Sayash Kapoor @sayash.bsky.social · 07/07/2026Thrilled to share that I am joining UC Berkeley as an Assistant Professor in the School of Information! I start in Fall 2027, and I am recruiting PhD students this cycle. List me in your application if you want to work with me! More on what I'm looking for (and a form to indicate interest) 🧵 5637
Reposted by Sayash KapoorMelanie Mitchell @melaniemitchell.bsky.social · 23/05/2026Another excellent post (from @sayash.bsky.social and @randomwalker.bsky.social) on questions to ask about announcements of AI capabilities. I highly recommend their newsletter! www.normaltech.ai/p/did-google...normaltech.aiDid Google’s AI agents really build an operating system for $916?The importance of independent evaluation 1328
Reposted by Sayash KapoorKaty Glenn Bass @katygb.bsky.social · 21/05/2026New from @randomwalker.bsky.social and @sayash.bsky.social: Responding to AI risks requires building broad resilience to harms through normal policymaking, rather than allowing the executive branch to exert extraordinary power over AI access and development. knightcolumbia.org/blog/do-ai-r...knightcolumbia.orgDo AI Risks Require Extraordinary Government Intervention? 0123
Reposted by Sayash KapoorJameel Jaffer @jameeljaffer.bsky.social · 21/05/2026New Trump EO is expected to "ask" AI cos to submit frontier models for agency review in advance of release. Here's a new essay--just published--by @randomwalker.bsky.social & @sayash.bsky.social about whether AI risks require extraordinary government intervention. knightcolumbia.org/blog/do-ai-r...knightcolumbia.orgDo AI Risks Require Extraordinary Government Intervention? 062
Reposted by Sayash KapoorAyla Taylor @aylataylor.bsky.social · 03/03/2026A silly little comic based on the opening section of AI Snake Oil by @randomwalker.bsky.social and @sayash.bsky.social. 13114
Reposted by Sayash KapoorLawfare @lawfaremedia.org · 12/02/2026Advanced AI will not automatically help consumers achieve desired legal outcomes at lower costs. Justin Curl, @sayash.bsky.social, & @randomwalker.bsky.social identify three bottlenecks that stand between AI capability advances and more accessible legal services in the latest Lawfare Research paper.lawfaremedia.orgAI Won’t Automatically Make Legal Services CheaperThree bottlenecks between AI capability and access to justice. 1134
Reposted by Sayash KapoorJason W. Burton @jasonburton.bsky.social · 19/05/2025(1/4) Ever wondered what tech policy might look like if it were informed by research on collective intelligence and complex systems? 🧠🧑💻 Join @jbakcoleman.bsky.social, @lukethorburn.com, and myself in San Diego on Aug 4th for the Collective Intelligence x Tech Policy workshop at @acmci.bsky.social! 11512
Reposted by Sayash KapoorPrinceton Center for Information Technology Policy @princetoncitp.bsky.social · 09/04/2025New commentary in @nature.com from professor Arvind Narayanan (@randomwalker.bsky.social) & PhD candidate Sayash Kapoor (@sayash.bsky.social) about the risks of rapid adoption of AI in science - read: "Why an overreliance on AI-driven modelling is bad for science" 🔗 #CITP #AI #science #AcademiaSkynature.comWhy an overreliance on AI-driven modelling is bad for scienceWithout clear protocols to catch errors, artificial intelligence’s growing role in science could do more harm than good. 02112
Reposted by Sayash KapoorKnight First Amendment Institute @knightcolumbia.org · 15/04/2025In a new essay from our "Artificial Intelligence and Democratic Freedoms" series, @randomwalker.bsky.social & @sayash.bsky.social make the case for thinking of #AI as normal technology, instead of superintelligence. Read here: knightcolumbia.org/content/ai-a...knightcolumbia.orgAI as Normal Technology 13717
Reposted by Sayash KapoorPrinceton Laboratory for Artificial Intelligence @princetonainews.bsky.social · 16/04/2025“The rush to adopt AI has consequences. As its use proliferates…some degree of caution and introspection is warranted.” In a comment for @nature.com, @randomwalker.bsky.social and @sayash.bsky.social warn against an overreliance on AI-driven modeling in science: bit.ly/4icM0hpbit.lyWhy an overreliance on AI-driven modelling is bad for scienceWithout clear protocols to catch errors, artificial intelligence’s growing role in science could do more harm than good. 064
