlife for like 3 months after my PhD so yeah, def get it. I hope this time is rejuvenating for you! 2021-07-30 03:05:56 @a_ramach Haha! That’s another one of mine too! 2021-07-29 22:34:27 Nothing like having a toddler to make you realize the phrases you use way too much. Today's phrase: "That'd be fun!" 2021-07-29 19:46:36 @Aaron_Horowitz @TheMichelleBao The olds don't get it, I guess. I also do not get it, minus the vanlife aspect of it, which I totally get and am jealous of. 2021-07-29 18:59:06 RT @samsaranc: Calling all Python/Data Engineering wizards!! Do you care about saving the Earth and want to use your tech expertise for… 2021-07-26 21:45:51 RT @JoyBonaguro: Omigod omigod omigod! I have some news! We are HIRING! The Office of the Chief Data Officer of #California (that's me :-)… 2021-07-26 15:12:12 @kdphd @WHOSTP @AlondraNelson46 @geomblog Awesome! 2021-07-24 01:46:06 @struthious I would be lying if I said that that possibility didn’t cross my mind. 2021-07-23 21:11:25 @quantadan Yep! Ready for the weekend and hoping it's calm and uneventful. 2021-07-23 21:05:32 Tues: man randomly spits on my face and bikes away. Thurs: man forcibly takes a package from me to "help me carry it" (despite telling him not to many times), follows me to apt lobby & It has been a WEEK 2021-07-22 22:09:31 @FarbodNezami @Abebab Yeah. I think everyone who saw/heard it felt that way. It was one of the most surreal professional experiences of my life. 2021-07-22 22:03:57 @itsclarkwithane @Abebab @FarbodNezami I mean, I wish I'd been taking notes because I can't even reproduce his argument relating the two it was so bizarre. But, for the sake of relaying the absurdity of the situation, I wish I'd written it down. 2021-07-22 21:40:31 RT @KLdivergence: @Abebab @FarbodNezami I once was at a meeting where a dude suggested WITH A STRAIGHT FACE that we should be concerned abo… 2021-07-22 21:12:58 @Abebab @FarbodNezami He then likened our lack of concern for his napkin "AI" to slavery, and I cannot think about that meeting so many years ago without my blood pressure going through the roof. 2021-07-22 21:10:42 @Abebab @FarbodNezami I once was at a meeting where a dude suggested WITH A STRAIGHT FACE that we should be concerned about the rights of his "AI", which was a decision tree for playing Monopoly that was so shallow you could write it on a napkin. 2021-07-22 16:13:26 @dan_p_simpson shit, nobody told me about 3... 2021-07-22 15:53:07 Came across this SEO for academic publications guide, and -- not saying I'm above it -- but it feels so .... counter to the romanticized version of academia many people imagine. https://t.co/zjvtGKUXnE 2021-07-21 19:40:36 @sanedragon @generativist Yes! That is absolutely weird. :) 2021-07-21 19:18:02 @christine_su these are gorgeous! 2021-07-21 19:17:42 @DrLukeOR Thanks for saying that. Yeah, I didn't realize until after the fact, but you're absolutely right-- it was really scary! I'm good now though. 2021-07-21 19:16:43 @nkrishnaswami wow, that cauliflower is beautiful 2021-07-21 16:03:09 @laureneoneal Wow. 2021-07-21 00:46:13 @AnisDamha I’m pretty sure it is. 2021-07-21 00:21:05 @sakrejda Yeah. No clue what his issue was with me. Really bizarre and disgusting. 2021-07-21 00:19:05 @code_star Idk man. Having experienced both, its pretty fucking disgusting. I think I’d take a day of revisions over that. 2021-07-21 00:13:31 Yesterday I got a peer review that requested no changes. Today a random dude, completely unprovoked, hocked a loogie right in my face on my walk home from work. Hey universe, next time I’ll take a few nitpicky comments if this is what it takes to cosmically balance my luck. 2021-07-20 21:54:42 @BaiocchiMike I'll wait to commit to that until it is fully accepted because this is too good to be true. A definite career first for me! 2021-07-20 14:48:40 RT @hemantmehta: This would be one hell of a day for MacKenzie Scott to announce another multi-billion-dollar set of donations. 2021-07-20 14:43:13 @ShariDeBaets 10% writing a damn good paper, 90% luck. 2021-07-20 02:29:19 @ruchowdh I know how I’m going to introduce myself in meetings from now on…. 2021-07-19 21:58:05 Found out from my two-year-old's daycare teacher that on her first day at her new school, she made her grand entrance to meet her new classmates by roaring loudly like a dinosaur. God, I hope she never loses that confidence. 2021-07-19 19:39:55 @ryanstravis 2021-07-19 19:36:08 @ruchowdh @RDBinns I think @s010n was doing some work to better define what "proxy" means. Maybe he'll have some refs? 2021-07-19 18:50:11 @bll_krtl Though there was, numerically, a Reviewer 2, I did not get a "Reviewer 2" on this one. Nominal Reviewer 2 had some suggestions for things we should cite but didn't request any major changes. 2021-07-19 18:48:21 Literally a review I just got on a paper I submitted: "This could have been one of the shortest reviews I have ever written: I think this paper is excellent and I cannot think of anything I would suggest to improve it." Miracles do happen. Thank you, Review 1. Thank you. 2021-07-19 15:23:47 @RDBinns @ruchowdh All I can say is this is a great question. I haven't seen it couched specifically in terms of monotonicity, but I bet there's some interesting work to be done examining that definition. I think the implication is usually more about correlation. 2021-07-16 21:15:53 @daniela_witten @onepeloton Own it. I am often "that person" (though not when it comes to Peleton that's a bridge too far) 2021-07-16 20:42:13 @hpcprogrammer This is true! I caught a glimpse of myself in the mirror the other day and realized my arms are jacked from picking up a wiggly 27 lb. weight so much. 2021-07-16 19:24:09 One of the few times I'm actually hoping to get ratioed :) Thanks, all, for all the tips. I'm going to try to compile some of these into a condensed thread later. https://t.co/aH32LK6vPq 2021-07-16 15:25:43 @lintool Is this not..... dangerous? I'm guessing maybe your kids are older than toddlers? 2021-07-16 15:25:16 @taraskaduk @nataliexdean My kid is two, but really, I'm asking anyone who's making it work with the constraints that having kid puts on one's schedule. 2021-07-16 15:23:40 @ajamesgreen Great idea! 2021-07-16 15:23:02 @AKrenelkaChase That is so hardcore! I'm so impressed you're able to do that. 2021-07-16 15:22:15 @hnorcross @nataliexdean Honestly, no. She hates being strapped into anything. I think she also gets motion sickness in the car, which doesn't help. I like the idea of giving her rewards every 10 minutes or so. Maybe I can make a positive association with riding in the stroller for her. 2021-07-16 15:07:48 @hnorcross @nataliexdean She's 2.5. She's never liked it. When she was little she just fussed to fuss, but now I think she doesn't like it because she wants to get out and run around herself. 2021-07-16 15:06:44 @dmacdonald85 @ShawnRadcliffe What is fasted cardio? 2021-07-16 15:06:11 @DederichLeda Thankfully I have a cooperative partner and a very flexible work situation! It just seems like there aren't enough hours in the day, and any time I've tried to exercise at home (e.g. yoga) it just doesn't work because my kid wants to "participate." 2021-07-16 15:05:00 @KevinKaichuang I actually kinda like 3. Maybe I could find an activity I like so much that it is preferable to sleep occasionally. Also I like the idea of working out during meetings where you're just listening. It's rare that happens for me but that's a good one. 2021-07-16 15:02:06 @SouthwarkBelle This is such a good point! 2021-07-16 14:57:46 @transportdata @kingsushigino Fair enough! 2021-07-16 14:51:38 @laurenfklein Alright, mine's 2.5 Maybe I'm almost getting there. The fact that I'm even asking this question might suggest that's the case, as a year ago, I don't think exercise was even on my radar. 2021-07-16 14:49:37 @kingsushigino OK, so you just get it in in bits and pieces throughout the day? Do you bike commute even in bad weather? 2021-07-16 14:48:44 @nataliexdean Yeah, stroller walks are pretty much all I've got, and my kid really doesn't like riding in the stroller, so I get about 15-20 minutes before the screaming begins. And, walking doesn't quite get those exercise endorphins flowing like I'm looking for anyway. 2021-07-16 14:46:49 @c_f_baumgartner I might have to try the lunch break jog. But, then aren't you all sweaty when you come back and when do you actually eat?? 2021-07-16 14:44:23 @jenwilson Problem for me is my daughter is a darter. She's very fast and has no fear of running straight into traffic. She's also very stubborn and does not particularly enjoy holding hands. Do you use a stroller for this? And, does your little one like the stroller? 2021-07-16 14:43:09 @LaurentFranckx Hahaha, at this point it seems to be this is going to be my solution, whether I like it or not. 2021-07-16 14:40:12 People with young children who work full time: how do you find time to exercise? It seems like it comes down to picking between sleep and exercise (faced with that decision, I'll always pick sleep). Tips actually, genuinely appreciated! 2021-07-15 17:42:42 @__lucab @ruchowdh Ha! this guy admits that it's pretty good! 2021-07-15 17:27:00 @__lucab @ruchowdh hold on... you're not one of the anti-pineapple-on-pizza people, are you? 2021-07-15 16:12:31 @evijitghosh That was one of the things I really appreciated about the application process when I applied also! 2021-07-15 14:56:53 @ecsalomon Oooh interesting. I guess I need to come up with a personal trademark reaction emoji... hmmmm... what should it be?? 2021-07-14 22:47:31 @ruchowdh PLEASE NO! I can barely keep up as it is! 2021-07-14 22:43:02 @chrisalbon A companion card defining alleatoric and epistemic uncertainty would be a nice compliment to this one. 2021-07-14 22:30:42 @chrisalbon So, like, measurement error? 2021-07-14 21:44:48 @chrisalbon What does it mean for crime to be deterministic? 2021-07-14 15:37:22 @__lucab I honestly don't even know what this is supposed to mean.... like, is the pizza really bad? Or are you drowning your sorrows in a slice of pizza? 2021-07-14 15:35:44 @binary_aaron I wish I could screenshot some of my other slack workspaces for you. It is a desert of emoji reactions. I'm seeing about 1 reaction per 8 messages or so, and it is usually just 2021-07-14 15:16:31 @Aaron_Horowitz I don't know if this is a paid account thing, or I just never noticed them before. But, the possibilities are endless with the available emoji reactions I'm seeing now. 2021-07-14 15:13:23 Really have to up my slack emoji reaction game now that I'm in tech. 2021-07-13 23:49:11 @ruchowdh Like why is there so much interesting stuff going on in the world and why oh why did all the criminal justice data ive been trying to get for literally years become available just as I’m winding down in that area?? 2021-07-13 23:46:35 Is it normal to both be super excited about your current work but also have serious FOMO (the kids definitely do not say that anymore, right??) for aaaaaaaallll the other stuff that’s going on? 2021-07-13 21:17:22 One more important detail: META stands for Machine learning Ethics, Transparency, and Accountability. 