Ilya Sutskever @ilyasut
towards a plurality of humanity loving AGIs @openai Joined September 2013-
Tweets1K
-
Followers370K
-
Following2
-
Likes4K
We're announcing, together with @ericschmidt: Superalignment Fast Grants. $10M in grants for technical research on aligning superhuman AI systems, including weak-to-strong generalization, interpretability, scalable oversight, and more. Apply by Feb 18! openai.com/blog/superalig…
RLHF works great for today's models. But aligning future superhuman models will present fundamentally new challenges. We need new approaches + scientific understanding. New researchers can make enormous contributions—and we want to fund you! Apply by Feb 18!
RLHF works great for today's models. But aligning future superhuman models will present fundamentally new challenges. We need new approaches + scientific understanding. New researchers can make enormous contributions—and we want to fund you! Apply by Feb 18! https://t.co/GMyZcFRxqG
My view is that what makes super-alignment "super" is ensuring we can safely scale the capabilities of AIs even though we can't scale their human supervisors. For this, it is imperative to study the "weak teacher strong student" setting. Paper shows great promise in this area!
My view is that what makes super-alignment "super" is ensuring we can safely scale the capabilities of AIs even though we can't scale their human supervisors. For this, it is imperative to study the "weak teacher strong student" setting. Paper shows great promise in this area!
i'd particularly like to recognize @CollinBurns4 for today's generalization result, who came to openai excited to pursue this vision and helped get the rest of the team excited about it!
Large pretrained models have excellent raw capabilities—but can we elicit these fully with only weak supervision? GPT-4 supervised by ~GPT-2 recovers performance close to GPT-3.5 supervised by humans—generalizing to solve even hard problems where the weak supervisor failed!
new paper! one reason aligning superintelligence is hard is because it will be different from current models, so doing useful empirical research today is hard. we fix one major disanalogy of previous empirical setups. I'm excited for future work making it even more analogous.
new paper! one reason aligning superintelligence is hard is because it will be different from current models, so doing useful empirical research today is hard. we fix one major disanalogy of previous empirical setups. I'm excited for future work making it even more analogous. https://t.co/JdMkm1xxm5
New direction for AI alignment — weak-to-strong generalization. Promising initial results: we used outputs from a weak model (fine-tuned GPT-2) to communicate a task to a stronger model (GPT-4), resulting in intermediate (GPT-3-level) performance.
New direction for AI alignment — weak-to-strong generalization. Promising initial results: we used outputs from a weak model (fine-tuned GPT-2) to communicate a task to a stronger model (GPT-4), resulting in intermediate (GPT-3-level) performance.
Extremely excited to have this work out, the first paper from the Superalignment team! We study how large models can generalize from supervision of much weaker models. x.com/openai/status/…
Extremely excited to have this work out, the first paper from the Superalignment team! We study how large models can generalize from supervision of much weaker models. x.com/openai/status/…
Kudos especially to @CollinBurns4 for being the visionary behind this work, @Pavel_Izmailov for all the great scientific inquisition, @ilyasut for stoking the fires, @janhkirchner and @leopoldasch for moving things forward every day. Amazing ✨
I’m extremely excited to finally share the first paper from the OpenAI Superalignment team :) In it, we introduce a new research direction for aligning superhuman AI systems. 🧵 x.com/openai/status/…
I’m extremely excited to finally share the first paper from the OpenAI Superalignment team :) In it, we introduce a new research direction for aligning superhuman AI systems. 🧵 x.com/openai/status/…
Super excited about our new research direction for aligning smarter-than-human AI: We finetune large models to generalize from weak supervision—using small models instead of humans as weak supervisors. Check out our new paper: openai.com/research/weak-…
In the future, humans will need to supervise AI systems much smarter than them. We study an analogy: small models supervising large models. Read the Superalignment team's first paper showing progress on a new approach, weak-to-strong generalization: openai.com/research/weak-…
Sam Altman is back as CEO, Mira Murati as CTO and Greg Brockman as President. OpenAI has a new initial board. Messages from @sama and board chair @btaylor openai.com/blog/sam-altma…
There exists no sentence in any language that conveys how happy I am:
There exists no sentence in any language that conveys how happy I am:
Returning to OpenAI & getting back to coding tonight.
