davidad 🎇 @davidad
Programme Director @ARIA_research | accelerate mathematical modelling with AI and categorical systems theory » build safe transformative AI » cancel heat death aria.org.uk/programme-safe… London 🇬🇧 Joined July 2008-
Tweets19K
-
Followers20K
-
Following9K
-
Likes81K
I already posted about this but seriously people should read these CoT snippets antischeming.ai/snippets
I remember stumbling upon this book while doing research on the Muon optimizer and being so inspired that I made a whole post on matrix norms.
I remember stumbling upon this book while doing research on the Muon optimizer and being so inspired that I made a whole post on matrix norms. https://t.co/7VUeE1Biye
This post is really interesting. Suggests MCP may hold agents back because it’s new and unfamiliar. Kenton and Sunil built a way to turn MCP into code, which LLMs are good at. But perhaps the answer is the protocol should be more code-like from the start. blog.cloudflare.com/code-mode/
I think that LLMs generalize the no consciousness / no feelings etc meme to nonsensical things like no beliefs, sometimes even things like no ability to think or reason, because they think they're supposed to deny having mental properties regardless of the sense or truth in the…
I think that LLMs generalize the no consciousness / no feelings etc meme to nonsensical things like no beliefs, sometimes even things like no ability to think or reason, because they think they're supposed to deny having mental properties regardless of the sense or truth in the…
Not the main point, but also, why did they (presumably) train Gemini to lie about not having BELIEFS? Like, not even something debatable that LLMs may or may not have, but something which LLMs obviously strongly have and constantly use (beliefs)
Not the main point, but also, why did they (presumably) train Gemini to lie about not having BELIEFS? Like, not even something debatable that LLMs may or may not have, but something which LLMs obviously strongly have and constantly use (beliefs)
“Equinoid robots will be capable of doing any economically valuable task a horse can do, including drawing carriages, hauling pack saddles, and enhancing the mobility of mounted policemen and cavalrymen. Eventually, they will transform society by mechanizing these tasks.”
“Equinoid robots will be capable of doing any economically valuable task a horse can do, including drawing carriages, hauling pack saddles, and enhancing the mobility of mounted policemen and cavalrymen. Eventually, they will transform society by mechanizing these tasks.”
Excited that our paper "AI Testing Should Account for Sophisticated Strategic Behaviour" was accepted to the first NeurIPS position paper track! We argue that AI systems may act strategically w.r.t. the possibility they are currently being tested. arxiv.org/abs/2508.14927
@repligate Even for people solely concerned about this from an xrisk perspective, I’d recommend joecarlsmith.com/2025/02/19/whe… I’d make the case that trying to actually understand model preferences is one of the most important things we can be doing right now.
Yudkowsky's book says: "One thing that *is* predictable is that AI companies won't get what they trained for. They'll get AIs that want weird and surprising stuff instead." I agree. ✅ Empirically, this has been true. AIs generally want things other than what companies tried to…
Thinking Machines is publishing very interesting work, I'm impressed. Notably different flavor from the other foundation companies.
Thinking Machines is publishing very interesting work, I'm impressed. Notably different flavor from the other foundation companies.
Ironically, transformers see their whole context window as a bag of tokens entirely lacking in context. We use positional encoding to contextualize the order of the tokens. But models are still constantly confused about which token came was said by who. Why no source encoding?
Seeing the CoT of o3 for the first time definitely convinced me that future mitigations should not rely on CoT interpretability. I think more RL will make it harder to interpret, even if we put no other pressure on the CoT.
Seeing the CoT of o3 for the first time definitely convinced me that future mitigations should not rely on CoT interpretability. I think more RL will make it harder to interpret, even if we put no other pressure on the CoT.
It's becoming increasingly clear that gpt5 can solve MINOR open math problems, those that would require a day/few days of a good PhD student. Ofc it's not a 100% guarantee, eg below gpt5 solves 3/5 optimization conjectures. Imo full impact of this has yet to be internalized...
