Chess Move
A chess move, seemingly simple, is a complex decision influenced by factors ranging from immediate tactical advantage to long-term strategic goals. Current research focuses on building AI systems that not only play chess at superhuman levels but also model human decision-making processes across different skill levels, employing architectures like transformers and Monte Carlo Tree Search (MCTS) alongside novel approaches like Mixture of Experts. This research contributes to a deeper understanding of human cognition, decision-making under uncertainty, and the development of more human-aligned AI, with potential applications in other complex decision-making domains beyond games.
Papers
August 2, 2022
December 22, 2021
November 17, 2021