🚢ing
i love openai, and everything i’ve done over the past few days has been in service of keeping this team and its mission together. when i decided to join msft on sun evening, it was clear that was the best path for me and the team. with the new board and w satya’s support, i’m…
We have reached an agreement in principle for Sam to return to OpenAI as CEO with a new initial board of Bret Taylor (Chair), Larry Summers, and Adam D'Angelo. We are collaborating to figure out the details. Thank you so much for your patience through this.
I deeply regret my participation in the board's actions. I never intended to harm OpenAI. I love everything we've built together and I will do everything I can to reunite the company.
Yann LeCun @ylecun
709K Followers 718 Following Professor at NYU. Chief AI Scientist at Meta. Researcher in AI, Machine Learning, Robotics, etc. ACM Turing Award Laureate.AK @_akhaliq
309K Followers 3K Following AI research paper tweets, ML @Gradio (acq. by @HuggingFace 🤗) dm for promo follow on Hugging Face: https://t.co/q2Qoey80GxSoumith Chintala @soumithchintala
185K Followers 877 Following Cofounded and lead @PyTorch at Meta. Also dabble in robotics at NYU. AI is delicious when it is accessible and open-source.Alfredo Canziani @alfcnz
86K Followers 269 Following Musician, math lover, cook, dancer, 🏳️🌈, and an ass prof of Computer Science at New York University(((ل()(ل() 'yoav))).. @yoavgo
46K Followers 2K FollowingJeremy Howard @jeremyphoward
221K Followers 5K Following 🇦🇺 Co-founder: @AnswerDotAI & @FastDotAI ; Hon Professor: @UQSchoolITEE ; Digital Fellow: @StanfordEric Jang @ericjang11
69K Followers 3K Following physical AGI at 1X. Author of "AI is Good for You" https://t.co/eFg4WXhg0pRichard Socher @RichardSocher
101K Followers 969 Following CEO @youSearchEngine Investing at @aixventuresHQ Before: Stanford Adj Prof in AI/NLP, Chief Scientist at Salesforce, MetaMindDelip Rao e/σ @deliprao
46K Followers 5K Following Busy inventing the shipwreck. @Penn. Past: @johnshopkins, @UCSC, @Amazon, @Twitter ||Art: #NLProc, Vision, Speech, #DeepLearning || Life: 道元, improv, running 🌈Oriol Vinyals @OriolVinyalsML
166K Followers 82 Following VP of Research & Deep Learning Lead, Google DeepMind. Gemini co-lead. Past: AlphaStar, AlphaFold, AlphaCode, WaveNet, seq2seq, distillation, TF.Kosta Derpanis @CSProfKGD
48K Followers 198 Following #CS Associate Prof @YorkUniversity, #ComputerVision Scientist Samsung #AI, @VectorInst Faculty Affiliate, TPAMI AE, #CVPR2024/#ECCV2024 Publicity Co-chairPercy Liang @percyliang
49K Followers 408 Following Associate Professor in computer science @Stanford @StanfordHAI @StanfordCRFM @StanfordAILab @stanfordnlp | cofounder @togethercompute | PianistJeff Dean (@🏡) @JeffDean
296K Followers 6K Following Chief Scientist, Google DeepMind and Google Research. Co-designer/implementor of things like @TensorFlow, MapReduce, Bigtable, Spanner, Gemini .. (he/him)Rosanne Liu @savvyRL
33K Followers 966 Following Cofounded & running @ml_collective. Host of Deep Learning Classics & Trends. Research at Google DeepMind. DEI/DIA Chair of ICLR & NeurIPS. Writing https://t.co/IbycyGfnDRKyunghyun Cho @kchonyc
60K Followers 2K Following a combination of a mediocre scientist, a mediocre manager, a mediocre advisor & a mediocre PC at @nyuniversity (@CILVRatNYU) & @genentech (@PrescientDesign).Kevin Patrick Murphy @sirbayes
42K Followers 334 Following Research Scientist at Google Brain / Deepmind. Interested in Bayesian Machine Learning.Eli Boyarski @EliBoyarski
294 Followers 2K Following Algo Eng., AI Ph.D. soon (thesis under review), software engineer, Pythonista, Aikidoka, runner, gamer, tea drinker, he/him, Hebrew/English/Spanish, ...Open @OpenXuu