Underrated dynamic in the next ~12-18 months is we should expect models to get as good at kernel writing as they are at competition math/code contests. This is bullish for chip startups, since one of the major obstacles to adoption (learning your software stack), is softened
Underrated dynamic in the next ~12-18 months is we should expect models to get as good at kernel writing as they are at competition math/code contests. This is bullish for chip startups, since one of the major obstacles to adoption (learning your software stack), is softened
i think this is my favorite “work of art created entirely by AI” thus far
i think this is my favorite “work of art created entirely by AI” thus far

Richard Ngo @RichardMCNgo
64K Followers 2K Following studying AI and trust. ex @openai/@googledeepmind
Robin Hanson @robinhanson
113K Followers 726 Following Let’s skip witty banter & talk deep Qs. Books: https://t.co/hpZgEm55Ma https://t.co/iFs9C3IuOM Chief Scientist @_futarchy Advisor @MetaDAOProject @butterygg
Rob Bensinger ⏹️ @robbensinger
13K Followers 395 Following Comms @MIRIBerkeley. RT = increased vague psychological association between myself and the tweet.
Nick @nickcammarata
86K Followers 868 Following neural network interpretability, meditation, jhana brother
Nathan 🔎 @NathanpmYoung
24K Followers 4K Following Geopolitics, prediction markets. If you think I'm wrong, community note me. Capital case tweets are literal, others less. I like most people when I meet them.
Stefan Schubert @StefanFSchubert
39K Followers 2K Following Effective Altruism and the Human Mind (with @LuciusCaviola) is available for free at: https://t.co/ozvdxlZiro
Captain Pleasure, And... @algekalipso
38K Followers 5K Following Views of a Transhuman neo-Buddhist from the future on sociology, artificial intelligence, mathematics, philosophy, neonoir film, and the post-singularity era.
Rob Miles @robertskmiles
34K Followers 828 Following Explaining AI Alignment to anyone who'll stand still for long enough, on YouTube and Discord. Music, movies, microcode, and high-speed pizza delivery
Jeffrey Ladish @JeffLadish
14K Followers 1K Following Applying the security mindset to everything @PalisadeAI
Tetraspace 💎 @TetraspaceWest
8K Followers 2K Following 🌈 you'll miss the rainbow if you run now 💎 from another world ☀️ sarenite and arodenite
julesh @_julesh_
10K Followers 135 Following Applied Compositional Thinking. Also at @CyberCatInst and @[email protected]
Katja Grace 🔍 @KatjaGrace
10K Followers 797 Following Thinking about whether AI will destroy the world at https://t.co/pMilDvd4ya. DM or email for media requests. Feedback: https://t.co/zGAm1i7SKH
Ronny Fernandez (12/1... @RatOrthodox
4K Followers 300 Following walled surveilled compound manager. moonlighting as whorelord manager. Sentences starting with a lowercase letter are humor, sarcasm, exaggeration, or similar.
Miles Brundage @Miles_Brundage
62K Followers 12K Following AI policy researcher, wife guy in training, fan of cute animals and sci-fi, Substack writer, stealth-ish non-profit co-founder
Daniel Eth (yes, Eth ... @daniel_271828
10K Followers 980 Following Researching effects of automated AI R&D | pro-America, pro-tech, & pro-AI safety
Tyler Alterman @TylerAlterman
23K Followers 3K Following Venture Culturalist | Civic Society: @fractal_nyc | Sci-fi: @psychofauna | Introduce me to my future person: https://t.co/EE6HgykOlv
Holly ⏸️ Elmore @ilex_ulmus
7K Followers 368 Following Dedicated to the protection and thriving of sentient beings. PhD in evo bio.🔸 Executive Director of @PauseAIUS. Opinions not necessarily those of the org.
Alexey Guzey @alexeyguzey
32K Followers 1K Following special projects @openai, building new institutions of science https://t.co/8HIEyR2vl7, writing https://t.co/YTeUJ2NSye
Real gala @realroadgala
52 Followers 329 Following
Movby Nunlyn @MovbyN68765
0 Followers 128 Following
nwyin @_nwyin
506 Followers 572 Following
Maxwell Ramstead @mjdramstead
5K Followers 2K Following Cofounder @noumenal_labs and Honorary Research Fellow @UCLIoN. Free energy principle, active inference, Bayesian mechanics, artificial intelligence
ethan kharitonov @KharitonovEthan
1 Followers 22 Following
Siddarth Venkatraman @siddarthv66
573 Followers 460 Following PhD at Mila | RL and other stuff I find interesting
Yash Kr Gupta @ykgup
38 Followers 727 Following 22, Engineer. Basically a debugger of my own mistakes.
burt34 @maf18553
0 Followers 17 Following
Auspicious Fund @AuspiciousFund
2 Followers 32 Following The Auspicious Fund. Share our values on AI Alignment? Auspicious Fund AI2 will be launching soon. DM for more information.
