Safe AI through Consensus Sampling

​Adam Tauman Kalai is a Research Scientist at OpenAI specializing in AI Safety and Ethics, with research interests spanning algorithms, fairness, AI theory, game theory, and crowdsourcing. He earned his BA from Harvard University and his PhD from Carnegie Mellon University, and has held positions as an Assistant Professor at TTIC and Georgia Tech and Senior Principal Researcher at Microsoft Research New England. He is also a member of Project CETI's science team. His work has been recognized by test of time and best paper awards, an NSF CAREER Award, an Alfred P. Sloan Fellowship, and the Majulook Prize.

​Adam will present his paper, Consensus Sampling for Safer Generative AI, which proposes using consensus across multiple AI models to provide provable safety guarantees. He will discuss how this architecture-agnostic approach addresses risks that are fundamentally undetectable by inspecting model outputs alone, and what it means for building AI systems with mathematically grounded safety properties.

​​Free food will be provided!

We will be accommodating virtual attendees. You can join the meeting here!