2021-07-13 20:37:33 @IEthics No. We're looking for masters and PhD students or recent grads. 2021-07-13 20:26:09 @AlexCEngler @ruchowdh @UChicagoCAPP @MassiveData_GU @McCourtSchool Thank you! 2021-07-13 19:54:06 FWIW, I'm loving it so far. Happy to respond to DMs if you have any questions. 2021-07-13 19:39:03 We're looking for fall interns. Come work with me at Twitter META! https://t.co/NY8O9xMQOF 2021-07-13 16:20:28 RT @brian_root: Amazing JOB ALERT. We are hiring our first ever Manager of the Digital Investigations Lab. Come lead our research team… 2021-07-13 01:18:48 RT @Aaron_Horowitz: AI bias folks: in a lot of ways, this is mostly translation work of your efforts. Would love for people to give it a re… 2021-07-12 23:09:03 @alexdamour I only regret that I have but one life to lose for petty terminology debates. 2021-07-12 23:04:36 @AlexCEngler I'd be totally ok with that as an "experiment." I have no beef with simulated data in general at all. I guess I'm more talking about things that I'd call toy examples here. 2021-07-12 22:55:24 @normdeviate Hold on, i see you're a statistician too. Maybe I'm viewing my grad school days through rose colored lenses. I think part of it was that the point of model development was genuinely for answering a scientific question, more often, rather than just beating some benchmarks. 2021-07-12 22:51:55 @normdeviate I come from statistics, so maybe it's a bit different there. But, honestly, I don't recall describing what we were doing as experiments when we already knew the answer. Maybe synthetic examples? Toy examples? I don't know, I'd have to go back and read some old papers. 2021-07-12 22:47:50 @willplayforfood Yeah, I don't think rigidity just to be stubborn is at all helpful (see, e.g., my recent adoption of the word "AI" for everything under the sun), but it really bothers me when I feel like the intent of a word choice is to be misleading. 2021-07-12 22:42:37 @willplayforfood I'd love to get you started on that. Maybe next coffee hour. :) 2021-07-12 22:37:07 @bearloga I'll expand on that. the difference I see is that my misuse of the term is just due to ignorance, whereas I suspect the overuse of the word experiment is meant to give the illusion of scientific investigation or impartiality where there is none. 2021-07-12 22:35:46 @bearloga haha, yes, YTA :) 2021-07-12 22:34:25 @IgorBrigadir Yes, or even more transparently, examples where there are like 6 data points and they are obviously chosen specifically to make a point. I don't mind people using them-- they are absolutely useful-- but in my mind, running your code on data like that is not an "experiment." 2021-07-12 22:26:04 To me, the word experiment connotes the intention to actually learn whether something is true, not demonstrate something you already know AITA for being cranky about machine learning papers that do "experiments" on data designed purposefully to showcase the merits of the method? 2021-07-12 16:11:34 @Navdeep_Gill_ @acidflask Ha! I was going to say something similar. Someone should really make a dog retching ring tone to use as an alarm. 2021-07-11 15:20:07 @daniela_witten There is no real maternity leave policy for research faculty at Penn. tenure track get teaching relief but research faculty don’t have any guaranteed reductions. This was absolutely something that gave me pause before accepting a job there. 2021-07-11 00:58:31 @YairZick @zacharylipton Good point! But yeah I’m not convinced that that is a case where the ai isn’t doing exactly what’s intended. 2021-07-11 00:54:43 @zacharylipton Wouldn’t it make sense to cut to ads at places people are unlikely to just leave once the ad starts?? I mean not from the POV of the person enjoying it but from the POV of getting them to watch it. 2021-07-10 22:49:08 @KennKnowles Might be!! 2021-07-10 22:44:23 @KennKnowles Lol. Are you sure they’re scheduled for more than an hour?? I noticed that meetings often ran over. 2021-07-10 17:03:06 @_joaogui1 Oh I’m just sensing a disturbance in the force 2021-07-10 17:02:16 @kat_heller Yikes. This is taking a very dark turn. I literally just meant that people seem exceptionally (and maybe slightly exaggeratedly?) happy. I’m sorry the situation is so bad :( 2021-07-10 17:00:58 So uhhhhhh the googlers are not ok, y’all 2021-07-10 14:14:29 @_joaogui1 Would have been much better with songs! 2021-07-10 14:14:08 @geomblog Yes. Among others. 2021-07-10 14:03:38 Why even bother making a Disney movie if there aren’t going to be songs in it?? 2021-07-09 22:23:14 @nkrishnaswami @mmitchell_ai Honestly, it's been something I've been trying to override in myself. I think (!!) I default to "I feel like" or other qualifiers, even when I know something to be true. Socialization is hard to override, though, and I still do it way more than I'd like to. 2021-07-09 22:16:48 @andrewthesmart @mmitchell_ai Fortunately, my instinct is to overuse exclamation points, and it's something I've had to intentionally temper in the past. Maybe I can finally use them to my heart's content!!!!!!!!!!!! 2021-07-09 22:04:07 godDAMN the tone of emails is extremely chipper in the corporate world 2021-07-09 22:00:35 @vm_wylbur nerd alert! 2021-07-09 21:58:40 @GermsAndNumbers May I propose half hour meetings for those remaining collaborations? 2021-07-09 19:09:28 @delagoya Nope! I'm at Twitter now! 2021-07-09 18:56:05 @SushGopalan @_brohrer_ @AngeBassa I hope I'm not giving away any trade secrets here, but this is the norm at Twitter-- meetings start at :05 & 2021-07-09 18:21:18 @mmitchell_ai Let me have my honeymoon period, M!! 2021-07-09 17:22:29 One big difference between tech and previous jobs I've noticed so far is the default to the half hour meeting. It's like a firehose, but I'm never going back. Hour-long meetings: never again. 2021-07-09 12:36:02 @merbroussard Jonathan Godfrey at Massey University in New Zealand. I don’t think he’s on Twitter. 2021-07-09 00:12:37 Here’s a hint for how not to pronounce my name https://t.co/GtCPvj9VWO 2021-07-08 15:46:31 RT @kjhealy: I'll be offering a 3-day "Data Wrangling with R" again in Sept. Last time was pretty enjoyable (my slides here notwithstanding… 2021-07-08 14:06:54 @austinvhuang Is it paw-sterior? I certainly hope not! 2021-07-08 02:42:51 @undersequoias @ruchowdh @geomblog Hahahaha that’s terrible but I’ll accept that pronunciation 2021-05-22 15:50:47 RT @pamela_herd: Human subjects ethics require we balance risks and benefits. There has *never* been a case of individual identification in… 2021-05-21 18:36:22 @MarkSendak I guess I should clarify that I have nothing but respect for Daniel Kahneman's work. In this case, I think he's applying some of his (admittedly laudable work) to an area in which he hasn't fully considered or appreciated some important dimensions. 2021-05-21 16:22:22 @acidflask Link? 2021-05-21 15:53:00 Ultimately, in order to trust an AI system, you first have to build trust in the people and institutions using it. That is much harder than simply improving performance metrics. /end 2021-05-21 15:52:59 Another day, another privileged person arguing that we should trust AI with more facets of our lives. The thing is, trust in AI is just as much about trust in the people and institutions using that AI as it is trust in the models themselves. " 1/n https://t.co/kWjjuHueUI 2021-05-20 19:54:29 RT @Rasha_Abdul: THREAD EXCITING announcement @AmnestyTech is setting up an Algorithmic Accountability Lab & 2021-05-19 22:15:58 @DrSLDesmarais @_PolicyResearch Congrats, Sarah! 2021-05-19 18:21:48 @ruchowdh The transparency on process here is really refreshing. 2021-05-17 19:12:18 I am currently doing some stuff on a Windows machine and opened up the command line... and, friends, I was immediately transported back to my childhood, entering the magic set of words in MS DOS to open Treasure Mountain! Oh, the nostalgia! I need to go catch me some elves now. 2021-05-17 18:34:03 @picturewing 2021-05-17 18:18:32 There are only two times I get patriotic: (1) the Olympics (sorry, not sorry) and (2) when someone amazing takes a job in government. Congrats, Suresh! The country is lucky to have you. USA! USA! USA! https://t.co/HHpwMvvq4l 2021-05-17 14:24:37 RT @rajiinio: By the way, if you're looking to see the absolute other side of the spectrum - a feature on algorithmic bias with Black peopl… 2021-05-17 14:22:30 @csdoctorsister @60Minutes @jovialjoy @timnitGebru @rajiinio This is absolutely egregious. 2021-05-16 17:52:12 @cheeflo Yes! It is exactly this same impulse. 2021-05-16 13:21:46 Is it even possible to pick up a toddler without letting out an audible grunt and a comment about how big they’ve gotten? So far, the evidence points to no. 2021-05-10 15:49:04 Someday I'd really like to go back and re-learn basic math like this, as a semi-guided "voyage of discovery." I bet there is tons of beauty I missed out on the first time when I learned it by just memorizing the right algorithms to solve problems. https://t.co/6ffEYGcOHR 2021-05-10 15:08:32 @jordantroutboy @KnightHennessy @NSF I lived right near Stanford for three years, and esp if you have a car (maybe an expense that's out of budget), the best entertainment is just getting out into the outdoors. CA is so naturally beautiful that if you're into hiking/biking/beaching, you'll find plenty to do. 2021-05-10 14:49:55 I was hoping this guy would come to Penn, but so excited to follow and see all that he does at Stanford! Congrats! https://t.co/TJET6o8PsR 2021-05-06 01:16:09 @fhuszar To be fair, I don’t hate them quite as much for closets. They offer basically no sound insulation which isn’t as much of an issue with a closet like it is with a door to, say, a hallway. 2021-05-05 20:16:08 @841io @roydanroy Yeah I have some pocket doors too and I like them! 2021-05-05 19:48:37 @roydanroy “Barn door in bedroom” should do it. Here are some examples https://t.co/bddr61CYd9 2021-05-05 16:49:36 Can we please just be done with the inside-the-house barn door trend already? This is going to be the thing that totally dates houses to this time period, and I think we can all agree we'd just like to forget about the last couple years. 2021-05-03 16:21:40 @rcalo honestly, yes. But it's pretty friggin' cool. 2021-05-01 00:00:05 RT @hspter: @mikeolson @ruchowdh @KLdivergence The advice online for plants is truly inscrutable. It's somehow too much information and not… 2021-04-30 22:32:21 @jennwvaughan I actually think it might be the opposite? I think pets and kids can tell you when they're full but you can easily overdo it on watering plants. I don't know. All I know is I have a house full of plant carcasses, so I'm doing something wrong. 