11th edition of AI Safety Camp

Cost-efficiently support new careers and new organisations in AI Safety.

Lightcone Infrastructure

Funding for LessWrong.com, the AI Alignment Forum, Lighthaven and other Lightcone Projects

Act I: Exploring emergent behavior from multi-AI, multi-human interaction

Community exploring and predicting potential risks and opportunities arising from a future that involves many independently controlled AI systems

MATS Program

Help us support more research scholars!

10th edition of AI Safety Camp

Finishing The SB-1047 Documentary

How California became ground zero in the global debate over who gets to shape humanity's most powerful technology

PIBBSS - General Programs funding or specific funding

Fund unique approaches to research, field diversification, and scouting of novel ideas by experienced researchers supported by PIBBSS research team

80,000 Hours

We provide research and support to help people move into careers that effectively tackle the world’s most pressing problems.

Apollo Research: Scale up interpretability & behavioral model evals research

Hire 3 additional AI safety research engineers / scientists

Next Steps in Developmental Interpretability

Addressing Immediate AI Safety Concerns through DevInterp

Research Staff for AI Safety Research Projects

Exploring feature interactions in transformer LLMs through sparse autoencoders

Help Apart Expand Global AI Safety Research

Incubate AI safety research and develop the next generation of global AI safety talent via research sprints and research fellowships

Scoping Developmental Interpretability

6-month funding for a team of researchers to assess a novel AI alignment research agenda that studies how structure forms in neural networks

Joseph Bloom - Independent AI Safety Research

Trajectory Models and Agent Simulators

Discovering latent goals (mechanistic interpretability PhD salary)

6-month salary for interpretability research focusing on probing for goals and "agency" inside large language models

Avoiding Incentives for Performative Prediction in AI

Deleted

Deleted

Building and maintaining the Alignment Ecosystem

12 months funding for 3 people to work full-time on projects supporting AI safety efforts

Garrett Baker salary to study the development of values of RL agents over time