Nathan Lambert Launches Trillium Labs, a Nonprofit for Open Research on RSI and Reward Hacking
natolambert · x · 2026-10-03
AI researchers Nathan Lambert and Tom Zick have unveiled Trillium Labs, a nonprofit devoted to open science of frontier AI, covered by Wired. It will publish open post-training recipes and expand into open infrastructure to study recursive self-improvement (RSI), reward hacking, and multi-agent systems.
Their theory of change: hard technical problems need more eyes. Lambert argues frontier labs' secrecy undermines scrutiny and that transparent, reproducible experiments are key to mitigating risk — calling the current closed trajectory "a step backwards." The lab is also recruiting.
More from AGI Musings
- OpenAI and Notre Dame Law School to co-host AI-for-humanity dialogue on Dec 4 — wjscheirer · 2026-11-19
- Anthropic Employees Debates Slowing AI as Historic IPO Nears, NYT Reports — JeffLadish · 2026-11-19
- NYT: Anthropic Shows Organizational Fractures Ahead of Historic IPO — kylekabasares · 2026-11-19
- The biggest advantage of a top lab isn't money — it's saving you years — madeofAjala · 2026-11-19
- Ex-Anthropic staffer casts superintelligence as Sauron's ring; Ladish pushes back — JeffLadish · 2026-11-19
- Proposal: A nonprofit to give visa safety to AI lab whistleblowers — AaronBergman18 · 2026-11-19