2021-04-30 21:51:50 I don't understand how I can manage to keep my dog alive for 9 years and my child for 2 years, but a couple of months with me and any plants are toast. 2021-04-23 20:14:45 RT @lizjosullivan: Hard to believe that this is my last day at @itsArthurAI, which makes this a tweet about... Some Personal News FAccT… 2021-04-22 14:53:08 RT @byjacobward: @KLdivergence @math_rachel As someone writing a general-interest book on AI who is in constant conversation with technical… 2021-04-22 14:09:08 RT @LeonYin: @geomblog @FAccTConference I came for the "ML" and stayed for the "FAccT." https://t.co/OyA5i6hzGY 2021-04-22 02:38:55 RT @jeremyjkun: @KLdivergence @math_rachel I'm even asked in design docs to write a section titled "Why shouldn't we do this?" I imagine no… 2021-04-22 02:38:53 RT @jeremyjkun: @KLdivergence @math_rachel Not technical, and yet Google's own engineering guidelines dictate that you should strive to fin… 2021-04-22 02:16:53 RT @KLdivergence: @math_rachel While I disagree with the perspective that a solution must be offered in order to point out a problem, it’s… 2021-04-22 01:41:26 @math_rachel I’ve seen this in my own work on predictive policing. Technical audiences are never satisfied with the fix being “just don’t do it.” Like certainly if we just throw moar reinforcement learning at the problem that will end racist policing 2021-04-22 01:35:59 @math_rachel While I disagree with the perspective that a solution must be offered in order to point out a problem, it’s worth noting that work in the AI ethics space often does point to a solution..... don’t build it. Or change policy. These often don’t count because they aren’t “technical”. 2021-04-22 00:50:51 RT @Linkletter: Proctorio is expanding their lawsuit against me and I urgently need your help. On the eve of the BC Supreme Court hearing… 2021-04-21 23:31:03 RT @LeonYin: I often think about how attending the FAT ML conference (now called @FAccTConference) at NYU in 2016 drastically changed my ca… 2021-04-21 13:45:35 @ARitaPena1 Thank you! Will check it out! 2021-04-21 13:30:09 Does anyone know where to get Portuguese language children’s books in the US? I want to buy some so my daughter’s au pair can read to her in Portuguese but all I am finding are poorly translated English books. https://t.co/PgUJrvRr12 won’t ship here apparently. 2021-04-21 01:27:22 @lauretig Unfortunately, no. My introduction was that I checked out a book on it from the Virginia tech library close to a decade ago but don’t remember the title/authors to track it down. May have been dempster and Shafer for all I know :) 2021-04-20 17:53:39 @alexhanna Realizing that this is kind of vintage mathematical notation makes this seem even better as a tattoo. I didn't realize that I was such a math hipster. 2021-04-20 16:28:26 Never have I been more thrilled to feel like I took a battering ram to the arm. A small price to pay to get back to normal! #CovidVaccine 2021-04-20 16:16:24 @atrishasarkar Yup. That's why I haven't made it much farther with it yet. My sense is also that it's more appealing theoretically than it is practically applicable. But, maybe not?? 2021-04-20 15:58:28 Oh, and anyone who responds to this automatically gets major nerd-cred in my book for being interested in Dempster-Shafer theory. It's such a weird, niche thing! 2021-04-20 15:53:52 Is anyone thinking about Dempster-Shafer theory in the context of automated decision systems? If not, I may add this to the long list of things that are probably cool and I'd like to work on but realistically probably won't. 2021-04-20 15:06:16 @ChadScherrer @AJWoStat I know, right!? And, perhaps more importantly, having that f or P there, really makes it hard to fit the components of Bayes Theorem into the limited finger real estate required for a knuckle tattoo. 2021-04-20 14:47:15 @roydanroy And, yes, apparently the 1990s is now vintage. 2021-04-20 14:46:01 @roydanroy This is just in reference to yesterday's thread about [A | B] meaning the "density of A given B". I got such a kick out of that, that I kind of want to both track down the origin of this notation and do this same exercise for other notations. https://t.co/4xmKhk6xyo 2021-04-20 14:05:51 @alexhanna so, you gonna get this tat or what? 2021-04-20 14:04:03 @PanakBill @AJWoStat But these actually have the P, it seems. I'm talking about notation that omits that P. 2021-04-20 14:02:34 Found my new thing! It's is niche, vintage mathematical notation. 2021-04-20 12:08:33 @AJWoStat @mattscho8 @tslumley @dan_p_simpson @mikebertolacci What the heck! I checked gelfand and smith yesterday! One of the first places I looked! Maybe I accidentally grabbed a different one?! 2021-04-19 23:53:38 @StatStaci5 @AJWoStat Awesome! Thank you! I wrote my dissertation on Bayesian spatial stats stuff though haven’t done much in that area since (it’s been about 10 years). No wonder it seemed common to me! 2021-04-19 23:40:21 @ruchowdh You’re welcome :) 2021-04-19 22:46:52 @tslumley @dan_p_simpson @mikebertolacci @AJWoStat YES!! I knew it was an applied mcmc thing! 2021-04-19 22:41:10 RT @tslumley: @dan_p_simpson @mikebertolacci @KLdivergence @AJWoStat It's in Gilks/Richardson/Spiegelhalter '101 Fun Things to do With MCMC… 2021-04-19 22:20:24 @StableMarkets @sam_power_825 @AJWoStat Yeah. I really like it too! I usually use it when writing stuff down on paper to save effort. 2021-04-19 21:51:30 @AJWoStat @dan_p_simpson This is my guess. Pretty sure I’ve got a copy at home. Will check when I get home. https://t.co/VST5UWCOmC 2021-04-19 21:49:54 @dan_p_simpson @AJWoStat This would make sense that I feel like it’s common but it seems like it might be kinda niche. I did do spatial stats for my dissertation. 2021-04-19 21:48:26 @sam_power_825 @StableMarkets @AJWoStat YESSSS!! We have our first official sighting!! 2021-04-19 21:33:51 @StableMarkets @AJWoStat Yes! I was thinking maybe some old Gibbs sampling papers from physics or something? 2021-04-19 21:23:42 @tslumley @AJWoStat @jonwakeblue @madeupinblue Same! I’ve used it but not in publications (unless you count posting drawings of bayed theorem knuckle tattoos publishing ). It’s driving me crazy not being able to find it anywhere. Thanks for the refs to others who might know! 2021-04-19 21:21:01 @AJWoStat @CortexNihilo I think this is probably zeroing in on the right corner of the literature though?? 2021-04-19 21:18:20 @matloff @AJWoStat I’m not asking for better notation. I’m asking if you have seen this particular notation used before and, if so, where. 2021-04-19 21:14:08 Paging @ioanna1182 @JCornebise you all must’ve seen this right?? 2021-04-19 21:01:01 Ok stats friends. I know this notation is a thing. @AJWoStat knows this notation is a thing. But we can’t find it anywhere!! Anybody know where [ | ] notation is used to denote conditional probability in general without specifying what the distribution actually is? https://t.co/doauoHeN2O 2021-04-19 20:31:18 @ruchowdh @alexhanna I’d be down for a Stein-estimator neck tattoo. 2021-04-19 17:11:49 @thejonullman Kinda looks like it, huh? The google streetview of the place had me pretty nervous, but it was actually super nice inside! But really, they could have vaccinated me in a back alley, and as long as it was legit, I wasn't about to complain. https://t.co/X9kuOIHnSQ 2021-04-19 16:30:23 @tslevi Hang in there! I hope your turn comes soon! 2021-04-19 16:21:39 @wraxtl Finally getting vaccinated FEELS like winning the lottery, so maybe so! 2021-04-19 16:18:48 One vaccine down! I. Am. So. Happy! #ByeByeCOVID https://t.co/M9jTiWXmBB 2021-04-17 12:54:57 RT @nadirahdixit: @jbouie @rmc031 because it *is* fun!! never ceases to amuse watching a certain type of Penn alum internally combust if th… 2021-04-16 22:40:42 I see you, people who faved this. And I have a specific sense of why this applies to each of you :) hang in there 2021-04-16 22:38:24 @SaranAhluwalia In the particular area I’m thinking of, at least two. Probably more. 2021-04-16 16:39:10 Me and Alix's dad have something in common-- we both like this piece! https://t.co/6GzQELLHZK 2021-04-16 15:48:38 RT @mdekstrand: If we build accurate models of the world as it is, but there are immoral biases between its present observable state and it… 2021-04-15 16:09:57 @SciOrestis You're on track to beat me! 2021-04-15 15:43:18 @DrVeronikaCH @geomblog This one's similar! 2018-2021, just one revision. 2021-04-15 13:57:38 @geomblog Ha! Journal publication turnaround time is bad but not *that* bad. 2021-04-15 13:53:39 I have a new PR for time from submission to publication at the same journal! https://t.co/epdZXXtfEy 2021-04-14 20:31:55 RT @mdekstrand: Why is critical race theory relevant to my work as a data scientist? Let's look at this picture 2021-04-14 19:54:24 @changhiskhan Ha! What! But actually I should factor that into my decision making process... 2021-04-14 19:41:05 @Ira_Sass Will try it out this summer 2021-04-14 19:11:10 April 14, 2021 at 3:08 PM ET is when I learned that it is not illegal to pick California poppies. https://t.co/CqzXute2dO 2021-04-14 19:06:26 @dmetaxak From what I saw, the heated blanket must've worked! Loved your talk! 2021-04-14 17:58:37 @geomblog @dmetaxak You too?? I already have a heater at my feet and sometimes it just isn't enough. Also will have to add it to my travel bag for in-person talks in the future :) 2021-04-14 17:56:20 @dmetaxak Wait, do you have this problem too? I am really liking this idea and wondering if when things get back to in-person I can bring a heated electric blanket to talks. 2021-04-14 17:52:17 I always give the WORST talks when I'm cold, but also bundling up in seasonally inappropriate clothing makes me seem like a weirdo. Decisions, decisions. 2021-04-14 17:45:33 RT @PHLPublicHealth: Get vaccinated today! Walk-up appointments are available today at the Center City Vaccination Center (FEMA clinic) un… 2021-04-13 18:52:51 RT @mdekstrand: Are you an undergrad interested in fair search & 2021-04-13 00:23:31 @andrewthesmart @geomblog Yep. 2021-04-12 22:07:59 Exhaustion from Apprehension about social isolation seeing humans again 2021-04-12 21:59:24 @roydanroy @alixtrot Maybe you and she would get along too! I've always said "honk" but, sure, they can say "bonk", I guess. 2021-04-12 21:54:11 @alixtrot My kiddo thinks geese say "bonk" and enjoys going to the park and yelling "BONK" at the geese. I feel like they might get along. 2021-04-12 21:39:30 @geomblog did you bug my zoom or something? This exact thing happened on a call I was on today! 2021-04-12 14:48:09 @jenniferdoleac @KrystaferRedden EXCUSE ME WHAT??? I just... can't imagine. 2021-04-10 16:57:22 RT @Aaron_Horowitz: Seriously, we and many others, modelled how this would play out back in March. Colleagues have been sueing all over the… 2021-04-10 16:04:55 RT @hrdag: Different Samples, Different Stories: HRDAG's @thegargiulian will give a keynote today at @StanfordODP Datathon! https://t.co/uw… 2021-04-08 20:54:31 @geomblog @nkrishnaswami I know :) I found that explanation very helpful but last time I checked it was gone from your website! 