Although today’s result wasn’t as “close” as the first match, where AlphaGo won by just half a point, Deepmind CEO Demis Hassabis said that Ke Jie played “perfectly” for much of the contest before he resigned, at least according to AlphaGo’s evaluations. “For the first 100 moves it was the closest we’ve ever seen anyone play against the Master version of AlphaGo,” Hassabis confirmed in the post-game press conference.
“Today’s game was different from the first,” Ke said, reported by The Verge. “AlphaGo made some moves which were opposite from my vision of how to maximize the possibility of winning. I also thought I was very close to winning the game in the middle but maybe that’s not what AlphaGo was thinking. I’m a little bit sad, it’s a bit of a regret because I think I played pretty well.”
#AlphaGo wins game 2. What an amazing and complex game! Ke Jie pushed AlphaGo right to the limit.
— Demis Hassabis (@demishassabis) May 25, 2017
AlphaGo is in China visting the Future of Go Summit, a five-day forum hosted by Google and the China Go Assocation. It’s brought together some of the world’s best Go players and AI experts to “explore the mysteries” of the ancient board game. The banner event is the match-up between AlphaGo and Ke, who will meet for the final time on Saturday, but the AI player is due to play two back-to-back matches on Friday. This will include a game of Pair Go, where Chinese pros face off against each other but alternate moves with AlphaGo, and a Team Go match, where the AI will battle a five-player team to test its “creativity and adaptability.”
This latest version of AlphaGo, which has been given the moniker Master, is said to use 10 times less computational power than the computer that beat Lee Sedol. All it needs is a single PC connected to Google’s cloud server.