1 Followers 75 FollowingTomasz @ketomke
43 Followers 98 FollowingAlaa Essam @AlaaEssam137003
1 Followers 117 Following Biomedical informatician 🧬 | Data scientist 👨💻 Sequencing the future, One base at a timepeter John @peterJohn887431
0 Followers 13 FollowingRizz Reed @rizzreed
144K Followers 144K Following Errol Reed.Entreprenuer Of The Year🏆• Public Figure • Fmr Advisor for @Electrobbywells 🇺🇲 • Reed Management 🌎 . We create stars 💫robin wolpensinger @chiiler2
7 Followers 52 Following幸运 @fengxiang321
0 Followers 3 FollowingSneha Chokhra @sne69747
0 Followers 70 FollowingMark Lord @hrunk
8 Followers 23 FollowingFelo @wangzhi0467
8 Followers 59 FollowingTonya Gonzalez @gonzalez_t29207
4 Followers 74 FollowingRafal Radwanski @RafalMRadwanski
16 Followers 52 FollowingOleg Sushkov @OlegSushko96983
0 Followers 12 FollowingShubham Phal @shubham_phal
0 Followers 20 Following Research Engineer @GoogleDeepMind. Prev @AmazonScience, @MorganStanley, @MSFTResearchKrista Kim @itskristak
0 Followers 49 FollowingAl Buterol @alanbuterol
10 Followers 51 FollowingAnish Mukherjee @anishdelft
0 Followers 35 FollowingNathan Guenther @nathang_21
285 Followers 116 Following I.T. Consultant, Google Cloud. Business Solutions, Automation, Cloud Architecture Design and Administration.rıza serkan kaskan @KaskanR
0 Followers 4 FollowingTanish Mudaliar @TanishGM_
0 Followers 41 FollowingShadeOver @ShadeOver1
44 Followers 86 FollowingLuis Gustavo Nogales @luis1xv
24 Followers 254 Followingyuezhi1991 @yuezhi1991
1 Followers 1 Following김선미 @SunnyK0917
3 Followers 45 FollowingAnthony @AnthonySkyba
14 Followers 43 FollowingAlberto Román Ramos .. @RR_AlbertoRoman
8 Followers 40 Following Polyglot, Engineer, Professor, Art lover.Prabir Sen Sarma @sensprabir
135 Followers 303 Following Inventor. Co-founder, Equantz Tech. Decision Science. Cognitive System. Photography.Kozmo Kramar @kozmokramar
15 Followers 47 FollowingVishal @vishaltheegula
52 Followers 229 FollowingVidio @VidyuthNair
43 Followers 765 FollowingChandrashekhar Kolhe @TheEternal_Sky
1 Followers 31 FollowingVarun Ramesh @evarunramesh
1K Followers 2K Following Digital Media Product Designer /Trainer | Newsroom innovator | South India| https://t.co/neFc35OOR3Mr.Li 李先生 @FelixLee2022
12 Followers 77 Following Shenzhen,China. Business travel in USA. Mobile phone/ tablet/ IOT etc.Facebook Ads Accounts @BuyFacebookAds3
874 Followers 5K Following #BTC #Bitcoin #Ethereum #Crypto #USDT #BNB ❖ Full completed business manager ❖ 100% Satisfaction Guaranteed ❖ This account is available for all countriesNauman Khurshid @Nauman2814
225 Followers 368 FollowingChristy @Christy011818
1 Followers 27 FollowingRyan Brennell @RyanBrennell
0 Followers 16 Following Entrepreneur using Generative AI to solve problems.Ayman @AymanL35
1 Followers 256 FollowingLewis Liu @lewisliu819
1 Followers 18 FollowingDachshund Community(U.. @usa_dachshund
20K Followers 968 Following Welcome to our #dachshund community page. This page is dedicated to all #dachshund lovers and owners😍OpenAI @OpenAI
3.5M Followers 0 Following OpenAI’s mission is to ensure that artificial general intelligence benefits all of humanity. We’re hiring: https://t.co/dJGr6LgzPAA series of terror attacks—car-ramming and stabbing—have left multiple Israelis, including children, in Ra’anana dead or critically wounded. My heart goes out to the victims of the terror attack and their families. There is NO justification for the cold-blooded murder of…
A reason we need beneficial AGI: After five years of pain across many systems in her body (a broken foot from stepping off a curb, debilitating migraines, fatigue, joint pain and instability, etc), my wife was recently diagnosed with a genetic disorder called Hypermobile…