2021-04-08 20:50:35 @ziebrah @geomblog @nkrishnaswami excellent point! 2021-04-08 19:38:33 @nkrishnaswami @geomblog Perhaps you could make a parable involving an onion that illustrates this point. Then we could refer to the idea as Venkatasubramanian's Onion. 2021-04-08 17:59:50 @geomblog How about the Venkatasubramanian law? You’ll know someone’s trying to invoke if when they have that long pause right before trying to say it :) 2021-04-07 17:45:16 @amcasari Machado’s. Auburn, CA. 2021-04-07 14:40:18 Colleague of the year award goes to @Aaroth, who just clued me in to a clinic near me that had covid vaccine availability and.... I GOT AN APPOINTMENT! 2021-04-06 22:41:51 @kdpsinghlab @mt_head Also agree. It seems pretty accurate there. 2021-04-06 21:52:11 @robanhk I can honestly say I've never heard anyone say that. But, yeah, that seems plausible enough. I've definitely heard "take advantage of" there, which for me can have both negative and positive connotations (the positive being exactly the setting you described) 2021-04-06 21:44:35 @robanhk Genuinely curious: in what non-jargon context is "exploit" not a more negative/aggressive alternative to "use"? In my mind it's always pretty negative, but I could be overlooking some cases. 2021-04-06 21:42:07 @kdpsinghlab Yes, like "use"! This reminds me of this douchebag my mom dated when I was a teenager who constantly tried to throw around "big" words to sound smart (narrator: he was not). E.g. it was *always* "utilize", never "use". 2021-04-06 21:36:55 @latticepolys Yeah, not surprised. These are all pretty related. 2021-04-06 21:36:21 @latticepolys a very fair point! 2021-04-06 21:32:54 @latticepolys I don't know, I think it also has a pretty negative connotation in relation to natural resources as well. 2021-04-06 21:32:10 @njwfish Not surprised. If econometrics isn't the source of all that is evil, I don't know what is. 2021-04-06 21:24:15 The comfort stats/ML has with framing "exploiting" as a good thing is really something, huh? I'm thinking of explore/exploit but even more so the ubiquity of the phrasing "we exploit X" where X is often some property of data/algorithm/mathematical setup. 2021-04-06 18:01:53 @mmitchell_ai @alexhanna @veenadubal Wow, this is.... I don't know what this is. I am just shocked this would actually work. 2021-04-06 17:50:04 @Ira_Sass I know, right? It's like, I didn't send you that money so you could turn around and spend it on filling my mailbox with paper that I then need to sort through and recycle! Put it towards **the mission**, not spamming me! 2021-04-06 15:51:20 @evijitghosh Yeah, honestly, that's a problem. Reviewers can be obnoxious about that. Though, I do think just going with simulation is the right call in a case where you really just want to showcase a new method in general but don't want to get into the weeds about fairness in context. 2021-04-06 15:36:49 @evijitghosh Unless you want to actually learn about an application area, I think it's probably best to just simulate! No shame in that! 2021-04-06 02:11:54 Is there a way to donate to organizations on the condition that they never send me a physical piece of mail? Ever. 2021-04-03 22:49:04 @egrefen I scrolled by this tweet as I’m sitting in the bathroom with my 2.5 year old and she drinks soapy bath water. Honestly I just gave up. I just ask her if it tastes good and she says yes so 2021-04-02 18:37:06 @sherrirose @JMichaelMcW @timothyjlayton @HPDSLab @HarvardGSAS Oops! I jumped the gun :) 2021-04-02 18:32:19 @sherrirose @JMichaelMcW @timothyjlayton @HPDSLab @HarvardGSAS Get a sharpie and cross out the "future"! 2021-04-02 16:16:12 @shortstein But that would require that you actually read their papers, and who has time to do that?! But for real, I (perhaps selfishly?) also think it makes sense to consider their background. If they're coming from, e.g., statistics, you probably want to look at journal pubs as well. 2021-04-01 22:19:05 RT @jw_lockhart: @FannisJengler @KLdivergence I certainly hope that people learn from the convo! But it is not surprising to me that people… 2021-04-01 20:25:09 @timnitGebru @databoydg Anyway, agree it is best to move on. I'd already muted OP long ago, so no action items for me here! 2021-04-01 20:24:04 @timnitGebru @databoydg Yep. Accurate. I'm a statistician, and this was my reaction https://t.co/O17mRH16JM 2021-04-01 00:55:32 @mdekstrand That bad, huh? 2021-03-31 23:11:55 RT @Ira_Sass: Happy trans day of visibility everyone! Aon some easy things to do to be trans-inclusive. 1. Educate yourself about pronouns… 2021-03-31 20:46:20 @osazuwa Disney plus has tons of stuff in both Portuguese Portuguese and Brazilian Portuguese. 2021-03-31 20:15:24 @ghoshd @David_NeverDave @AllbriteAllday Maybe? Even more songs to sing to that tune! 2021-03-31 17:20:04 Currently reading @AllbriteAllday's paper "If You Give a Judge a Risk Score", but can't not read the title to the tune of "Do You Want to Build a Snowman?" 2021-03-31 15:21:35 < 2021-03-31 15:16:37 @acidflask @AlexGDimakis hell yeah I am! 2021-03-31 14:39:41 RT @Meagan_Levinson: Such an excellent piece from @karen_ec_levy. https://t.co/ZCIjWJG0Dx 2021-03-31 13:26:29 RT @AlexGDimakis: Thanks to CLIP zero-shot classification this can be *objectively* settled: Our AI overlords call them Manta Ray plots. ht… 2021-03-31 13:26:18 @AlexGDimakis Hold on, so, wait. I just googled "manta ray plots" and pretty much the only thing that comes up is my tweet. Did it train on that? Or, does CLIP just "think" like me? I don't know how I feel about that. 2021-03-31 00:05:32 Well well well, look who it is! @wsisaac making an appearance in @signmagazine https://t.co/n0iSCnRrpX 2021-03-29 17:50:17 @mmitchell_ai @haydenfield I'd like to hear the story of a butchered TED talk :) 2021-03-28 14:22:58 @geomblog Lol please don’t! After oh 15 years of being a statistician I still can never remember which number is which. Learn from our mistakes! 2021-03-28 01:00:18 logged in to Twitter to see that we that we are doing the whole “algorithmic bias is just a data problem and we can fix it” thing again. Getting my tweeting thumbs ready, but actually... Nope. Not today, satan. Logging out for the rest of the weekend. 2021-03-28 00:02:53 @rajiinio I thought we were done with this but I guess not. Thank you for your service, deb. 2021-03-27 21:41:49 @josephtlucas I haven’t but I’ll check it out! 2021-03-27 19:13:52 @GM_TweeterBird Thank you for settling it! Basically it’s a mut :) I didn’t realize they could mix like this and the blue bill was throwing me! Thank you! 2021-03-27 17:22:55 @mom2phd The face doesn’t have the white stripe like the ruddy duck. Maybe a ruddy duck/ mallard cross though? 2021-03-27 17:22:03 @mikeloukides I’m also thinking mallard hybrid.will look into mallard/merganser cross 2021-03-27 16:47:54 @margootjeg I’m thinking it’s a mallard-scaup hybrid 2021-03-27 16:45:41 @margootjeg They were also about I’d guess 1.4x the size of the regular mallards and swam with the front submerged a bit more as though leaning forward whereas the mallards were more flat 2021-03-27 16:39:04 @margootjeg This is the best I got coloring-wise. Nothing super obvious though I do think some of them were kind of light reddish-brown in the front. One thing that maybe also points to hybrid is that there was a lot of variety in this group in terms of looks. https://t.co/m8Uq6FcONw 2021-03-27 16:30:44 @margootjeg THIS seems reasonable to me! Definitely not a mallard but mallard-like! Any guesses as to the cross? Maybe ruddy duck? 2021-03-27 16:13:20 @v__s_ @marthaharbison @mollycule__ I’m specifically talking about this guy with the blue beak. There were mallards there too but these guys were bigger and obviously different https://t.co/rzjqzGBQyi 2021-03-27 16:06:10 @alex_riina Here’s a zoom in of the ones I mean. Never seen a mallard with a blue beak. https://t.co/Ao1UYbhnFI 2021-03-27 15:59:56 @davidrpoliti And yeah definitely some Canada geese back there too. I’m talking about the birds in the foreground. 2021-03-27 15:59:28 @davidrpoliti Hmm possibly? The tail wasn’t very pin-like but that was one of the possibilities we considered 2021-03-27 15:58:48 @alex_riina I’m sure they weren’t mallards though. There may be a few mallards mixed in but these dudes were obviously different. 2021-03-27 15:57:41 @alex_riina Yep! Philly! Down in the wissahickon 2021-03-27 15:46:17 Should have hash tagged before. #birdersoftwitter #birds #birdwatching #ducks #BirdTwitter 2021-03-27 15:42:48 @mdekstrand Only ironically and improperly to make fun of my oldness with my 20-something friends 2021-03-27 15:42:12 Help an amateur birder out. Cant find these ducks in my book! https://t.co/2LGRY1zxAU 2021-03-26 20:25:27 @jbakcoleman @dakkaynak It is DEFINITELY 2 for me. Always 2021-03-26 18:53:12 @rdpeng It didn’t work for the Ever Given and a screaming, determined toddler is at least as strong as a container ship. 2021-03-25 17:05:00 @balazskegl Have you heard of the field of "statistics"? It's kind of niche, but you might find your people over there. 2021-03-25 16:18:24 @dan_p_simpson @mdekstrand @jazzystats I've already sent it around to some collaborators, so, yes! definitely useful! 2021-03-25 15:54:24 @BlancheMinerva @IgorBrigadir Yeah, I don't think what we decide on this is going to change much in terms of the analysis due to the small sample. In this case, it's more about treating data that represents humans with the care it deserves. But, also, I want to learn for future cases where there's more! 2021-03-25 15:36:29 @BlancheMinerva @EconAndrew This is data from one of those open data portals. I haven't had much luck in the past getting additional info out of these types of things on, e.g., details of collection process, but it's definitely something we're going to try (about some other variables a well). 2021-03-25 15:32:42 @btskinner @jazzystats That looks great. Thank you! 2021-03-25 15:31:56 @BlancheMinerva @sakrejda Yes. 2021-03-25 15:21:10 @sakrejda @BlancheMinerva Indeed. 2021-03-25 15:20:48 @geomblog @scheidegger @kdphd Excellent point, Suresh. I wonder if I could find any pie factories to collaborate with to invent some new statistics for the the big problems of the day in pie making. 2021-03-25 15:09:40 @IgorBrigadir I responded up-thread with a problem description! Definitely interested if you think that this idea might work with that sparse of data in this use case. 2021-03-25 15:08:38 @IgorBrigadir There are like 5 people in that group. I don't think resampling will get us too far with that few. 2021-03-25 15:06:33 @BlancheMinerva I was thinking female and non-female because my understanding is that gender diverse people tend to be treated with suspicion by police in a way that is more similar to men than women. But... that's as far as I got... 2021-03-25 15:05:22 @BlancheMinerva Thank you so much! So, the use case has to do with modeling police stops and likelihood of search following stop (and effects over time and space). There are a few people whose gender is listed as non-binary, though truly not enough to get reasonable estimates for that group. 