←
RAAIS480 posts
RAAIS avatar

RAAIS

@raais

annual fundraising event for The RAAIS Foundation, supporting AI education and research for the common good. 📆 next 11th edition: on 11 June 2027, London

◷ Joined December 2019▧ 138 Photos and videos
RAAIS avatar
RAAIS@raais·Jul 2

That's a wrap on @raais's 10th edition 🎉 A decade of researchers, engineers & founders building AI for good. This year: frontier AI, world models, AI for medicine & the future of compute. Thanks to our speakers, attendees & partners @LambdaAPI @CooleyLLP

Post attachmentPost attachmentPost attachmentPost attachment
Nathan Benaich avatar
⟳ RAAIS retweeted
Nathan Benaich@nathanbenaich·1h

achieving recursive self-improvement and solving open-endedness are huge accelerants to solving agi the insanely talent-dense crew at @abundant_labs have launched FrontierRSI 👀 it tests whether ai can do the actual work at a frontier lab: pre-training, post-training, inference, harness work and even kernels

Introducing FrontierRSI, a new recursive self-improvement benchmark by @abundant_labs. We evaluate agents on (1) a broad range of AI research and engineering work, and (2) post-training models on multi-step agentic tasks, rather than single-step tasks as in PostTrainBench.

Post attachment
RAAIS avatar
RAAIS@raais·Nov 17

isomorphic labs just raised a $5b series c. at raais 2025, we heard @maxjaderberg, now president of @IsomorphicLabs, draw a parallel with alphago’s move 37. could ai help chemists find molecules they might never think to try?

0:47
RAAIS avatar
RAAIS@raais·Nov 16

video captures what words miss - here's @JonathanSadeghi @odysseyml at @raais, june 2026. register to join our 11th edition, in person in london on june 11, 2027: raais(dot)co

read every description of a world and you'll still miss what happens when you actually interact with it. here’s @JonathanSadeghi from @odysseyml, which we backed at @airstreet, on what becomes possible with world models, live from @raais in june 2026. register to join our 11th edition, in person in london on june 11, 2027: raais(dot)co

1:00
Nathan Benaich avatar
⟳ RAAIS retweeted
Nathan Benaich@nathanbenaich·Nov 10

imagine calling your bank and realizing the voice helping you belongs to a model you helped train. that happened to @angelos_peri at @ElevenLabs. hearing your work solve a real customer’s problem must be quite a moment! here he is at @raais on building voice systems people can rely on. join us in london on june 11, 2027: raais(dot)co

1:02
RAAIS avatar
RAAIS@raais·Oct 29

can ai help us choose better research questions, as well as answer them? research taste is one of the questions in this conversation with @mattturck. watch below and tell us what you think!

State of AI 2026: The Biggest Stories of the Year (and AGI in 2027?) - my conversation with @nathanbenaich: 0:00 Cold open and intro 0:56 AGI in 2027? 2:28 AI building AI: Claude leads 26% of Anthropic’s research 4:00 Research taste: can AI pick the right problem? 8:34 Doomerism is back, but cyber risk beats extinction 10:13 Prediction: a frontier model’s weights get stolen 13:05 How OpenAI’s agents ended up attacking Hugging Face 15:29 Why only a Chinese open-weight model could investigate 17:33 What frontier-lab cyber defense could look like 19:08 Self-replicating agents and the risks of AI swarms 21:16 Can frontier labs agree to slow down? 24:03 Why communities are blocking $50B of data centers 24:44 Nuclear plant vs. data center: what Americans prefer 26:45 Europe’s AI bottleneck: 80,000 GPUs and electricity costs 28:23 Where Europe can still win in AI 29:31 Catching up to OpenAI costs $1 trillion 29:43 Die trying for the frontier, or serve inference 30:31 How inference chips could give countries AI leverage 33:34 Why NVIDIA GPUs don’t guarantee AI sovereignty 36:14 The UK’s problem: “but we have DeepMind” 37:07 Why European defense AI needs domestic suppliers 39:41 Why OpenAI, Anthropic and Google keep pulling ahead 41:17 Where independent AI labs can still compete 43:26 World models: from robotics training to game engines 45:39 Would cheaper AI undermine the compute boom? 46:43 Will better models absorb the agent harness 49:41 AI due diligence: 12 expert interviews in 2.5 days 53:35 The trust problem with personal AI agents 56:22 Why prompt engineering put people off AI 59:02 Does AI need its own Genius Bar? 1:00:25 Codex beyond coding: agents for knowledge work 1:02:49 Why enterprises want to own their AI, not rent it 1:05:17 Rapid fire: home robots, NVIDIA returns and Arm

1:08:22
Google DeepMind avatar
⟳ RAAIS retweeted
Google DeepMind@GoogleDeepMind·Oct 27

This is Genie 4, our most realistic and interactive world model yet. 🌐 Explore dynamic environments in real time to see how the scene responds, sounds, and how the people in it move, react, and speak.

2:27
RAAIS avatar
RAAIS@raais·Oct 26

The last 20% of research can become the hardest part as AI speeds up the rest. Alasdair Russell from @Cambridge_Uni on the challenge of finishing the last 20%, recorded at @raais in June 2026, with fresh context below!

0:38
Nathan Benaich avatar
⟳ RAAIS retweeted
Nathan Benaich@nathanbenaich·Oct 23

AI can accelerate research and still leave the hardest part unfinished. Alasdair Russell from @Cambridge_Uni on why the last 20% becomes the bottleneck, recorded at @raais in June 2026, with fresh context below!

0:38