Reposted by Sayash KapoorDr Abeba Birhane @abeba.blacksky.app · 09/04/2025Science is not collection of findings. Progress happens through theories.As we move from findings to theories things r less amenable to automation. Proliferation of scientific findings based on AI hasn't accelerated—& might even have inhibited—higher levels of progress www.nature.com/articles/d41...nature.comWhy an overreliance on AI-driven modelling is bad for scienceWithout clear protocols to catch errors, artificial intelligence’s growing role in science could do more harm than good. 312049
Sayash Kapoor @sayash.bsky.social · 03/02/2025I spent a few hours with OpenAI's Operator automating expense reports. Most corporate jobs require filing expenses, so Operator could save *millions* of person-hours every year if it gets this right. Some insights on what worked, what broke, and why this matters for the future of agents 🧵 63910
Reposted by Sayash KapoorMelanie Mitchell @melaniemitchell.bsky.social · 19/12/2024Excellent post discussing whether "AI progress is slowing down". www.aisnakeoil.com/p/is-ai-prog... And if you're not subscribed to @randomwalker.bsky.social and @sayash.bsky.social 's great newsletter, what are you waiting for?aisnakeoil.comIs AI progress slowing down?Making sense of recent technology trends and claims 05615
Reposted by Sayash KapoorArvind Narayanan @randomwalker.bsky.social · 18/12/2024Excited to share that AI Snake Oil is one of Nature's 10 best books of 2024! www.nature.com/articles/d41... The whole first chapter is available online: press.princeton.edu/books/hardco... We hope you find it useful. 513930
Sayash Kapoor @sayash.bsky.social · 16/12/2024More than 60 countries held elections this year. Many researchers and journalists claimed AI misinformation would destabilize democracies. What impact did AI really have? We analyzed every instance of political AI use this year collected by WIRED. New essay w/@random_walker: 🧵 38336
Reposted by Sayash KapoorArvind Narayanan @randomwalker.bsky.social · 15/12/2024My wife noticed an article with my name on it in her feed and said “Did you write this? It doesn’t sound like you.” I was surprised to see an alarmist headline attributed to me and @sayash.bsky.social by Wired: “Human Misuse Will Make Artificial Intelligence More Dangerous” 14812
Reposted by Sayash KapoorStefano Maffulli @maffulli.net · 15/12/2024Misinformation is not an AI problem. @sayash.bsky.social and @randomwalker.bsky.social highlight that "while generative AI reduces the cost of creating misinformation, it does not reduce the cost of distributing it". Always worth reading them. substack.com/app-link/pos...substack.comWe Looked at 78 Election Deepfakes. Political Misinformation is not an AI Problem.Technology Isn’t the Problem—or the Solution. 094
Reposted by Sayash KapoorJameel Jaffer @jameeljaffer.bsky.social · 13/12/2024Is artificial intelligence supercharging the problem of political misinformation? In a word: no. New from @randomwalker.bsky.social & @sayash.bsky.social at @knightcolumbia.org. knightcolumbia.org/blog/we-look...knightcolumbia.orgWe Looked at 78 Election Deepfakes. Political Misinformation Is Not an AI Problem. 4248
Reposted by Sayash KapoorKnight First Amendment Institute @knightcolumbia.org · 13/12/2024“Fixes to the information environment depend on structural and institutional changes rather than on curbing AI-generated content.” @sayash.bsky.social & @randomwalker.bsky.social find that AI hasn’t "fundamentally changed the landscape of political misinformation”: knightcolumbia.org/blog/we-look...knightcolumbia.orgWe Looked at 78 Election Deepfakes. Political Misinformation Is Not an AI Problem. 02213
Reposted by Sayash KapoorSara Hooker @sarahooker.bsky.social · 04/12/2024AI amplifying biorisk has been a major focus in AI policy & governance work. Is the spotlight merited? Our recent cross-institutional work asks: Does the available evidence match the current level of attention? 📜 arxiv.org/abs/2412.01946 25912
Reposted by Sayash KapoorLuiza Jarovsky, PhD @luizajarovsky.bsky.social · 29/11/2024🚨 [AI BOOK CLUB] "AI Snake Oil: What AI Can Do, What It Can’t, and How to Tell the Difference" by @randomwalker.bsky.social & @sayash.bsky.social is a MUST-READ for everyone interested in AI, and it's our 🎉 15th selected book: 📖 About the book: 22610
Reposted by Sayash Kapoorklaudia jaźwińska @klaudia.bsky.social · 27/11/2024I have a new piece out with @aisvarya17.bsky.social in @columjournreview.bsky.social in which we test how OpenAI's new search feature surfaces and attributes news content. Our findings were not promising for news publishers (1/9) www.cjr.org/tow_center/h...cjr.orgHow ChatGPT (Mis)represents Publisher ContentChatGPT search — which is positioned as a competitor to search engines like Google and Bing — launched with a press release from OpenAI touting claims that the company had “collaborated extensively wi... 817584
