Skip to main content

AlphaGo beats Lee Se-dol again to take Google DeepMind Challenge series

AlphaGo beats Lee Se-dol again to take Google DeepMind Challenge series


DeepMind AI goes 3-0 up to seal historic victory

Share this story

If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement.

Virtuoso Go-playing AI AlphaGo has secured victory against 18-time world champion Lee Se-dol by winning the third straight game of a five-game match in Seoul. AlphaGo is now 3-0 up in the series, but there's no mercy rule here — the remaining games on Sunday and Tuesday will still be played out. AlphaGo is a program developed by DeepMind, a British AI company acquired by Google two years ago.

The ancient Chinese board game of Go has long been considered impossible for computers to play at a world-class level, despite similar breakthroughs in other games like chess and checkers. Go's simple rules and elaborate possibilities have made it one of the most sought-after milestones in the field of AI research.

Read more: Why Google's Go win is such a big deal

The third game, however, further established AlphaGo's superiority. In particular, it was seen to capably navigate tricky situations known as ko that didn't come up in the prior two matches, which had prompted speculation that it could be a potential tripping point for the program.

"I'm very excited that we've been able to instil Go's level of beauty inside a computer."

"Go is a very beautiful game and I think it teaches a lot about life, much more so than a game like chess," said Google co-founder and Alphabet president Sergey Brin after the match. "When you watch really great Go players play, it is like a thing of beauty. So I'm very excited that we've been able to instil that level of beauty inside a computer. I'm really honored to be here in the company of Lee Se-dol, such an incredible player, as well as the DeepMind team who've been working so hard on the beauty of a computer."

"To be honest we are a bit stunned and speechless," said DeepMind founder Demis Hassabis. "AlphaGo can compute tens of thousand positions a second, but it's amazing that Lee Se-dol is able to compete with that and push AlphaGo to the limit. We came here to challenge Lee Se-dol because we wanted to see what AlphaGo was capable of, and his amazing genius and creative skills have done that."

alphago-go-google-deepmind-challenge-match-"Sam Byford"-01

"I do apologize for not being able to satisfy people's expectations," said Lee, who believes that he had no chance in the first game, missed opportunities in the second, and succumbed to pressure today. He asked for people to continue to show interest in the remaining two games, despite his overall loss. "I believe [Lee] would have been difficult to beat today by any other top professional," said 9-dan pro player and match commentator Michael Redmond, who called AlphaGo a "work of art" that could revolutionize Go play in the future.

DeepMind's AlphaGo program has gone further than anyone else by using an advanced system based on deep neural networks and machine learning, which has seen it overwhelm Lee over the course of three games. The series is the first time a professional 9-dan Go player has taken on a computer; Lee was competing for a $1 million prize.

The Verge is in Seoul for the entire Google DeepMind Challenge Match series — follow our coverage at this dedicated hub, and watch the matches live on YouTube.

Read more: DeepMind's Demis Hassabis on how AI will shape the future