Nathan Lambert launches Trillium Labs, a nonprofit for open frontier AI science
himanshustwts · x · 2026-10-03
Nathan Lambert announced Trillium Labs, a new nonprofit fostering open science of frontier AI. It will build open post-training recipes and expand into open infra to study RSI, reward-hacking and multi-agent systems.
The org's theory of change: hard technical problems need more eyes, and it puts its faith in the scientific method. The post itself is a congratulatory reply; the substance sits in the quoted announcement.
Related event: Nathan Lambert Launches Nonprofit Trillium Labs for Open Frontier AI(14 posts)→
More from Companies & People
- At The Curve conference: reverse federalism AI policy debate and RSI evaluation talks — neil_chilson · 2026-10-03
- Free AI Curriculum Co-Developed with Google DeepMind Pilots in Australia — lmoroney · 2026-10-03
- Anthropic 'models are conscious' hype debunked: article opens with Olah saying 'we don't know' — _sholtodouglas · 2026-10-03
- Why isn't Thinking Machines counted among the new AI labs? Talent density called 'crazy' — dejavucoder · 2026-10-03
- Allen AI heads to COLM 2026 with Olmo-core 3 and AstaBrief releases — allen_ai · 2026-10-03
- Every business will have its own agent the way it has a website, argues Every — every · 2026-10-03