'Go' champion Lee Sedol manages first win against supercomputer

The AlphaGo program won the first three of a 'best-of-five' contest against Mr Sedol

go, player, lee, sedol, google, artificial, intelligence, supercomputer

South Korean professional Go player Lee Sedol reviews the match after finishing the fourth match of the Google DeepMind Challenge Match. Image: Lee Jin-man / AP/Press Association Images

A champion of the complex board game 'Go' has won a game against a supercomputer - his first win after three losses.

The AlphaGo program has already beaten South Korea's Lee Sedol in a best-of-five contest in Seoul.

Mr Sedol (33) is the holder of 18 international titles in the ancient Chinese game.

Even the designers of the programme - which was developed at the Google subsidiary 'DeepMind' - said they were surprised by the computer's victory.

However, in the fourth round Sedol has managed to avoid a whitewash by scoring a win.

The Guardian reports Mr Sedol believed he had found a pair of weaknesses in the AlphaGo artificial intelligence, and that the machine struggled to deal with unexpected moves.

Korean commentator Song Taegon said, “it seems Lee Sedol can now read AlphaGo better and has a better understanding of how AlphaGo moves. For the 5th match, it will be a far closer battle than before since we know each better".

Go takes place on a 19-by-19 grid, with two players taking turns placing white and black stones to take territory on the board. The game has a massive amount of possible board positions, which has made it a source of fascination for mathematicians and players alike.

Mark Keane is Chair of Computer Science at UCD. Speaking to Newstalk's Futureproof yesterday, he said AlphaGo's win is a significant step given the complexity of the game:

You can listen to more of Mark's comments on the Futureproof podcast.