Reposted by Sayash KapoorArvind Narayanan @randomwalker.bsky.social · 27/11/2024New short paper on the limits of one type of inference scaling, by @benediktstroebl.bsky.social @sayash.bsky.social & me. The first page has the main findings and message. (The title is a play on Inference Scaling Laws.) More on the limits of inference scaling coming soon. arxiv.org/abs/2411.17501 45510
Reposted by Sayash KapoorDream Wilson @dreamstatic.bsky.social · 23/11/2024Current read. Very informative so far. Not quite in the way most folks I engage with would expect. I knew predictive AI was bad but I did not know just how bad it is. Or how much government and corporations use it and screw up lives in the process. 1269
Reposted by Sayash KapoorOve Christensen @ovechr.bsky.social · 23/11/2024Don't leave the future of AI to the tech companies. The authors of AI Snake Oil cut through the hype of AI-bullshit. Not to neglect the good stuff but in order to make the reader aware of the (snake)pits of overselling the technology. @sayash.bsky.social www.linkedin.com/pulse/dont-l... #skolechatlinkedin.comDon't leave the future of AI to the tech companiesBook review: Arvind Narayanan & Sayash Kapoor, AI Snake Oil. What Artificial Intelligence Can Do, What It Can't and How to Tell the Difference. 1142
Reposted by Sayash KapoorDirk Eddelbuettel @eddelbuettel.com · 20/11/2024Another banger just picked up from my local library: @randomwalker.bsky.social and @sayash.bsky.social on "AI Snake Oil". 091
Reposted by Sayash KapoorEerke Boiten @eerkeboiten.bsky.social · 19/11/2024I have recently read the book "AI Snake Oil" by @randomwalker.bsky.social and @sayash.bsky.social. Before buying it, I expected it to be the best book I'd read on AI, it most definitely is. I heartily endorse it to anyone wanting to find out what AI is all about and see what is real and what is hype 1142
Reposted by Sayash Kapoorsasha @sashax.bsky.social · 18/11/2024This book was surprisingly good! Excited to go to this. 151
Sayash Kapoor @sayash.bsky.social · 18/11/2024Folks in San Francisco: my AI Snake Oil book talk is *today* at 5:30pm at Book Passage (SF ferry building). Come through to discuss the future of AI, why AI isn't an existential risk, how we can build AI in/for the public, and what goes into writing a book. Looking forward to seeing some of you! 1111
Reposted by Sayash KapoorStefanie Sirén-Heikel @mssiren.bsky.social · 18/11/2024This easily approachable book by @randomwalker.bsky.social & @sayash.bsky.social should be compulsory reading for anyone implementing AI and automation. Especially everyone in the EU wishing to push full throttle on applying AI for everything and anything. www.aisnakeoil.comaisnakeoil.comAI Snake Oil | Sayash Kapoor | SubstackWhat Artificial Intelligence Can Do, What It Can’t, and How to Tell the Difference. Click to read AI Snake Oil, a Substack publication with tens of thousands of subscribers. 2278
Reposted by Sayash KapoorAlan the Baltimoron @bmorons.bsky.social · 18/11/2024Just finishing @randomwalker.bsky.social and @sayash.bsky.social's book AI Snake Oil. It's really good - clear eyed and thoughtful. Their chapter on content moderation is a must read for anyone engaging with social media, and vital user education for us all in the growth of a platform like this. 062
Reposted by Sayash KapoorAlexander Wuttke @kunkakom.bsky.social · 15/11/2024I learned much from this book on AI. - How to Identify Trustworthy / Overhyped AI Applications - How AI should (not) be regulated - The Reproducibility Crisis in AI Research - AI Research Culture Great work by @randomwalker.bsky.social , @sayash.bsky.social 1558
Reposted by Sayash KapoorBecca Ricks @baricks.bsky.social · 16/11/2024(3) AI Snake Oil by Arvind Narayanan @randomwalker.bsky.social & Sayash Kapoor @sayash.bsky.social. Building on a 2019 talk, a key takeaway for me was that AI snake oil is appealing because it's being marketed to broken institutions that are already overworked. press.princeton.edu/books/hardco...press.princeton.eduAI Snake OilFrom two of TIME’s 100 Most Influential People in AI, what you need to know about AI—and how to defend yourself against bogus AI claims and products 162
Reposted by Sayash KapoorKenneth Marino @kennethmarino.bsky.social · 16/11/2024Going to do my part to try to build content on Bluesky by trying to post a paper I like every day. First is one of the best papers I’ve read in a while doing essential work in examining how we evaluate agents: arxiv.org/html/2407.01... From @sayash.bsky.social @randomwalker.bsky.social at Princetonarxiv.orgAI Agents That Matter 481