2021-03-25 14:59:01 Hi Twitterverse, can anyone point me to some references on best practices for building models with data where there are **very** few non-binary people represented? Problem is, very sparse data for those folks precluding reasonable model for them, but also don't want to drop them! 2021-03-25 14:48:33 @scheidegger @geomblog @kdphd I meant wasting my time in the sense of doing statistics is not getting me closer to my ultimate goal of pie taste-testing. 2021-03-25 01:05:00 At this point, I'd welcome some shrimp in my breakfast cereal. It would really break up the monotony of just sitting around at home all the time not having shrimp in my breakfast cereal. 2021-03-24 23:46:47 @kdphd @scheidegger @geomblog Haha! Yes! definitely fishing for pie :) 2021-03-24 23:37:54 @scheidegger @geomblog @kdphd I've wasted my time doing statistics for so long, I'm probably not qualified even if I could figure out how to apply. 2021-03-24 23:30:58 @tylerlu Sometimes. And sometimes proving a theorem or two in the paper you want to write would require abstracting away or distracting from the interesting bits you specifically want to write about. 2021-03-24 22:50:09 @geomblog @scheidegger @ccanonne_ I think by news influencers they mean "friends". Where on earth did you find this? 2021-03-24 22:48:49 @scheidegger @geomblog @kdphd Unfortunately, I don't know where to apply for the job that was my childhood aspiration: pie taste-tester. 2021-03-24 22:38:33 Just this. https://t.co/UJQjZtxuk7 2021-03-24 22:29:21 @geomblog @kdphd You're right. Nothing matters. Why are we doing any of this again? 2021-03-24 20:50:12 @geomblog @kdphd Well, at least in the end, you got to say what you wanted to say? 2021-03-24 19:50:01 @NinosIndex LOL, granted, maybe that was not one of those times :) 2021-03-24 19:11:36 Sometimes you've just gotta write the paper you want to write rather than the paper that's going to have an easier time getting in. 2021-03-23 15:36:34 @kdphd Congratulations on the excellent work. And, maybe even more... congrats on the perseverance! 2021-03-23 14:44:29 RT @kdphd: It has been 7 YEARS since @geomblog @scheidegger and I started writing this paper, 5 since we put it on arxiv, and it's FINALLY… 2021-03-22 19:20:10 @brennan_mike That is so cool! 2021-03-22 14:14:11 @Alexis_AnYee Thank you! 2021-03-22 13:49:54 RT @mdekstrand: at some point my students will probably get tired of me showing them this slide. but it continues to be relevant. https://t… 2021-03-22 13:29:50 @geomblog @Aaron_Horowitz @GermsAndNumbers I’m not even in the queue :( 2021-03-21 23:28:46 @GermsAndNumbers I feel the same! I think you might be one of my only friends who I know does NOT have it based on this tweet 2021-03-21 20:53:13 @yisongyue Lol haha yes oops 2021-03-21 20:17:12 Not seeing nearly enough complaints about iclr reviews. 2021-03-20 15:52:48 @malleshpai @Aaroth @geomblog @Aaron_Horowitz But then I’d have to be an economist 2021-03-20 01:56:16 @tswaterman 2021-03-20 01:41:14 @tswaterman I considered it. If I had an account ready to go I would have. 2021-03-20 01:24:38 @forestgregg Thank you so much, forest. 2021-03-20 01:15:24 @Aaroth @geomblog @Aaron_Horowitz Great! Cant wait for the complaining to start. I’m happy to extend my current complaining to participate in this round! 2021-03-20 01:03:14 RT @steindavidb: @KLdivergence 50 years of using the supremicist rhetoric of the Axis powers is probably enough. It's pretty easy to stop u… 2021-03-20 00:45:07 Excitedly awaiting 2/2.... https://t.co/a4rxcY3SXm 2021-03-20 00:05:05 @miguelusque There will be no follow-back though. That’s for sure 2021-03-20 00:00:16 @hadleywickham So far, no luck. We can’t all be as bold as you, Hadley. 2021-03-19 23:39:15 @joel_c_miller I don’t have patients but apparently the authors of that paper have patients with patients 2021-03-19 23:23:27 @geomblog @andrewthesmart @scheidegger Whoaaaaaaa what was it like @scheidegger ? 2021-03-19 23:05:44 @andrewthesmart I’m conflicted. I’m both SO ready to go back to normal life and also super apprehensive about having to interact with real humans in person 2021-03-19 23:04:01 @meghapsimatrix Will be on the lookout... 2021-03-19 23:02:48 @miguelusque Yeah I’m not mad about it. Anyone is welcome to follow. Just semi awkward. 2021-03-19 22:56:21 Sitting here hoping for him to like this tweet just so this can be as awkward as humanly possible.... 2021-03-19 22:32:19 LOL that feeling when the editor who just rejected your paper follows you on Twitter while you're complaining about it. Hi there 2021-03-18 02:56:31 @vm_wylbur So awesome! 2021-03-17 23:14:09 @geomblog @hannahsassaman Ok so her favorite type of whale is “big whale” and she had no comment about the car inside of a whale. She looked confused and then handed me a book. 2021-03-17 19:56:50 @geomblog Um isn't the point of this that what the garden needs varies from day to day?? 2021-03-17 14:11:59 @afeight27 Yes but she only kinda likes them. 2021-03-17 13:58:50 RT @theduckiest: I think this is the most important story I’ve ever worked on. 2021-03-17 03:17:42 @DrLukeOR She’s 2 :) 2021-03-17 02:34:27 @hannahsassaman @geomblog My guess is that belugas are the only kind she’s familiar with but I’ll ask. She’s constantly saying things that I have no idea where she possibly learned it, so for all I know she’s gonna be like “duh. Sperm whales.” 2021-03-17 02:32:09 @geomblog @hannahsassaman Will report back tomorrow. 2021-03-17 02:27:04 @hannahsassaman I know right!! Whales are awesome! And it is hilarious to hear a tiny voice out of nowhere say totally matter of fact “I kinda like whales” 2021-03-17 02:12:01 At various point during the day today, my daughter announced unprompted that she “kinda likes” -cars -whales -juice -peacocks And being the total mom that I am, I need to share this with you so I can bask in the delusion that you, dear follower, find this as cute as I do. 2021-03-16 13:29:11 RT @AngeBassa: “Data are just a bunch of qualitative conclusions arranged in a countable way.” https://t.co/bOUBec17GX 2021-03-16 00:32:50 @JubaZiani @JamesJohndrow Thanks so much for the invitation! I look forward to talking to you all about our new work! 2021-03-16 00:29:05 RT @JubaZiani: This week, our speakers for the toc4fairness seminar series are @JamesJohndrow and @KLdivergence! More info about the talk/s… 2021-03-15 23:46:18 @DMulliganUCB This is some mom of the year shit right here 2021-03-15 15:19:10 @JubaZiani Celebratory drink time?? Congrats! 2021-03-13 22:29:59 A fun little poll for your Saturday afternoon. (My money is on logistic regression going all the way) https://t.co/z5nJMFb08y 2021-03-13 19:51:36 @kdpsinghlab I have no idea what you’re talking about https://t.co/1dz80I0Eow 2021-03-12 18:43:22 @timnitGebru I'm so sorry, Timnit. This is absolutely disgusting. 2021-03-11 19:00:22 @extremefriday thank you! < 2021-03-11 18:53:39 Third panel moderated of the week! I didn't realize it, but I kind of love doing it.... though I strangely find it to be at least as much pressure as actually being a panelist. https://t.co/MEvbaLMYkR 2021-03-11 14:01:41 RT @Phil_Lewis_: Happy 4 year anniversary to the best interview of all time https://t.co/ugUdcHzjpV 2021-03-10 22:16:27 Whew! That double header moderating the #facct2021 town hall followed by paper session 26 with @emilymbender, Tony Sun, and Molly Roberts has me absolutely wiped. NAP TIME! 2021-03-10 22:00:27 Thank you to @Abebab and @rajiinio for their work running the #facct21 socials! https://t.co/Xd9U4u867D 2021-03-10 20:36:47 @emilymbender Put those questions into slide-o, people! I'd love to ask the authors in the Q& 2021-03-10 17:11:21 of* 2021-03-10 16:04:27 It's the last social hour off #facct2021. If you'd like to come say hi, go grab the link from Circles here: https://t.co/6gPOzx4Qc7 We've had some great discussions on past days. Hope some of you all can swing by! 2021-03-10 14:50:47 Last day of #facct21! I'm gonna need a few days to rest and synthesize after this firehose of great research. Anyone else? 2021-03-10 02:29:37 @luke_stark @drjessehoey This was a great presentation. I really enjoyed it. 2021-03-10 01:29:48 @CalebMiles16 Why do you think I talk so damn fast? 2021-03-09 18:59:51 RT @mdekstrand: following @KLdivergence's example, I'm going to open up my zoom room for the noon social break. Look in Hallway Chat for th… 2021-03-09 17:34:05 BRB, going to go read every damn thing Issa Kohler-Hausman has ever written. #facct2021 2021-03-09 15:40:17 @KenButler12 Haha, I actually do do this... I like doing things like knitting, puzzles, or cleaning up while listening to podcasts super fast. 2021-03-09 14:51:19 @TomTheDataDude I'm currently listening to one at 1.5x but listened to another one at 2x. Depends on the speaker for me. 2021-03-09 14:49:08 I kind of love the pre-recorded online conference format because I can watch the talks on 2x speed. 2021-03-09 13:46:24 RT @being_saige: @FAccTConference is changing my perspective forever. I am learning more than I have ever learned at any academic event. #f… 2021-03-09 13:45:51 Another dream job alert! https://t.co/taCoisWBal 2021-03-09 02:47:37 @ziebrah You’re the real deal, best paper award winner! 2021-03-09 02:14:27 @sameer_ @yisongyue @Caltech Same. I’m kind of afraid to ask. 2021-03-09 00:02:03 @SheridanLGrant I think that's a lovely idea 2021-03-08 23:37:40 @StatStas @ruchowdh Harry Potter and the Cursed ML terms 2021-03-08 22:21:54 @ruchowdh Was thinking the same thing :) 2021-03-08 22:21:01 @geomblog @ccanonne_ @joaomlribeiro @NoahSD I was wondering the same thing :) 2021-03-08 22:09:45 @rajiinio Oh, for sure. Honestly, it was *already* embarrassing last time this happened with Cynthia Rudin's face regenerator thing. At this point, it's straight up negligence. 2021-03-08 22:01:47 @rajiinio At this point, it's hard to call it an "oops", isn't it? There's no excuse for not noticing this at this point. 2021-03-08 21:50:09 @cheraghchi I never said "unbiased coin" is exclusive to CS. Taking that Webster definition and applying it in this case is quite a stretch. 2021-03-08 21:48:35 @NeuroStats @tdietterich I've only just recently been noticing it, but it's everywhere! 2021-03-08 21:40:59 @cheraghchi Not sure what dictionary you're using, but that's not what I'm seeing. Perhaps you're referring to a ML text, which is circular. I'm saying it doesn't make sense for the field to call it that and it sounds like you're saying it does make sense because that's what they call it. https://t.co/tRDVQV70n3 2021-03-08 21:37:51 @ThankUBayesGod Mine too. This is a self-critique. Wish I'd realized sooner. 