Center for Human-Compatible AI

Technical Research

UC Berkeley research center developing the conceptual and technical foundations for provably beneficial AI systems that are uncertain about human preferences and defer to them.