Center for Human-Compatible AI (CHAI)
Building exceptional AI for humanity — reorienting AI research towards provably beneficial systems.
Year Founded:
2016
Geography:
North America
Address:
University of California, Berkeley, CA, USA
Visit WebsiteTo develop the conceptual and technical wherewithal to reorient the general thrust of AI research towards provably beneficial systems.
AI areas they serve
Areas of Focus
AI alignment
Provably beneficial AI
Value alignment
Inverse reinforcement learning
Cooperative AI
Human-robot interaction
AI safety theory
AI governance and policy advice
AI Impact Areas
Upcoming Goals
Continue developing provably beneficial AI systems
Build political neutrality evaluations
Advance research on value alignment, inverse reward design, and cooperative AI
Train next generation of AI safety researchers
2023: Listed "100 Most Influential People in AI"
Russell received ACM AAAI Allen Newell Award
2019: Stuart Russell published "Human Compatible" explaining CHAI's mission to broad audience
Received $17M+ in funding from Open Philanthropy
Recognized by Founders Pledge as one of four recommended organizations for AI safety
2016: Founded by Stuart Russell with initial $5
5M grant from Open Philanthropy/Good Ventures
Faculty spans UC Berkeley, Cornell, University of Michigan, and Princeton