2021-03-08 21:34:30 @cheraghchi which still implies a 50/50 split is unbiased. Why is a 50/50 split the "unbiased" state? 2021-03-08 21:25:30 Sure, they signed up for their photos to appear in an article that contextualizes their stories.... but did they sign up to be the poster boys of every random fair ML method that random people they've never spoken to come up with? 2021-03-08 21:22:12 Anyone ever wonder whether the two guys whose pictures appear at the top of the ProPublica Machine Bias article know that their pictures appear in pretty much every "fair ML" presentation ever and how they feel about that? https://t.co/szN1XuWpwP 2021-03-08 21:18:46 @cheraghchi Why should they define it as -1/+1? 2021-03-08 21:06:03 @Ira_Sass @ccanonne_ Ah, ok, can't argue with wikipedia! 2021-03-08 18:02:17 On deck to speak at #FAccT2021 is @YESHICAN. I've heard her speak before, and she is fantastic. Tune in now! 2021-03-08 17:27:26 RT @struthious: Shameless self-promotion: the paper we (@ziebrah and I) co-authored won one of the FAccT 2021 best paper awards. Come ask u… 2021-03-08 14:39:01 Day 1 of the FAccT academic program! 2021-03-08 14:27:22 @hspter @data_nurse eagerly awaiting the answer! 2021-03-08 01:20:07 @hlynur @hspter Thank you 2021-03-08 00:58:55 @hspter Strange, possibly very stupid idea.... but burn it? In my experience burning doll hair smells very different than burning human hair. 2021-03-07 20:50:28 RT @AlvinGrissomII: It's been a while since I've had the pleasure of the awkward silences as a bunch of socially awkward nerds standing in… 2021-03-07 01:57:10 @BM789 Uh oh. What do you end up hauling around for kids that age? 2021-03-07 01:56:01 @aaronkbr Sounds luxurious! I’m almost there! 2021-03-06 21:30:15 As soon as my daughter is out of diapers I am buying myself the stupidest, most pointlessly tiny purse just because I can. 2021-03-06 21:28:10 @AmeliaMN @hspter Same! 2021-03-06 02:27:19 @acidflask @dataandpolitics Same. I had no idea. And I have a Chinese last name! 2021-03-05 19:12:21 RT @m_c_elish: If you weren’t able to join #FAccT2021 CRAFT sessions, check out all the incredible TUTORIALS now posted to the conference h… 2021-03-05 17:45:51 @geomblog You're talking about (im)possibility, aren't you? That paper is so good. 2021-03-05 14:24:36 RT @BrownSarahM: Another secret joy of @FAccTConference : the volunteers!!! @SciOrestis recruited, prepared and scheduled so many! They've… 2021-03-05 00:44:46 @LucyStats Almost. I think *on expectation* it would fall inside of 95 of the intervals. But I’ll take it! 2021-03-05 00:42:46 RT @LucyStats: Proper confidence interval definition in the NYTimes, love to see it! “If scientists came up with confidence intervals for… 2021-03-04 23:24:53 @dimvar @dmetaxak Whew! Glad to know you’re on board. 2021-03-04 22:41:52 @scheidegger @841io I'm having a *much* harder time with Spanish than I thought I would... 2021-03-04 22:37:44 @ccanonne_ @geomblog @scheidegger @841io Sorry, we already did French :) 2021-03-04 22:36:34 @geomblog @scheidegger @841io YES. Speaking of, I have some in my fridge that I need to cook. Mmmmmm 2021-03-04 22:33:50 @scheidegger @841io And, yes, I still don't remember how to do all the diacritical marks on my keyboard so you'll have to continue to deal with my poor written portuguese. 2021-03-04 22:33:12 @scheidegger @841io I mean, dude, I had to spell cheese... I had to take some artistic license here. 2021-03-04 22:27:50 @841io OK, while i'm at it... QUEIJO-- QUestoes de Etica, Igualdade, Justica, e Opacidade @scheidegger thoughts? 2021-03-04 21:53:31 @hspter Nope. I just really like cheese. 2021-03-04 21:46:15 @841io I can make cheese acronyms all day, folks. Step on up if you need any more! 2021-03-04 21:44:55 @841io BRIE- Bien social ,Responsabilité, Interprétabilité, Éthique 2021-03-04 20:57:07 This post brought to you by the letters F, T, E, and A and my lunch. https://t.co/bqxaRFjbG6 2021-03-04 20:46:59 @kdpsinghlab I already knew you were the absolute king of stats dad jokes... but maybe that's just dad jokes in general? :) 2021-03-04 20:44:33 @HetanShah I don't know if this is apocryphal but when I was growing up, I heard it was once discussed creating a public transit system similar to the Bay Area's BART for my area.... Foothill Area Rapid Transit. 2021-03-04 18:07:16 @bhecht What an absolute joy to get to work with you and, really, everyone in organizing this conference. 2021-03-04 17:55:59 RT @geomblog: What @BrownSarahM is not saying (and should) is that she's been responsible for much of the very nice social environment that… 2021-03-04 17:48:05 RT @bhecht: @KLdivergence being wonderfully eloquent as usual! I’ve also hit FAccT leadership “retirement age” and I feel exactly the same… 2021-03-04 15:26:25 So, as it turns out, yoga with a 25 pound wiggly weight jumping on your back (i.e. a toddler) is an extremely effective core workout. 2021-03-04 15:20:31 @mdekstrand @FAccTConference @rajiinio @emilymbender Should be a-double-c, right? 2021-03-04 14:17:51 @rajiinio Our discussions last year were one of the highlights of the conference for me! These types of hallway (or I think it was more like doorway) conversations are one of the things I hope Circles can facilitate, but it's gonna be tough this year. 2021-03-04 02:06:53 RT @jeanluc_t: I often regret being interdisciplinary. The physicists tell me my paper would be more appropriate in a math journal, and vi… 2021-03-04 01:12:59 Excited for #facct2021 but it’s strangely bittersweet. It’s my third year on the executive committee so it’s my “retirement from FAccT leadership” year. I’m ready to pass the baton, beyond proud of how the community has grown, but sad to not see you all in person this last year 2021-03-03 19:07:06 @evijitghosh @FAccTConference @shakir_za Wonderful to meet you, Avijit! 2021-03-03 17:54:36 Ahhhh greater than OR EQUAL TO zero. I wish I wore a hat so I could throw it angrily on the ground right now. 2021-03-03 17:35:16 @_aaronmiles Honestly, normally I enjoy it. What a letdown this time though. I found some neat identities I was sure were going to give a non-trivial bound and, then, when everything fell out... yeah... not so much. 2021-03-03 17:30:19 Math: Not even once. 2021-03-03 17:23:09 Just spent the last couple hours proving the probability of < 2021-03-02 23:16:37 @mayhewsw @terrible_coder Sorry should have known 2021-03-02 22:30:09 @terrible_coder @mayhewsw Oops, was it a joke? Wouldn’t be the first time I’ve missed the mark on Twitter. :) 2021-03-02 22:19:14 @mayhewsw I think you can do this in vim without commenting every line. https://t.co/oPURqnFf18 2021-03-02 22:14:05 @hspter Brilliant! I think I have one left! 2021-03-02 14:10:12 @PierreEJacob @RoyalStatSoc Congratulations Pierre! 2021-03-02 01:59:07 RT @BrownSarahM: If you're also excited about @FAccTConference, remember the conference site is already open if you're registered! You can… 2021-02-28 15:00:20 @mdekstrand This has to be a joke. This is an abomination. 2021-02-27 22:51:53 @PhDemetri I nearly always get the old “oh I took stats in college. Hated it. I was so bad at it. “ 2021-02-27 01:10:52 RT @hannahsassaman: This is absolutely core and something that organizers have focused on for a long time - it isn’t what the tools say, it… 2021-02-26 18:50:32 "It’s not just about what these tools say, not just about the bias in them and how they communicate that bias, but what they do. We have to audit the use of these tools for the impact they have in the systems." -@hannahsassaman https://t.co/emlaTIBRAb 2021-02-26 18:27:08 RT @PrisonPolicy: Since July, 66% of the jails we've been tracking have had population increases, suggesting that the early reforms institu… 2021-02-26 17:02:40 RT @AlexCEngler: These two paragraphs by @ruchowdh & It strik… 2021-02-26 17:01:48 @AlexCEngler You nailed it. This was the precipitating twitter thread :) 2021-02-26 16:42:54 @ioanna1182 @betanalpha @five9a2 @learnfromerror Thanks! Nice work to @shiraamitchell @alexdamour Eric Potash and @s010n as well! 2021-02-26 15:29:08 @mchardcastle @RDBinns off key *human* singers of all genders? I am so confused. That seems a bit.... extreme. 2021-02-26 15:12:01 My op-ed with @ruchowdh is out in @techreview! We discuss how differing views on what "counts" as an algorithm muddies the water. When it comes to policy, focus on impact. https://t.co/EdFUnQjvzZ 2021-02-26 15:03:34 Help her out! And also follow her. She's hilarious in person so maybe we could get some more @ioanna1182 content if she had more followers? https://t.co/BWm0KPzt2c 2021-02-26 14:55:04 @mchardcastle @RDBinns Ok wait. So the intent was that they could eat human singers who were off key or birds?? 2021-02-26 14:53:16 @John_J_Howard @ruchowdh Thank you! In my mind, yes, all are algorithms! Though the term seems to be causing some confusion with the broad class of things of varying complexity it covers 2021-02-26 14:45:59 RT @ndiakopoulos: "What matters is the potential for harm, regardless of whether we're discussing an algebraic formula or a deep neural net… 2021-02-26 14:05:04 RT @ruchowdh: In our new @techreview piece, @KLdivergence and I dive into unclear definitions of the term “algorithm” drives confusion in c… 2021-02-25 20:37:10 @GermsAndNumbers @piliffq Can the journal founder publish his own papers in the journal? 2021-02-25 17:47:42 RT @lorin_crawford: I will say this over and over again: I would not be here today without Theresa Price, @BlkCollegeExpo, and my HBCU @CAU… 2021-02-25 17:47:10 @lorin_crawford @BlkCollegeExpo @CAU @CBSThisMorning @jamieyuccas You're famous! I mean, more so than you already were :) 2021-02-25 15:12:59 Does anyone else have trouble getting the comments in @overleaf to go away? I open them, and then click seemingly randomly around the page until they go back into hiding. 2021-02-25 13:03:30 RT @deaneckles: Nice example of using domain knowledge & 2021-02-25 12:57:19 RT @BHedtGauthier: An important discussion about the history and future of statistics, featuring @ScarlettBellamy @goodmanthebrain @TukufuZ… 2021-02-25 12:38:01 @Corey_Yanofsky It’s a giant friggin sword :) 2021-02-24 19:59:08 @andrewthesmart I'm about to give a talk in 2 minutes, and I might just put this advice to use. 2021-02-24 19:46:17 Post an image of yourself as the final boss without downloading any new pics. https://t.co/mW9ob8bp9I https://t.co/Bmfpfcx6m5 2021-02-24 14:53:49 @Binette228 @joftius @ajwroblewski78 I agree with this. Bad statistics is bad statistics. 2021-02-23 20:19:37 @kjw_chiu @alex_peys @AnthonyLeeZhang If your interest is stats you could also just do a PhD in stats. 