A Google computer has beaten a grandmaster of the popular board game ‘Go’ for the third time in a row, marking a significant breakthrough for artificial intelligence over humans.
South Korea’s Lee Sedol, one of the world’s best Go players, went head-to-head with Google’s AlphaGo programme to decide the best of five matches of the Chinese board game.
In the highly anticipated showdown in Seoul, Google developers deployed software programmed to approach the complex mathematical tasks associated with the 3,000-year-old game in a “human-like” way.
Some Go professionals commented that AlphaGo displayed unorthodox, dubious moves that initially befuddled humans but made sense in hindsight.
Players in the game place black and white stones on a board while trying to remove their opponent’s pieces. Because there are so many combinations of moves, the game is said to be harder for a computer to master than chess.
AlphaGo uses two sets of “deep neutral networks” that allow it to crunch data in a more human-like fashion – ignoring millions of potential moves that humans would instinctively know were pointless.
It also employs algorithms that allow it to learn and improve from previous games, and predicts winning the game from each move, something co-creator David Silver describes as “more akin to imagination”.
“AlphaGo played consistently from beginning to the end while Lee, as he is only human, showed some mental vulnerability,” said one of Lee’s former coaches, Kwon Kap-Yong.
Despite losing the series, Lee is scheduled to play twice more against AlphaGo, today and on Tuesday.