AI300
  1. 31.
    The only thing I don't get is why it has to be a Mac mini.›
    Ben Burtenshaw (Hugging Face Researcher) · LLM score 20 · 4 days ago
  2. 32.
    @2024dion @deredleritt3r The fact that it still seems 50-50 by 2028 surprises me! I would have thought that evidence would have tipped it one way or another by now.
    Shane Legg (Chief AGI Scientist at DeepMind) · LLM score 30 · 5 days ago
  3. 33.
    @ariccio Use it for when the main agent doesn't need to see the work that a skill did, only its output.›
    Boris Cherny (Creator of Claude Code) · LLM score 40 · 6 days ago
  4. 34.
    Honestly, I think memory is the biggest blocker to continual learning right now.›
    Michael Elabd (DeepMind Researcher) · LLM score 85 · 6 days ago
  5. 35.
    Is it better to go to academia or industry for research in AI or cognitive science? It's the most frequent question I get asked by PhD students, and I've finally written up some of my thoughts on the answers, as an epilogue to my more research-focused posts on cog sci and AI: https://t.co/qoIqBKGqX6
    Andrew Lampinen (Research Scientist at DeepMind) · LLM score 75 · 6 days ago
  6. 36.
    Il est essentiel de développer à la fois une réglementation efficace, des garde-fous sociaux et des solutions techniques comme celles que nous développons à @LawZero_ pour éviter les risques de l’IA avancée tout en développant ses applications positives.›
    Yoshua Bengio · LLM score 40 · 6 days ago
  7. 37.
    @kevinkshah @korbencopy I'd give it a 50% probability of happening within 5 years of Minimal AGI.›
    Shane Legg (Chief AGI Scientist at DeepMind) · LLM score 70 · 6 days ago
  8. 38.
    @GDalegend @korbencopy We get to 2028 and have something quite close to Minimal AGI, but not quite.
    Shane Legg (Chief AGI Scientist at DeepMind) · LLM score 20 · 6 days ago
  9. 39.
    @tablefourthree @BogdanIonutCir2 @korbencopy It's actually surprising to me that my probability hasn't shifted yet, given that 2028 is so close.›
    Shane Legg (Chief AGI Scientist at DeepMind) · LLM score 20 · 6 days ago
  10. 40.
    @korbencopy 50% chance of Minimal AGI by 2028.›
    Shane Legg (Chief AGI Scientist at DeepMind) · LLM score 70 · 6 days ago
  11. 41.
    At Deepmind ever since we maxed out SWE-bench , I’ve felt that there hasn’t been a difficult enough benchmark that is aligned with how people want to use AI.›
    Ronak Malde (DeepMind Researcher) · LLM score 75 · 7 days ago
  12. 42.
    @_sholtodouglas @jekbradbury @GoogleDeepMind @andy_l_jones @OpenAI @kevin_wang3290 A word on comp: I know folks who become quants to make money but 5 years later ask what they're doing with their life.›
    Noam Brown (OpenAI Research Scientist) · LLM score 20 · 8 days ago
  13. 43.
    @_sholtodouglas @jekbradbury @GoogleDeepMind @andy_l_jones Labs like @OpenAI also hire researchers straight out of undergrad, like @kevin_wang3290, though the bar is high.›
    Noam Brown (OpenAI Research Scientist) · LLM score 60 · 8 days ago
  14. 44.
    @_sholtodouglas @jekbradbury @GoogleDeepMind @andy_l_jones was a semi-retired quant when he wrote a paper comparing the impact of scaling pretraining vs scaling test-time compute (before TTC scaling was cool).›
    Noam Brown (OpenAI Research Scientist) · LLM score 70 · 8 days ago
  15. 45.
    We’re excited to be working with @JNJInnovation to accelerate the path to new medicines.›
    Demis Hassabis (CEO of DeepMind) · LLM score 60 · 8 days ago
  16. 46.
    @MillionInt The engineer presented a single grain of etched silicon.›
    Igor Babuschkin (Cofounder of xAI) · LLM score 65 · 9 days ago
  17. 47.
    @supreme__sharma There's no good answer.›
    Keller Jordan (OpenAI Researcher) · LLM score 20 · 10 days ago
  18. 48.
    @eliebakouch In terms of hypothetical disagreement-winrates between these different cultures, I’d estimate that my (heavily muted) FYP has +500 ELO vs.›
    Keller Jordan (OpenAI Researcher) · LLM score 30 · 10 days ago
  19. 49.
    The internet has become real life only in the sense that *public* cultural systems have been largely replaced by it›
    Keller Jordan (OpenAI Researcher) · LLM score 20 · 11 days ago
  20. 50.
    @aref_vc The formula for the auto-compact buffer is really simple: (contextWindow - maxOutputTokens) - (small constant)›
    Boris Cherny (Creator of Claude Code) · LLM score 80 · 12 days ago
  21. 51.
    @Wind_AndSnow Ah, you want permissions: { allow: "Read(~/.config/runmenu/*)"}
    Boris Cherny (Creator of Claude Code) · LLM score 80 · 12 days ago
  22. 52.
    @aref_vc Reduce your CLAUDE_CODE_MAX_OUTPUT_TOKENS or rm the env var, and you'll decrease the buffer to 45k
    Boris Cherny (Creator of Claude Code) · LLM score 70 · 12 days ago
  23. 53.
    @DominikPeters Claude usually only uses Plan agents.›
    Boris Cherny (Creator of Claude Code) · LLM score 70 · 12 days ago
  24. 54.
    Now in Claude Code: when you accept a plan, Claude automatically clears your context, so your plan gets a fresh context window.›
    Boris Cherny (Creator of Claude Code) · LLM score 70 · 12 days ago
  25. 55.
    Amazing continual learning paper out of DeepMind 🚨›
    Michael Elabd (DeepMind Researcher) · LLM score 85 · 13 days ago
  26. 56.
    TranslateGemma is our new collection of open translation models for edge devices built on our amazing Gemma 3 open model.›
    Demis Hassabis (CEO of DeepMind) · LLM score 80 · 14 days ago
  27. 57.
    Really fun working with the @cursor_ai team to help launch Composer 2.0 on GB200s and B200s!›
    Dan Fu (VP of Kernels at Together) · LLM score 30 · 14 days ago
  28. 58.
    METR hasn’t evaled GPT 5.2 yet but I bet it would do extremely well https://t.co/8xZ4fUoAr0
    Noam Brown (OpenAI Research Scientist) · LLM score 20 · 15 days ago
  29. 59.
    One day we’ll be able to decompose the loss curve of a neural net into all of the quanta it learns along the way - this is one of my fav streams of fundamental research.›
    Sholto Douglas (Researcher at Anthropic) · LLM score 80 · 15 days ago
  30. 60.
    For AI to be truly useful, it needs to understand you.›
    Demis Hassabis (CEO of DeepMind) · LLM score 40 · 15 days ago
More