Go is a game in which there are more possible moves than a computer running conventional programming approaches can reasonably process. Such a barrier has inspired AI researchers from Google's DeepMind to develop AlphaGo, an AI-based program that can teach itself how to play a game that requires some level of instinct to succeed.
Last year, AlphaGo managed to defeat one of the world’s top Go players, Lee Se-dol, by defeating him in four out of five games, and winning the match. But today, the AI went further to defeat the current world number one Go player, Ke Jie, at DeepMind's "Future of Go Summit" in Wuzhen, China. After the match, Ke Jie said through an interpreter:
“I was deeply impressed. There was a cut (a type of Go move) that quite shocked me, because it was a move that would never happen in a human-to-human Go match.”
This match was only one of many to feature AlphaGo as a player in the summit. Ke Jie will face it again on Thursday and Saturday, even though he said he doesn't want to subject himself to the "horrible experience" of competing against AlphaGo ever again. Also, the AI will challenge other Go players, including several games involving one or more players collaborating with it.
Sources: MIT Technology Review & Phys.org
14 Comments - Add comment