The Frontier Research Club is where frontier AI work gets pressure-tested. Fourteen sessions in, the room draws researchers from DeepMind, Meta, OpenAI, and Anthropic, faculty and PhDs from Stanford and Berkeley, and the engineers running these systems in production — convened around concrete work: papers, methods, results.
The bias is toward the questions that decide whether a result is real: What are the assumptions? How was it evaluated? Where does it fail? What would convincing evidence actually look like?
Short talks, long arguments. Each session features two to three presentations…

Hosted by Kristopher Floyd