Event Website
https://docs.google.com/presentation/d/1DdH4JLnBdtTMPtMoFGhDSlZYuSqPn4fUionWpxD5CYw/edit#slide=id.p
Start Date
12-12-2018 10:50 AM
Description
Two years ago, an artificially intelligent program defeated the world’s top player at Go, a complex board game that humans have been playing for millennia. The match opened up possibilities for professional Go players to hone their skills by emulating the program’s tactics. This breakthrough in artificial intelligence begs a more general question: is it possible for programs to acquire tactics innovative enough to teach humans new strategies in classic games such as a chess, checkers, and Go? To answer this question, this work proposes an analysis of the interactions between human players and artificial agents, and specifically whether or not the prior can learn new strategies from the latter. This work then proposes an experiment to test this observational learning between humans and machines through the game of chess.
Click below to download individual papers.
Old Dogs, New Tricks: Can Artificially Intelligent Programs Teach Humans New Ways to Approach Classic Games?
Two years ago, an artificially intelligent program defeated the world’s top player at Go, a complex board game that humans have been playing for millennia. The match opened up possibilities for professional Go players to hone their skills by emulating the program’s tactics. This breakthrough in artificial intelligence begs a more general question: is it possible for programs to acquire tactics innovative enough to teach humans new strategies in classic games such as a chess, checkers, and Go? To answer this question, this work proposes an analysis of the interactions between human players and artificial agents, and specifically whether or not the prior can learn new strategies from the latter. This work then proposes an experiment to test this observational learning between humans and machines through the game of chess.
https://digitalcommons.lmu.edu/honors-research-and-exhibition/2018/section-01/13