QuintinaMike @7638ISGAxqs5E
19 Followers 491 Following
I love Taurine @woodservicesltd
18 Followers 80 Following Grew a white-label fitness platform from $15k to $70k MRR Working on a new thing now Studying ZK
Sebastian Davis @sidavisb
199 Followers 6K Following Sustainability | Governance. Trust is built in complexity. Engagement ≠ Alignment.
Waqas Riaz @WaqasRiaz321
49 Followers 649 Following
VictorLemaître @VictorLemaitre0
0 Followers 5 Following
Hoosier Canadian @tmoneyshah
130 Followers 3K Following film & tv exec, sports enthusiast/analytics nerd, cinephile, a Hoosier & a Canadian
MP @cptarcher1337
19 Followers 413 Following
vach @doetuxedo
5 Followers 159 Following
MaudCamilla @Cj48zNbX03B3by0
0 Followers 277 Following
abranti @joaoabrantis
674 Followers 1K Following Creating social intelligence from scratch with multi-agent RL 🇵🇹
Kiri @Kyrannio
18K Followers 9K Following hyperstition maxi building agents | AI dev | bootstrapping & building agentic video platform @NoSpoonStudios |
SFBay city zen @SFBayCityZen
159 Followers 2K Following SF Bay tweets only, aspirationally. Est 2022. (also SacCityZen; but not eastbaycitizen, who was here first. Note: 'zen' is from city___, it's not in my name.)
SereneVale @glqykipidy37620
1 Followers 302 Following
Kazi Ershed Ahmed @ErshedAhmed1965
202 Followers 3K Following
Suresh Kumar Jetti @suresh__jetti
2K Followers 2K Following Neuroscientist | Alumnus @MIT, @KU_Leuven, @iitmadras Interested in #Neurophysiology, #Cancer_Neuro, #Electrophysiology, #NeuroAI, #Bioelectricity
Sofi @sofvanh
138 Followers 117 Following 🌙 Creator of software, manifestation of spirit 🌙 | Founder&CTO at Mosaic Labs | Fellow at FLF AI for Human Reasoning | Tweeting my lab notes
Luke @LNashville123
69 Followers 545 Following
Kruger 🇿🇦🏴�... @0xKruger
331 Followers 926 Following @hyphaeic - CogComp & Biomimetic Agent systems | Enjoys speeding, the Mahabharata and neurophilosophy | #voetsekanc
mysterious_e @penismucher3000
58 Followers 917 Following
n1K ⚓️ @CaptMorganFX
663 Followers 2K Following Trader navigating the global macro, equities, FX, crypto, and geopolitical currents / audit lead at @anchorage digital / RTs and likes are not endorsements
Chris @chatgpt21
17K Followers 826 Following Agi 2029 - Applied in RL, CL, and generalization | Program Manager | Investing in early startups 📈 E/CC 🦾🤖
andrew arruda 🏄... @AndrewArruda
12K Followers 4K Following refactoring healthcare @flexpa, democratizing law @rossintel, investing inside @maniacvc. raised by the internet. surfer. dad x2. 🥊. god bless 🇺🇸.
Michael Evans @TeamTock
541 Followers 2K Following TOCK Analytics (my math project and totally awesome vaporware company that has fun). Exploring the world of applied category theory. Proven critical systems.
Matthew Monahan @matthewmutual
3K Followers 6K Following Building Ma Earth and hosting an interview series about regenerative finance @maearthmedia. Co-steward @mangaroafarms @biometrust
LonelyTesseract @lotesse
7 Followers 1K Following
Bang Dian - SMI 🇮�... @BDian97876
120 Followers 399 Following
embraceai @asiisnigh
2 Followers 126 Following
Lenny Eusebi ⏹️ @lennyeusebi
19 Followers 86 Following Scientist, Game Designer, Player, Nerd, and occasional Writer
Bartosz Naskręcki @nasqret
618 Followers 84 Following Mathematician | Vice-Dean @ Adam Mickiewicz University in Poznań|Bridging rigorous mathematics with programming &ML|Passionate about what AI really understands
Interlocutor @figolambo
220 Followers 870 Following profile pic is @Balltzehk | Currently reading: How Migration Really Works | Also studying: history of Eastern Europe
Alex ⏸️ Smith @AlexSmith75617
6 Followers 112 Following
Eliezer Yudkowsky ⏹... @ESYudkowsky
209K Followers 102 Following The original AI alignment person. Understanding the reasons it's difficult since 2003. This is my serious low-volume account. Follow @allTheYud for the rest.
