South Korean professional Go player Lee Sedol, right, puts a stone against Google's artificial intelligence program, AlphaGo, as Google DeepMind's lead programmer Aja Huang, left, sits during the match in Seoul on Wednesday.(Photo/Agencies)
Google's computer program AlphaGo on Wednesday unexpectedly defeated world human Go champion Lee Sedol of South Korea in the first of their five-game match, taking a 1-0 lead with an automatic victory by a wide margin.
Lee, the world champion of the ancient Chinese board game Go in the past 10 years, was shockingly defeated by AlphaGo, a computer program developed by Google's London-based artificial intelligence (AI) arm.
It marked the first time that AI beats world human champion at Go, which has been regarded as the last game humans can dominate in matches with computer programs.
AlphaGo beat European Go champion Fan Hui by 5-0 in October last year, defeating a human Go professional for the first time.
Lee and AlphaGo played a tight game by the middle of the match and showed aggressive plays. Lee put his first position in the right flower spot, and about one and a half minutes later, AlphaGo placed its first marker in the left flower.
One of DeepMind developers put AlphaGo's markers on the board, shown through Internet broadcasting, on behalf of the computer program.
From the very beginning, Lee played anomalously by placing his fourth marker at an irregular position, which a commentator said it was the first anomaly he has ever seen from the South Korean Go professional.
AlphaGo responded "calmly" to the anomaly, playing aggressively to Lee who in turn took counterattacks against the algorithm, the commentator said.
Lee's slight domination by the middle of the match was reversed as AlphaGo infiltrated into the right areas of black stone put by the human champion with its 102nd placing of white stone.
Lee tried to change the situations in his favor, but he finally admitted his defeat after his 186th placing of black stone. As he got black stone which allows him to put the very first marker, Lee had to win 7.5 areas more than AlphaGo for victory, according to the Chinese Go rules.
Go originated from China more than 2,500 years ago. It involves two players who take turns putting markers on a grid-shaped board to gain more areas on it. One can occupy the markers of the opponent by surrounding the pieces of the other.
The five-game match between Lee and AlphaGo will run through next Tuesday. The second match will be held in Seoul from 1 p.m. local time (0400 GMT) on Thursday.
Both human champion of the ancient Chinese board game Go and Google's computer program have played aggressively in the middle of a "match of century," commentators said Wednesday.