Sign In

Communications of the ACM

ACM TechNews

Would You Like to Play a Game? New AI Teaches Itself the Rules

View as: Print Mobile App Share:


Paris Diderot University researcher Lukasz Kaiser has developed an artificial intelligence (AI) program that can watch two-minute videos of simple board games being played, learn the rules, and then challenge human opponents.

The program is a set of subroutines that work in concert--a visual-analysis system provides data to a game-learning algorithm, and both are linked to the open source game engine Toss. The first tests of the AI program--on tic-tac-toe, Connect Four, Go-Moku, Pawns, and Breakthrough--were on a laptop with a single-core processor and 4GB of RAM.

"My background is in logic and in finite model theory, so I'm mostly a theoretical computer scientist, but I was looking for applications of the finite model theoretic methods, so I looked toward AI," Kaiser says. "It turned out that, indeed, it is possible to apply some of the theoretic methods to get better results in learning and to illustrate it by game-playing programs."

Although Kaiser says the program has no barrier in principle to learning how to play more advanced games, a more complex rule set would make it far more difficult.

From Network World
View Full Article

Abstracts Copyright © 2012 Information Inc., Bethesda, Maryland, USA 


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account