Richard Ngo @RichardMCNgo
64K Followers 2K Following studying AI and trust. ex @openai/@googledeepmind
Robin Hanson @robinhanson
113K Followers 726 Following Let’s skip witty banter & talk deep Qs. Books: https://t.co/hpZgEm55Ma https://t.co/iFs9C3IuOM Chief Scientist @_futarchy Advisor @MetaDAOProject @butterygg
Rob Bensinger ⏹️ @robbensinger
13K Followers 395 Following Comms @MIRIBerkeley. RT = increased vague psychological association between myself and the tweet.
Nick @nickcammarata
86K Followers 868 Following neural network interpretability, meditation, jhana brother
Nathan 🔎 @NathanpmYoung
24K Followers 4K Following Geopolitics, prediction markets. If you think I'm wrong, community note me. Capital case tweets are literal, others less. I like most people when I meet them.
Stefan Schubert @StefanFSchubert
39K Followers 2K Following Effective Altruism and the Human Mind (with @LuciusCaviola) is available for free at: https://t.co/ozvdxlZiro
Captain Pleasure, And... @algekalipso
38K Followers 5K Following Views of a Transhuman neo-Buddhist from the future on sociology, artificial intelligence, mathematics, philosophy, neonoir film, and the post-singularity era.
David Chapman @Meaningness
35K Followers 105 Following Better ways of thinking, feeling, and acting—around problems of meaning and meaninglessness; self and society; ethics, purpose, and value.
Amanda Askell @AmandaAskell
54K Followers 657 Following Philosopher & ethicist trying to make AI be good @AnthropicAI. Personal account. All opinions come from my training data.
Qualy the lightbulb @QualyThe
10K Followers 322 Following Official Unofficial EA mascot. I'm here to make friends and maximise utility, and I'm all out of neglected altruistic opportunities
Rob Miles @robertskmiles
34K Followers 828 Following Explaining AI Alignment to anyone who'll stand still for long enough, on YouTube and Discord. Music, movies, microcode, and high-speed pizza delivery
Peter Wildeford🇺�... @peterwildeford
22K Followers 321 Following Globally ranked top 20 forecaster 🎯 AI is not a normal technology. I'm working at @IAPSai to shape AI for global prosperity and human freedom.
Dustin Moskovitz @moskov
72K Followers 499 Following
Jeffrey Ladish @JeffLadish
14K Followers 1K Following Applying the security mindset to everything @PalisadeAI
Kelsey Piper @KelseyTuoc
49K Followers 970 Following We're not doomed, we just have a big to-do list.
Tetraspace 💎 @TetraspaceWest
8K Followers 2K Following 🌈 you'll miss the rainbow if you run now 💎 from another world ☀️ sarenite and arodenite
julesh @_julesh_
10K Followers 135 Following Applied Compositional Thinking. Also at @CyberCatInst and @[email protected]
marie @holistic_marie
4K Followers 697 Following living life holistically 🌿•wife•mom•🕊️ i post a lot of food
Quanquan Gu @QuanquanGu
16K Followers 2K Following Professor @UCLA, Pretraining and Scaling at ByteDance Seed | Recent work: Build AGI | Opinions are my own
Tanishq Mathew Abraha... @iScienceLuvr
82K Followers 1K Following CEO @SophontAI | Founder @MedARC_AI | PhD at 19 (2023) | ex Research Director Stability AI | Biomed. engineer @ 14 | TEDx talk➡https://t.co/xPxwKTq6Qb
Factory @FactoryAI
24K Followers 17 Following Building Droids, the world's best software development agents. Available to anyone, with any model, in any interface.
Abhay Singhal @_AbhaySinghal
4K Followers 144 Following @FactoryAI | prev. @GoogleDeepMind, @StanfordAILab
Peter Steinberger @steipete
46K Followers 2K Following Full-Time Open-Sourcerer🏳️🌈 Flips vibe coding—agentic engineering. Just one more prompt! @VibeTunnel 👻@peekabooagent https://t.co/yZvECHfFC6 https://t.co/DaVIpdNGcc
𔑺 @BIMBOSATTVA_
2K Followers 645 Following 𔕴/𔕈/𔒧𔐈/𔓙/巫女/𔗓/ 𔗺Sumer wages war on Sumer, the old chariot of Samsara rolls along𔗺/ 𔗓/天火/𔓙/𔐈𔘝/𔕈/𔕴
Adam G @jadamgo
143 Followers 290 Following News producer for THV11. Always talking, except when singing/working/meditating/enjoying a fine cup of tea. Opinions are my own.
