Facebook Twitter 新浪微博 Instagram YouTube Monday, Mar. 14, 2016
Search
Archive
English
English>>

Go grandmaster defeats computer after three losses

(China Daily)    07:13, March 14, 2016

Go grandmaster defeats computer after three losses

The world's top Go player Lee Sedol (R) puts the first stone against Google's artificial intelligence program AlphaGo as Google DeepMind's lead programmer Aja Huang (L) looks on during the fourth match of Google DeepMind Challenge Match in Seoul, South Korea, in this handout picture provided by Korea Baduk Assosication and released by Yonhap on March 13, 2016. [Photo/Agencies]

A South Korean Go grandmaster scored his first win over a Google-developed supercomputer on Sunday — a surprise victory after three humiliating defeats in a high-profile showdown between man and machine.

Lee Se-dol thrashed AlphaGo after a nail-biting match that lasted for nearly five hours. It was the fourth of the best-of-five series in which the computer clinched a 3-0 victory on Saturday.

Lee struggled in the early stages of the fourth match but gained a lead toward the end, eventually prompting AlphaGo to resign.

One commentator said Lee's victory was proof that artificial intelligence has not surpassed humans completely, while another said the fourth match revealed AlphaGo’s weakness, which had also been shown in the previous game.

Lee, 33, is one of the greatest players in the modern history of the ancient Chinese game, with 18 international titles to his name — the second most in the world. He earlier predicted a landslide victory over artificial intelligence, but was later forced to concede that AlphaGo was "too strong".

After his second defeat, Lee had vowed to try his best to win at least one game.

The best-known artificial intelligence victory to date came in 1997, when the IBM-developed supercomputer Deep Blue beat chess champion Garry Kasparov.

But Go, played for centuries mostly in East Asia, had long remained the holy grail for developers of artificial intelligence due to its complexity and near-infinite number of potential configurations.

AlphaGo uses two sets of deep neural networks that allow it to crunch data in a more humanlike fashion — dumping millions of potential moves that human players would instinctively know were pointless.

Its performances during the games with Lee stunned many Go experts, who described its moves as so unconventional that no human player would ever make them.

(For the latest China news, Please follow People's Daily on Twitter and Facebook)(Editor:Liang Jun,Bianji)

Add your comment

We Recommend

Most Viewed

Day|Week

Key Words