Log inSign up
Sholto Douglas
1,082 posts
user avatar
Sholto Douglas
@_sholtodouglas
Scaling RL @AnthropicAI, ex @DeepMind - working towards intelligence too cheap to meter
Joined February 2018
1,828
Following
47K
Followers

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up
  • Pinned
    user avatar
    Sholto Douglas
    @_sholtodouglas
    Feb 28, 2025
    Very excited to announce that I joined Anthropic at the start of the month! Everything I'm seeing says that we are on trend for AGI in 2027. If the trend lines continue - then we have a shot at building a radically better world, but the path there is fraught with very real
    501K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Oct 3, 2025
    Should we have ‘keep thinking’ merch?
    user avatar
    sam mcallister
    @sammcallister
    Oct 3, 2025
    GOOD MORNING NEW YORK CITY COME DO YOUR BEST THINKING AT OUR THINKING SPACE IN THE WEST VILLAGE SAY NO TO SLOP
    193K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 17, 2025
    We're sorry - and we'll do better. We're working hard on making sure we never miss these kind of regressions and rebuilding our trust with you.
    user avatar
    Claude
    Anthropic
    @claudeai
    Sep 17, 2025
    We've published a detailed postmortem on three infrastructure bugs that affected Claude between August and early September. In the post, we explain what happened, why it took time to fix, and what we’re changing:
    531K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Dec 6, 2023
    Replying to @0interestrates and @nearcyan
    Name order was randomised (except for the first 6 names which spell out Gemini) - Sergey was in with us basically every day, often pairing!
    731K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 29, 2025
    Claude 4.5 is the best coding model in the world - and the qualitative difference is quite eerie. I now trust it to run for much longer and to push back intelligently. As ever - everything about how its trained could be improved dramatically. There is so much room to go. It is
    user avatar
    Mike Krieger
    @mikeyk
    Sep 29, 2025
    We asked every version of Claude to make a clone of Claude(dot)ai, including today’s Sonnet 4.5… see what happened in the video
    00:00
    161K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 22, 2025
    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
    user avatar
    Lisan al Gaib
    @scaling01
    Sep 22, 2025
    Tri Dao says Claude Code makes him 1.5x more productive and that it's quite helpful at writing Triton kernels
    254K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 21, 2025
    Or “hey why haven’t we solved continual learning yet”
    110K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 25, 2025
    Incredible work - this should immediately become one of the most important metrics for policy makers to track. We’re probably only a few months from crossing the parity line. Huge props to OAI for both doing the hard work of pulling this together and including our scores.
    user avatar
    Tejal Patwardhan
    @tejalpatwardhan
    Sep 25, 2025
    Understanding the capabilities of AI models is important to me. To forecast how AI models might affect labor, we need methods to measure their real-world work abilities. That’s why we created GDPval.
    163K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 20, 2025
    People often talk about the future of software engineering as looking like starcraft (managing swarms of agents). E.g. I just found out one friend uses 9 claude codes at once! I think this would be really fun as an actual PvP game - but I'm not sure what the right
    110K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 18, 2025
    As is tradition
    user avatar
    Nat McAleese
    @__nmca__
    Sep 18, 2025
    everything reminds me of him 😭
    67K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Jul 6, 2025
    “Where did you meet the bootleg peptide dealer?” “At the underground robot boxing ring” Visceral future-shock
    92K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Jul 29, 2025
    smh my ranking is below my best ever fencing world ranking (43rd), I've got some work to do
    user avatar
    TBPN
    @tbpn
    Jul 29, 2025
    Introducing The Metis List: a live leaderboard of 100 top AI researchers, ranked by their peers. We include: - Education - Citation count - Current company - # of Dwarkesh appearances - Notable work Check it out at MetisList dot com
    95K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Mar 25, 2024
    High context people should write more 'requests for research / project' to give talented but unknown people threads to pull on and break into their worlds. Loosely "If you did a world class blog post on X, I know many people who'd be excited to meet you".
    141K
  • user avatar
    Sholto Douglas
    @_sholtodouglas
    Sep 18, 2025
    Replying to @ainergiz
    Next version insanely better is the plan
    57K