Learning to Cooperate and Compete via Self Play

Explore the intricacies of multi-agent artificial intelligence in this lecture from the 2023 Cooperative AI Summer School. Delve into the world of imperfect-information games as Noam Brown, a renowned researcher at OpenAI, shares insights on learning to cooperate and compete through self-play. Discover the groundbreaking work behind Libratus and Pluribus, the first AI systems to defeat top human players in two-player and multiplayer no-limit poker. Gain valuable knowledge from Brown's expertise, which has earned him accolades such as the Marvin Minsky Medal for Outstanding Achievements in AI and recognition as one of MIT Tech Review's 35 Innovators Under 35. Uncover the scientific breakthroughs that led to Pluribus being named one of the top 10 scientific achievements by Science Magazine. Learn from Brown's distinguished career, including his time at Facebook AI Research and his award-winning PhD work at Carnegie Mellon University, as he presents cutting-edge concepts in cooperative and competitive AI strategies.