Communications of the ACM
Chess Engine Sacrifices Mastery to Mimic Human Play
A new artificial intelligence chess engine was trained to play like humans, rather than out-playing them to win.
Credit: Aleksey Arkhipov/Shutterstock
A team of researchers from Cornell University, Canada's University of Toronto, and Microsoft Research has developed an artificial intelligence chess engine that is trained to play like, rather than beat, humans.
The Maia chess engine was taught to mimic human behavior through training on individual human chess moves, instead of the larger problem of winning the game.
The researchers found Maia matched human moves within each skill level over 50% of the time, an accuracy rate higher than those of the popular chess engines Stockfish and Leela.
Cornell's Jon Kleinberg said, "Our model didn't train itself on the best move; it trained itself on what a human would do. But we had to be very careful—you have to make sure it doesn't search the tree of possible moves too thoroughly, because that would make it too good. It has to just be laser-focused on predicting what a person would do next."
From Cornell Chronicle
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found