2021-02-23 18:32:11 @2plus2make5 I'm not even going to vote because this is just so wrong. Green apple or go home. 2021-02-23 15:08:21 Just read everything @uhlily writes. Seriously. https://t.co/tkbdhNPzIn 2021-02-23 04:06:13 RT @EvanRosenman: Happy to finally put a preprint of my 3rd dissertation paper, "Designing Experiments Informed by Observational Studies,"… 2021-02-22 17:48:27 @ruchowdh @humanetechne @TwitterEng it is?! 2021-02-22 16:59:37 RT @WiDS_Worldwide: "All along the way, I’ve gone where my interests have taken me...What is the problem I’m really interested in working o… 2021-02-20 22:57:49 @mdekstrand Ah I see. Yeah mine is in latex :( What you’ve done seems much more advanced than what I’m doing 2021-02-20 21:33:20 @mdekstrand This looks fantastic. Teach me! Mine is in need of an update! Badly... 2021-02-19 22:26:52 @mmitchell_ai Omg. I am so sorry! 2021-02-19 18:45:06 RT @KevinKaichuang: What if I flip a coin 1e6 times, throw out half the heads, train an unbiased estimator on the remaining data, and then… 2021-02-19 14:51:40 @Sci_Tai Yeah, I saw that. I'm glad they're making progress there, though selfishly, it doesn't help me much with my 2 year old. 2021-02-19 02:02:17 @JennK3s @Tamara_DeBeuf @corinederuiter Ummm that was “I look forward....” thanks autocorrect 2021-02-19 02:01:52 I don’t know how I missed this info but the covid vaccine isn’t for people under 16.... and they aren’t even testing it for littles yet!!! and here I was feeling optimistic we’d all be vaccinated by summer 2021-02-19 01:54:49 @JennK3s @Tamara_DeBeuf @corinederuiter Thank you for taking the time to read! iBook forward to your comments! 2021-02-17 20:45:22 RT @guacamolebio: holy fuck data cleaning is like 90% of the work you can’t split the work up by saying which graphs someone should make ju… 2021-02-17 15:06:13 @thienan496 @pmddomingos I just spent several minutes trying and ..... I cannot. 2021-02-17 14:35:53 @MeganTStevenson @davidsabrams This looks awesome! I can’t wait to dig in 2021-02-17 14:31:34 RT @MeganTStevenson: How "risky" does someone need to be to justify pretrial detention under current law? @sandy_mayson and I lay out the l… 2021-02-17 13:54:04 @sherrirose This is part of the reason it galls me too. 2021-02-16 19:51:09 RT @_5u11y: @KLdivergence Infosec has had rockstars and it's been pretty harmful. Some of us have been promoting plumbers in the field as t… 2021-02-15 23:29:30 @JubaZiani This looks great! 2021-02-15 19:42:19 @jbakcoleman Hold on. Wtf does this even mean? Of course plant and animal based food contains genetic material.... I’m so confused. Should we just be eating rocks? 2021-02-15 18:17:54 @acidflask What is the policy on remote? Any options for people who, say, live in Philadelphia and could come to the office one or two days a week or so? 2021-02-14 20:15:26 @deaneckles Not to be that jerk but the Krustal-Wallis test 2021-02-13 18:16:46 @Aaron_Horowitz I think it’s pretty easy to spot the difference :) it’s one of those you know it when you see if type of things 2021-02-13 02:09:14 @Aaron_Horowitz @GermsAndNumbers I second this question.... 2021-02-12 21:39:36 @andrewthesmart Not at all! 2021-02-12 21:36:54 @RobotsSonia Honestly, depends on the situation. Yes, I do interrupt when I feel that I can, but there are plenty of situations where this happens and I feel like I can't, for whatever reason. 2021-02-12 19:06:02 @BrownSarahM Maybe this is why I tend to end meetings early. :) Don't want to do this to anyone else! 2021-02-12 19:01:16 @timnitGebru Seems like the person with a fixation is the person tracking down everyone connected to you on Twitter. 2021-02-12 18:51:03 So, I apparently subconsciously interpret excessive droning on in meetings as an act of aggression. Like, a display of power. Nothing gets my fight or flight response revving like someone talking at me for far longer than is useful. 2021-02-12 01:12:16 @zschutzy Oh I’m so jealous! Congrats! 2021-02-12 00:47:12 @zschutzy Did you get a vaccine?! 2021-02-11 17:41:08 @jhaushofer OK, economists are officially the worst (I already knew that anyway, but this seals it). But, computer scientists are pretty bad too... See e.g. https://t.co/yTGNXjIjSd, where you can expand any institution and see yourself ranked next to your colleagues! 2021-02-11 15:20:32 Having this issue where the word "height" keeps randomly appearing in tables (see link below for other example, not my own). LaTeX, you are a beautiful mystery. https://t.co/JkPhnLYkm6 2021-02-11 14:06:12 @jhaushofer Link?? 2021-02-11 13:11:37 @jhaushofer Please no. There are charts?? 2021-02-11 04:55:00 @CaitDoogan They would. 2021-02-10 18:47:41 @_5u11y I like this a lot. A whole lot. 2021-02-10 18:42:59 @grimalkina Good to know. 2021-02-10 17:56:58 @Stephen_Lynch Also wondering this! Is rock still the cool thing? Only time I listen to music is driving around in my mom-car listening to music from the early 2000s now, so I really don't know what's cool these days. 2021-02-10 17:50:11 @stuartbuck1 So considered. Any rockstar properties I attribute to her are unrelated to her ability to do statistics. 2021-02-10 17:42:19 @SeeTedTalk Yes, I've heard it here too a lot. But, I pick on CS a lot these days, so I thought I'd go back to my roots and pick on my home field for a change :) 2021-02-10 17:41:25 @dillonthehuman And what about actual rockstars? Are they like, "Man, Jimi Hendrix was a total 10x developer"? :-p 2021-02-10 17:39:33 @AndrewMoody_ In the military, what would get someone called a rockstar? 2021-02-10 17:38:12 @dillonthehuman YES! What about the geologists? They absolutely *should* use the term. 2021-02-10 17:34:48 Just wondering, do all fields call people who do good work in their area "rockstars" and similar? This has always struck me as... well... strange, because even the coolest statisticians are closer to the nerd end of the nerd-rockstar spectrum. 2021-02-10 14:19:33 @geomblog @TaliaRinger @ckfinite @FAccTConference Newest version is here!... though we paid to have it be open access and it appears not to be. I guess I'll go look into that... https://t.co/Wgar8huY4F 2021-02-09 16:57:01 @2plus2make5 Thank you, Emma! 2021-02-09 16:01:35 @hannahsassaman I would also like a 5% stress differential for living in Philadelphia. The stress of dealing with the drivers (and roads!) around here ought to justify at least 5% :) 2021-02-08 01:14:46 @EvanRosenman Huh. I was exactly this moment years old when I learned that. 2021-02-07 21:46:19 @HooverGreen I didn’t pay to see that movie but I still feel like I was charged too much it was so bad 2021-02-06 19:51:00 RT @jennwvaughan: While I appreciate that the @nytimes devoted a full section to how brutal the past year has been for moms, these articles… 2021-02-06 18:44:11 @niljanaakpinar @mdekstrand @FAccTConference @mariadearteaga @achould Congrats, you three! 2021-02-05 22:41:28 @yudapearl RAI is "Risk Assessment Instrument". It is defined in literally the first sentence of the paper you are commenting on. 2021-02-05 20:20:56 @yudapearl Really liked the interactive visualizations in the first link, btw! 2021-02-05 20:20:42 @yudapearl In this case, we weren't trying to make any causal statements-- just evaluate the between-individual variability that is not explained by covariates. Though counterfactual analysis has been applied to RAIs, imo, the approach is under-utilized and could contribute a lot. 2021-02-05 19:00:11 @cbrew Honestly, not even fully sure I understand this, but that won't stop me from using it :) 2021-02-05 16:48:20 RT @mdekstrand: In this paper we make three main contributions to the study of provider-side fairness in recommender systems (with likely… 2021-02-05 03:38:18 @mdekstrand @andrewthesmart Best. Phrasing. Ever. 2021-02-04 20:15:03 RT @lauramoy: In the article, I propose a taxonomy of five major equity problems that appear repeatedly across different police technologie… 2021-02-04 20:14:12 @lauramoy @UIllLRev Congratulations, Laura! This is awesome! 2021-02-04 18:15:35 RT @Ira_Sass: @KLdivergence tired: including the word "desiderata" in paper wired: including the entire text of the 1921 prose poem "Deside… 2021-02-04 18:15:14 @BaiocchiMike I'm sorry, Mike. Are you in this tweet? 2021-02-04 17:52:08 @acidflask AUTOMATIC REJECT 2021-02-04 17:43:24 Crap, just realized I did not include the word "desiderata" in my latest fair ML paper. Major oversight on my part. Huge violation of community norms. 2021-02-04 01:17:31 @jxnlco @LaytonEWilliams It’s devastating, I know. 2021-02-04 00:30:52 @LaytonEWilliams Grant Imahara. 2021-02-03 22:23:19 @RDBinns It's one of those papers that as I'm reading it, I'm like "yes, yes, very good, all makes sense" and then if you ask me to tell you what it said, I'm like "uhhh so.... it's complicated" 2021-02-03 22:13:27 @RDBinns Uh, I mean the Dawid one... not my own.... just in case that wasn't clear :) 2021-02-03 22:13:02 @RDBinns You just linked to one of my current favorite papers! 2021-02-03 19:56:14 @mdekstrand For anyone who's interested, I found this paper super interesting and clear (at least while I was reading it, though later hard to regurgitate what I'd learned... that's how you know you've hit something deep :) https://t.co/B2dosT5WsE 2021-02-03 19:43:30 @mdekstrand Not nearly enough discussion of the philosophy of probability, though! So much to be said there! 2021-02-03 19:23:10 @vm_wylbur @JamesJohndrow @DavidDunson7 Thanks, PB! 2021-02-03 19:07:00 @hannawallach Bummer! Would be a complete Dream. Come. True. to collab. Hit me up if your plate ever clears?? 2021-02-03 19:04:38 @MeganTStevenson I'd guess there's probably a sprinkling of those things. Probably some irreducible error thrown in there too. 2021-02-03 18:59:36 @eleafeit @JamesJohndrow made a similar observation re: online marketing data over the course of this project 2021-02-03 18:42:37 @DavidDunson7 @JamesJohndrow Ditto! Let's do it again :) 2021-02-03 18:39:49 If there are any takers, send me an email! I'm not that hard to find. (And if for some reason I am, my DMs are open) 2021-02-03 17:53:04 @Aaroth @geomblog You are so right. 2021-02-03 17:48:23 @geomblog I could get Jac to do vehicle-related captchas all day. Might be a good rainy day activity :) 2021-02-03 17:27:51 Btw, I think this paper hints some ways in which Bayesian statistics can contribute to the algorithmic fairness literature more generally outside of risk assessment that haven't yet been fully explored. If anyone's interested in thinking about this further, don't be a stranger! https://t.co/KrUDlZSXl6 2021-02-03 16:36:43 @ryanstravis @JamesJohndrow @DavidDunson7 Thank you! 2021-02-03 14:49:40 RT @deaneckles: Valuable use of repeated observations + random effects models to quantify individual heterogeneity missed by only looking a… 2021-02-03 14:43:52 @Ira_Sass Thanks, Ira! And, thanks for the super careful read through. The paper is better because of it! 2021-02-03 02:47:50 @mdekstrand Thanks, Michael! I'll be interested to get your feedback. :) 2021-02-03 02:45:06 And @Aaron_Horowitz!! 