Trillium Labs launches open AI research nonprofit with $30M training budget
Nathan Lambert and Tom Zick, former researchers at Ai2 and Harvard, founded Trillium Labs, a nonprofit focused on transparent AI research. The lab will publish detailed experiments on topics like recursive self-improvement (RSI) and reinforcement learning, allowing outside scrutiny and replication. Lambert argues closed development by labs like OpenAI and Anthropic limits community oversight,…
Key points
- Trillium Labs publishes experiment details to enable replication and scrutiny of AI research
- Focus areas include recursive self-improvement (RSI) and reinforcement learning for model behavior
- Nonprofit raised undisclosed funds, targeting $40–$100M total with $30M earmarked for training
Trillium Labs launched today with undisclosed funding from Schmidt Sciences and Halcyon Futures, aiming to raise $40–$100 million total. It plans to spend $30 million on training over the next 18 months. The initiative follows recent debates over AI risks, including an Anthropic researcher’s warning about RSI’s existential threat. The lab’s open approach contrasts with China’s model-sharing practices and Stanford’s Marin project, reflecting a growing industry divide over transparency versus controlled access.
The story so far
9 episodes →- Trillium Labs launches open AI research nonprofit with $30M training budgetthis story
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- Researcher tests GPT-6 Astra in WoW with agent-wow framework · 1 src
- GitHub releases KernelOPT for agentic GPU kernel optimization · 1 src
- Karpathy shares how to make AI outputs like aircraft manuals · 1 src
- Allen Institute for AI open-sources AstaBrief 8B for faster scientific reports · 1 src
- Datalab releases OmniExtractBench to audit PDF extraction accuracy · 1 src
Comments
via GitHub Discussions