Reposted by Sayash KapoorDr Abeba Birhane @abeba.blacksky.app · 12/11/2024"It turns out that the algorithm is mostly just assessing need, that is, how long patients would survive without a transplant." this is a damming essay about UK's liver transplant matching algo www.aisnakeoil.com/p/does-the-u...aisnakeoil.comDoes the UK’s liver transplant matching algorithm systematically exclude younger patients?Seemingly minor technical decisions can have life-or-death effects 28542
Reposted by Sayash KapoorArvind Narayanan @randomwalker.bsky.social · 12/11/2024New essay in the AI Snake Oil newsletter with Angelina Wang, @sayash.bsky.social and Solon Barocas — we analyze the UK's liver transplant matching algorithm that's been reported to make it nearly impossible for patients under 45 to get a liver. www.aisnakeoil.com/p/does-the-u...aisnakeoil.comDoes the UK’s liver transplant matching algorithm systematically exclude younger patients?Seemingly minor technical decisions can have life-or-death effects 02214
Reposted by Sayash KapoorShannon Vallor @shannonvallor.bsky.social · 06/08/2024A bit over the moon to see The AI Mirror featured in The New Yorker today (alongside the good company of @randomwalker.bsky.social and @sayash.bsky.social)! www.newyorker.com/culture/open...newyorker.comIn the Age of A.I., What Makes People Unique?More than ever, we’re challenged to define what’s valuable about being human. 0172
Reposted by Sayash KapoorPrinceton University Press @princetonupress.bsky.social · 10/04/2024From two of TIME’s 100 Most Influential People in AI, AI Snake Oil by @randomwalker.bsky.social & @sayash.bsky.social publishes on September 24. 🎉 Learn what #AI can do, what it can’t, & how to tell the difference. Preorder your copy of this important book today: princeton.press/gpl5al2h 042
Sayash Kapoor @sayash.bsky.social · 10/04/2024I'm ecstatic to share that preorders are now open for the AI Snake Oil book! The book will be released on September 24, 2024. @randomwalker.bsky.social and I have been working on this for the past two years, and we can't wait to share it with the world. Preorder: princeton.press/gpl5al2h 13712
Sayash Kapoor @sayash.bsky.social · 16/11/2023Humans in the loop are not enough to fix algorithms. When push comes to shove, companies choose profits over people. The latest example: UnitedHealth forced employees to cut off care based on an algorithm's prediction. If they disagree, they're fired.statnews.comUnitedHealth pushed employees to follow an algorithm to cut off Medicare patients' rehab careUnitedHealth executives pressured clinical staff to follow an algorithm to cut off patients’ rehab care, leading to coverage decisions that may violate Medicare regulations. 03224
Sayash Kapoor @sayash.bsky.social · 09/11/2023If you have been anywhere near AI discourse in the last few months, you might have heard that AI poses an existential threat to humanity. In today's Wall Street Journal, @randomwalker.bsky.social and I argue that claims of x-risk rest on a tower of fallacies (no paywall: archive.is/Jv60s)wsj.comDoes AI Pose an Existential Risk to Humanity? Two Sides Square OffYes, say some: The threat of mass destruction from rogue AIs—or bad actors using AI—is real. Others say: Not so fast 082
Reposted by Sayash KapoorArvind Narayanan @randomwalker.bsky.social · 24/08/2023A VentureBeat interview with @sayash.bsky.social and me about the AI Snake Oil book and Substack! venturebeat.com/ai/princeton... www.aisnakeoil.comventurebeat.comPrinceton University’s ‘AI Snake Oil’ authors say generative AI hype has ‘spiraled out of co...In a VentureBeat Q&A, Princeton University's Arvind Narayanan and Sayash Kapoor, authors of the upcoming "AI Snake Oil," discuss AI hype. 0129
Reposted by Sayash KapoorKnight First Amendment Institute @knightcolumbia.org · 31/07/2023INMA included @knightcolumbia.org's paper "How to Prepare for the Deluge of Generative AI on Social Media," written by @randomwalker.bsky.social & @sayash.bsky.social, on their AI reading list. www.inma.org/blogs/smart-data-initi…inma.orgAI reading list: vacation editionSometimes a deep, thoughtful dive into the biggest topic in media — and beyond — requires the leisurely pace of summer days. So, here’s your summer AI reading list. 153
Reposted by Sayash KapoorMargaret Mitchell @mmitchell.bsky.social · 07/07/2023If you're a journalist or news editor covering AI, please consider reading this. Thanks to Hilke for including me, Ari & @sayash.bsky.social for their work... www.cjr.org/analysis/how-to-report-…cjr.org 12515