South Korean Lee Sedol has won his first match against a computer programme developed by a Google subsidiary in the ancient board game Go, denying a clean sweep for the artificial intelligence in a five-match series.
Lee, one of the world’s top players and the holder of 18 international titles, recovered from three consecutive losses against the AlphaGo programme, which was developed by DeepMind.
“This win is invaluable and I would not trade it for anything else in the world,” a jubilant Lee told reporters after the match, thanking fans for their support.
The 33-year-old professional player has admitted he underestimated AlphaGo’s skills, but also said the programme was not perfect.
DeepMind founder Demis Hassabis told reporters the loss was a valuable learning tool and would help identify weaknesses in the programme that his team need to address.
“It’s a real testament to Mr Lee’s incredible fighting spirit and he was able to play so brilliantly today after three defeats,” Mr Hassabis said.
Rules of Go
Go, which is most popular in countries such as China, South Korea and Japan, involves two contestants moving black and white stones on a square grid, with the aim of seizing the most territory.
Experts did not expect an artificial intelligence programme to beat a human professional for at least a decade, until AlphaGo beat a European champion player last year.
Lee was considered a much more formidable opponent, however.
Google executives say Go offers too many possible moves for a machine to win simply through brute-force calculations, unlike chess, in which IBM's Deep Blue famously beat former world champion Garry Kasparov in 1997.
They said AlphaGo has instead sought to approximate human intuition, by studying old matches and using simulated games to hone itself independently.
The fifth and final match is scheduled for Tuesday.
Reuters