What Does Machine Learning Offer Game Theory - And Vice Versa

Explore the intersection of machine learning and game theory in this 36-minute talk by Alex Peysakhovich from Meta AI. Delve into how deep learning techniques can simplify complex scenarios into low-dimensional game theoretic representations, potentially advancing the development of cooperative and coordinated AI agents. Examine applications in poker AI, the Prisoner's Dilemma, and market design. Investigate the role of state space, coordination, and equilibrium allocations in bridging these two fields. Gain insights into the theoretical and computational aspects of combining machine learning with game theory, and discover how this fusion can lead to the creation of cooperative policies and high-level strategies in multi-agent reinforcement learning and bandit learning scenarios.

Introduction
Poker AI
Prisoners Dilemma
State Space
Coordination
Market Design
Equilibrium Allocations
Theory
Computation
Cooperative Policy
HighLevel Strategies