Science

Google DeepMind computer AlphaGo beaten by human after 3 tries

A champion Go player scored his first win over a Go-playing computer program on Sunday after losing three straight times in the ancient Chinese board game, saying he finally found weaknesses in the software.

South Korea's Lee Se-dol was victorious after dropping 3 straight matches against AlphaGo

South Korean professional Go player Lee Sedol, left, shakes hands with Google co-founder Sergey Brin, right, after winning the fourth match of the Google DeepMind Challenge Match against the company`s artificial intelligence program, AlphaGo, in Seoul, South Korea, Sunday. (The Associated Press)

A champion Go player scored his first win over a Go-playing computer program on Sunday after losing three straight times in the ancient Chinese board game, saying he finally found weaknesses in the software.

Lee Sedol's victory over AlphaGo is a reminder that Google's Go-playing program has room for improvement despite winning the first three matches in the best-of-five series and the $1 million prize, which will be donated to charity. The program — the first computer system to defeat a top Go player — was developed by Google DeepMind two years ago.

"This one win is so valuable and I will not trade this for anything in the world," Lee, one of the best Go players in the world, said with a smile after entering the post-match news conference at a Seoul hotel to applause from journalists.

Lee had said earlier in the series, which began last week, that he was unable to beat AlphaGo because he could not find any weaknesses in the software's strategy.

But after Sunday's match, the 33-year-old South Korean Go grandmaster, who has won 18 international championships, said he found two weaknesses in the artificial intelligence program.

Lee said that when he made an unexpected move, AlphaGo responded with a move as if the program had a bug, indicating that the machine lacked the ability to deal with surprises.

Lee Se-Dol apologized to fans for his 'powerless display,' but said he might have a better chance in the final two matches because the extreme pressure to win had been lifted following his defeat to AlphaGo. (Lee Jin-man/Associated Press)

AlphaGo also had more difficulty when it played with a black stone, according to Lee. In Go, two players take turns putting black or white stones on a 19-by-19-line grid, with a goal of putting more territory under one's control. A player with a black stone plays first and a white-stone player gets extra points to compensate.

Lee played with a white stone on Sunday. For the final match of the series, scheduled for Tuesday, Lee has offered to play with a black stone, saying it would make a victory more meaningful.

South Korean commentators could not hide their excitement three hours into Sunday's match, when it became clear that Lee would finally notch a win. AlphaGo narrowed the gap with Lee, but could not overtake him, resigning nearly five hours into the game.

Go fans whose pride had been crushed by Lee's earlier defeats cheered the result. Prior to the series, Go fans, many of them in Asia, believed that the game would prove too complex for the machine to master.

Because there are near-infinite board positions in Go and top players rely heavily on intuition, the popular Asian board game has remained the holy grail for the artificial intelligence community for about two decades, after chess was conquered by computers.