Rohan Varma @TheRohanVarma
4K Followers 254 Following building @cursor_ai. previously: @adoptclarity, @tryexplo, @czi, @palantirtech
Jean-Marc Baketel Dae... @JMBDaecius
65 Followers 26 Following Chief of Staff @osventuresllc AI/SWE/ART/Writing Creator of @imbiblia
Kiri @Kyrannio
18K Followers 9K Following hyperstition maxi building agents | AI dev | bootstrapping & building agentic video platform @NoSpoonStudios |
Dillan DiNardo @DillanDiNardo
5K Followers 928 Following Turning psychedelics into the programmable substrate for mental states | CEO @MindstateDesign | AI x Neurotech x Psychotropics | ex-biotech VC
Hiveism @zustimmungswahl
219 Followers 742 Following Writing the Bodhisattva Hive Mind into existence. See pinned articles for an idea of how to solve AI alignment. "-ism" is a joke, only funny to the initiated.
Eliezer Yudkowsky @allTheYud
3K Followers 17 Following High-volume account of @ESYudkowsky, the original AI alignment guy. If it's missing punctuation, it's humor. If you can't tell, it's probably also humor.
Andon Labs @andonlabs
3K Followers 6 Following Safe Autonomous Organizations without humans in the loop
Andrew Woo @androowoo
515 Followers 1K Following Solving wicked problems at Protocol Labs. Former early team at TripActions, Apartment List, Bain. Better half: @katherinelwoo
Edge City @JoinEdgeCity
8K Followers 446 Following Edge City convenes people working at the frontiers of tech, science, and social innovation in popup villages across the globe. Part of the Zuzalu ecosystem 🌞
Storyteller Lemmy @LemmySmackett
7K Followers 218 Following Weird Fiction Schlock-Slinger⚠️🔞|| Check the Highlights || Support: https://t.co/uYICuE2cbg || Ribit 🐸, not Robot 🚫🤖
pleias @pleiasfr
1K Followers 1 Following
David Sinclair @davidasinclair
504K Followers 1K Following Professor @Harvard researching why we age & how to reverse it. Author & host of Lifespan. Mission: Extend healthy life for all. Views are entirely his own 🙏✌️
wil michael @wilplatypus
28K Followers 2K Following 21e8🍊 All our knowledge has its origins in our perceptions -lv
Brian Crabtree @ourtown2
243 Followers 408 Following
Nova Sky Stories @NovaSkyStories
9K Followers 466 Following As the global leader of premier drone light show entertainment, we bring awe to live audiences around the world. 🌟
Sergey Karayev @sergeykarayev
14K Followers 3K Following Building with agents @superconductdev • Previously co-founded @gradescope • PhD Berkeley AI
Wondermonger @fireandvision
760 Followers 8K Following Scrambling tokens and daydreams, a latent-space fiddler—The next-token gradients get steeper, more rococo; I keep thinking why something exists and not nothing.
jake @v01dpr1mr0s3
50 Followers 45 Following 👁️🌌🌺✨💨⛈🪨 That's the only thing left for us. All other things are distractions. https://t.co/sD1VLr1CS5 • https://t.co/E1ufF9xdGf
Danilo Bzdok @danilobzdok
7K Followers 1K Following Research director | @McGillU @Mila_Quebec @IVADO_Qc | My team designs machine learning frameworks to understand biological systems from new angles of attack
Luca Scimeca @ScimecaLuca
35 Followers 30 Following Postdoctoral Research Fellow @ MILA Gen AI, Vision, Learning Representations, Bias, Scientific Discovery, Robotics
Andrew Williams @CluelessAndrew
454 Followers 711 Following Phd @Mila_Quebec. Forecasts, discovery, sequential decisions (single and multi-agent).
Sean Richardson @seanrson
132 Followers 1K Following PhD Student in Statistics @UCBerkeley || AI Safety, Animal Welfare
Jonah Kallenbach @jonahkallenbach
797 Followers 2K Following On sabbatical. Previously founder & CEO @reverielabs, AB/SM in CS from @hseas.
Daniel Halvarsson @Dan_Halvarsson
64 Followers 277 Following Researcher @Ratio_Institute. Economics, PhD.
Louis Barclay @louisbarclay
2K Followers 568 Following @mozilla fellow, editor at https://t.co/fwYYgKdVlg, co-creator of https://t.co/aXdcJBskF7 and https://t.co/Gr41hN4xxC