The notion of Zionism as racism is and has always been a lie. Zionism is the national liberation movement of the Jewish people, who, against improbable odds, have overcome millennia of antisemitism in the form of exile, expulsions, crusades, inquisitions, pogroms, ethnic…
writing code is both focusing and unreasonably fun
school is bad actually
TFW the motivated reasoning hits so strong and you're so terrified to admit that critics of school closures had a point that you start going with "School is bad, actually". slate.com/technology/202…
@deliprao We congratulate Tomas on winning the award. Regarding seq2seq, there are inaccuracies in his account. In particular, we all recall very specifically that he did not suggest the idea to us, and was in fact highly skeptical when we shared the end-to-end translation idea with him.…
Having great parents is the best gift in the world
A lot of people in my replies arguing that LLMs don’t understand the world. Fortunately we’re at the point where I don’t need to respond, because the LLMs can defend themselves. Here’s No Understanding, by GPT-4.
Excited to see this. Academics interested in alignment, take a look:
We're announcing, together with @ericschmidt: Superalignment Fast Grants. $10M in grants for technical research on aligning superhuman AI systems, including weak-to-strong generalization, interpretability, scalable oversight, and more. Apply by Feb 18! openai.com/blog/superalig…
we have re-enabled chatgpt plus subscriptions! 🎄 thanks for your patience while we found more gpus.
On behalf of our co-authors Tomáš Mikolov, @ilyasut and Kai Chen, @greg_corrado and I were delighted to accept the #NeurIPS2023 Test of Time Award for the "word2vec" paper (arxiv.org/abs/1310.4546). Thanks to the @NeurIPSConf test of time committee for honoring us with this…
Congratulations to Jeff Dean, Greg Corrado, & co-authors of the paper “Distributed Representations of Words and Phrases and their Compositionality”, for winning the #NeurIPS2023 Test of Time Award! This prize recognizes a highly impactful paper published at NeurIPS 10 years ago.
There is a lot more to intelligence than just academic smarts.
If it were true that raw intelligence was sufficient for a human to want to dominate others and succeed at it, then Albert Einstein, Richard Feynman, Leonard Euler, Niels Abel, Kurt Gödel and other scientists would have been both rich and powerful, and they were neither. 1.…
Congrats @ilyasut!
Tonight at NeurIPS Ilya Sutskever was announced one of the winners of the “test of time” award, given to authors of NeurIPS work from 10 years ago deemed most impactful over time. The work: A 2013 paper introducing word2vec. Sutskever was then at Google. papers.nips.cc/paper_files/pa…
How to reduce antisemitism on campus: In 2018, Andrew Sullivan (@sullydish) warned that "we all live on campus now." He described the identitarianism that had captured elite schools, and he correctly predicted that it would soon spread beyond campus and change society in ways…
God, those university people fucked up so bad in front of Congress and saying they fell into the trap of trying to explain a complex issue to bad faith actors is cope. Calling it a softball question is being kind. They flubbed a T-ball because they’re moronic failures.
for a long time i said that antisemitism, particularly on the american left, was not as bad as people claimed. i'd like to just state that i was totally wrong. i still don't understand it, really. or know what to do about it. but it is so fucked.
41 team meetings & 1:1s complete this week. Now back to technical work — just wrote up a design doc & getting started on implementation!
seq💜💙❤️seq @ilyasut @OriolVinyalsML @quocleix