TWO : NIL to the computer. That was the score, as The Economist went to press, in the latest round of the battle between artificial intelligence (AI) and the naturally evolved sort. The field of honour is a Go board in Seoul, South Korea—a country that cedes to no one, least of all its neighbour Japan, the title of most Go-crazy place on the planet. To the chagrin of many Japanese, who think of Go as theirs in the same way that the English think of cricket, the game’s best player is generally reckoned to be Lee Sedol, a South Korean. But not, perhaps, for much longer. Mr Lee is in the middle of a five-game series with AlphaGo, a computer program written by researchers at DeepMind, an AI software house in London that was bought by Google in 2014. And, though this is not an official championship series, as the scoreline shows, Mr Lee is losing.
Go is an ancient game—invented, legend has it, by the mythical First Emperor of China, for the instruction of his son. It is played all over East Asia, where it occupies roughly the same position as chess does in the West. It is popular with computer scientists, too. For AI researchers in particular, the idea of cracking Go...Continue reading
Source: Science and technology http://ift.tt/1SBRZxU
EmoticonEmoticon