2021-02-03 02:37:07 If you've gotten this far (first, thank you) and you're still confused, this is a tricky topic that's hard to do justice in tweets. I promise the paper does a better job of explaining it all. Please take a look! :) 2021-02-03 02:37:06 This has been a long-standing debate in this area, which is fundamentally unresolvable without a dataset with multiple observations per person. Well, guess what we got.... 2021-02-03 02:37:05 Really excited about some recent work (hot off the arxiv presses) on individual heterogeneity in risk assessment models https://t.co/GZwHEBeWfv with @JamesJohndrow and @DavidDunson7 2021-02-03 00:37:27 @acidflask Congratulations! 2021-02-02 15:59:39 @AJWoStat I mean, it was a really neat idea! 2021-02-02 15:48:04 @HBWHBWHBW This is my steady state. 2021-02-02 15:45:27 @Aaroth @GermsAndNumbers will remember this. It was around the same time he invented hash tables, i.e. 2012 :) 2021-02-02 15:42:35 @geomblog https://t.co/7EIQ7MwFEP 2021-02-02 15:40:15 @Aaroth One time I re-invented propensity scores. On the one hand, if I had thought of this earlier, I could have been Don Rubin (well, ahem, I would have been different in some ways). On the other, I would have had to have invented it before I was born, so I can't be too upset about it 2021-02-02 15:36:31 You know that feeling when you have a really good research idea that seems so good it's impossible that no one's thought of it before... and yet, you can't seem to find it anywhere? It's kind of a combination of excitement and fear. 2021-02-02 15:32:36 @jseabold Right!?!! If ever there were a time to be frolicking outside marveling at the beauty of the world, it is now. Instead, I'm sitting inside staring at a screen. :/ 2021-02-02 15:00:35 @skoularidou I'm good with that. 2021-02-02 14:20:36 May I humbly suggest the convention that when there’s heavy snow, we call a snow day and cancel all meetings, regardless of whether we’re working from home? https://t.co/ars5px0teh 2021-01-29 23:11:36 Why am I always the LAST person to exit out of a zoom meeting? Are all of you just sitting there with your mouse hovering over the end button, or what? 2021-01-29 19:18:42 @Aaron_Horowitz @rajiinio @mmitchell_ai I strangely feel proud that my research shit posting is being acknowledged. 2021-01-29 19:14:22 @Aaron_Horowitz @rajiinio @mmitchell_ai 2021-01-29 17:08:29 @kdphd @geomblog Sad! I miss you all too :( 2021-01-27 18:18:15 @iskander Congratulations! They're beautiful! 2021-01-27 17:40:49 @LissackMichael @pmddomingos @timnitGebru Stochastic Parrots has nearly 160 citations. In the context of this conversation about ML citing too few papers, if anything, this should be a counter-example of shoddy work by the field. 2021-01-27 17:23:41 @tdietterich @geomblog And watch me *not* be a computer scientist in 3 years when re-appointment comes up... lol 2021-01-27 17:20:35 @PogrebnyakE I think on that website, a citation is how many times each paper was cited by another paper. A reference is how many papers each paper cites. So, the first is "incoming" citations. The latter is "outgoing". 2021-01-27 17:11:25 @geomblog Yes, this is step 2: figure out how to structure my work around conference deadlines. I still am in the mode of just calling it good when I think it's done, but I am getting the sense that won't serve me in this field. 2021-01-27 17:03:44 @tdietterich Yeah, it very well might be. But, the tendency towards journals or conferences is itself a cultural norm, so I still find it really interesting the difference in expectations here. 2021-01-27 17:00:27 @BrownSarahM Certainly interesting context! But also, that was a choice by the field that apparently stats did not make. 2021-01-27 16:00:28 For reference, this is the latest issue of JASA, a flagship journal for stats. It doesn't have a nice figure on number of refs per paper, but some clicking around leaves the impression of about 35-70 citations per paper. https://t.co/2mbrFwFcSi 2021-01-27 15:54:57 Yes, I am fully aware that conference papers tend to be shorter. But, STILL! 2021-01-27 15:54:42 In my quest to become a real computer scientist, I've been looking into community norms, like expectations for number of references per paper. The difference between ML and stats is huge. I usually think of 30-40 citations as the low end in stats. https://t.co/T7kikbREmW 2021-01-25 18:24:16 @KenButler12 That's how it was when I applied too 2021-01-25 16:06:33 One thing I didn't anticipate about being back academia: the sheer number of cold emails I get from people (students) trying to get into the program. I literally want to help *every* *single* *one* of them, and I can't. So, I mostly just ignore them and feel terrible about it. 2021-01-25 02:41:18 @kat_heller @DrLukeOR Hi 2021-01-24 00:09:06 @jingyingyang @PartnershipAI @GeorgiaTech Congratulations! 2021-01-22 23:23:54 @hannahsassaman YES PLEASE 2021-01-22 22:14:22 @AJWoStat Obviously so do I. 2021-01-22 20:58:33 I swear, if Renaissance Fair organizers are not planning plague themed events for 2021 (assuming it's safe to go this year), they are missing a huge opportunity. 2021-01-22 20:03:33 @Aaron_Horowitz Are you seriously subtweeting me? 2021-01-22 15:08:54 One kids is so many kids. https://t.co/KuNlRO1f2S 2021-01-21 18:14:41 I've literally seen papers where the positive outcome is described as "getting bail". Meanwhile.... https://t.co/LqNIlV82me https://t.co/Fz60aaiu1s 2021-01-21 17:46:28 Been reading some old fair ML papers recently that, of course, use the COMPAS data set and.... you all realize that for many being "granted bail" is not actually a great outcome, right? Maybe you, ML researcher, assume you'd be able to pay it, but that's not true for everybody. https://t.co/HVNFIzaX2h 2021-01-21 16:37:56 This is awesome. It would be great if this working group can consult with academia and civil society to figure out what data the folks in those spaces need to study, provide evidence of, and remedy the inequity this mandate is meant to address. https://t.co/OsXheVXNSG 2021-01-21 13:20:17 @HooverGreen @_5u11y @StatMegan @vm_wylbur I love it! 2021-01-20 18:15:27 @2plus2make5 @KevinKaichuang 2021-01-20 17:00:37 For the third inauguration of a new administration in a row, I'm crying. The first was tears of joy, the second tears of fear. Today, these are just tears of relief. 2021-01-20 13:49:28 The smile on my face right now as I watch Trump leave for the last time to YMCA blasting in the background. “Young man!” 2021-01-20 03:43:00 @brianchristian @wsisaac @geoffreyirving That’s so kind! But if you put in the work to write a book I’m happy to pay for it. And I will. 2021-01-19 19:33:15 I'm concerned. https://t.co/PDkVwFpycu 2021-01-19 19:32:38 @zschutzy Do they have different expiration dates? 2021-01-19 19:32:20 @zschutzy Getting down to the important business post-proposal, I see? :D 2021-01-19 18:16:25 @jaydaronson That sounds like a good thing? I read "declaring genocide" as similar to "declaring war" and.... yeah. 2021-01-19 17:46:10 @jaydaronson jeez, what did I miss? 2021-01-15 14:34:15 @Aaron_Horowitz Narrator: it is not. 2021-01-14 22:03:58 @BrownSarahM Wow! You are far too kind. I am so excited to collaborate with you. I'd been hoping this would happen since we spoke at KDD in Halifax :) 2021-01-14 16:40:58 @KylaChasalow @karen_ec_levy @FAccTConference On my list it goes! Looks awesome! 2021-01-13 17:54:51 @rcalo @MartinKaste I mean, Ryan, he's got a point. Seems pretty accurate to me. Do you DENY THE FACTS??? 2021-01-13 16:36:33 @rcalo That was supposed to be "when I inevitably write 2020" when I date things... 2021-01-13 15:07:13 @rcalo To be honest, it will be easier to pop in the 3 than cross out the zero when I inevitably write when I date things 2021-01-12 16:39:18 @Aaron_Horowitz And what a great thing it was that you did! Our collabs have been a highlight of what was otherwise a pretty craptastic year. 2021-01-12 16:23:58 Say what you will about the cesspool that is Twitter dot com... but for me it has been great for collaborative research/work during COVID. Today, I'm working on an op-ed, and tomorrow/Thursday I have meetings to kick off new collabs-- all of which were dreamed up right here. https://t.co/A7BiX1cUSd 2021-01-10 20:30:46 @scheidegger @kdphd Uhh hi. Maybe I can get in on this?? 2021-01-10 13:49:26 @geomblog @Aaron_Horowitz Move along. Nothing to see here. 2021-01-10 13:49:17 @geomblog @Aaron_Horowitz Probably an algorithm picked it :) 2021-01-05 21:46:32 @ARGohdes Honestly I think slow is the way to go if you can afford it 2021-01-05 20:57:58 Nearly 4 months later and this paper I've been working on for over a year (and is only 25 pages) is nearly complete. I may not crank out work that fast, but the stuff I do put out there I am proud of. https://t.co/xgsoviZSR3 2021-01-03 03:42:24 @ruchowdh @geomblog Link? 2021-01-03 03:22:42 @ruchowdh @geomblog Lol what? The only words I understood in all of that were frequentism and Bayesianism. 2021-01-03 03:19:54 @generativist I did warn you... 2021-01-02 19:00:57 @hspter I’m on it! 2021-01-02 18:55:24 @hspter I’ve tweeted in support of Chesa. 2021-01-02 18:50:45 It’s been a long December but there’s reason to believe maybe this year will be better than the last. 2021-01-01 12:55:15 It’s 2021! Is everything suddenly great now, or.... 2020-12-31 18:27:32 Building on the best papers of 2020 requests: tell me about something you read in 2020 whose premise or conclusions you disagree with but you have to admit they make a good point? 2020-12-31 01:33:28 @malleshpai @Aaroth @JubaZiani @zstevenwu @Ira_Sass Shows what I know! 2020-12-30 22:14:28 @hannahsassaman @easears @bscottmercer That sounds lovely. Will signal you to set up! 2020-12-30 22:07:57 @Ira_Sass @JubaZiani @zstevenwu @Aaroth LOL, you have a spreadsheet!?! I love it! 2020-12-30 22:07:27 @malleshpai @Aaroth @JubaZiani @zstevenwu @Ira_Sass I love the simplicity of calling it "Indian Restaurant". Not great SEO though to find that specific one. 2020-12-30 22:06:42 @easears @hannahsassaman @bscottmercer You sent me an extensive list! Once the pandemic is over I plan to hit them all up, though I guess I could do takeout before then... 2020-12-30 21:50:41 @JubaZiani @zstevenwu @Ira_Sass @Aaroth I thought Thanal was a step up from meh but not great. I think I got spoiled living in RDU. The Indian food there was really, really good. 2020-12-30 21:48:30 @zstevenwu @Ira_Sass @Aaroth OK, locals... I guess I should have asked this sooner, but where is the good Indian food? I've had some up near me (Germantown) and it was..... not great. Yes, I've tried Tiffin. It was meh. 2020-12-30 21:47:10 @malleshpai @Aaroth @zstevenwu big yikes 2020-12-30 21:39:24 @malleshpai @Aaroth @zstevenwu It was Fung's Kitchen. 2020-12-30 21:32:57 @malleshpai @Aaroth @zstevenwu Honestly, since it was already paid for, I ate on campus 99% of the time and sampled very little of Houston's restaurants. I did used to go out for dim sum occasionally on the weekends. I just texted my friend to try to remember the name of the place! 2001-01-01 01:01:01