While only a moral victory, the world’s reigning Go champion was able to avoid being swept in his match with Google’s AI.

When Google’s DeepMind Health built its Artificial Intelligence (AI) AlphaGo Program, it was said that it would be the only computer program to conclusively beat humans at the ancient Chinese board game, Go. However, the best human Go player brought this projection to a halt, at least for a moment.

Lee Sedol, who is a Go world champion and probably the game’s best player, had been pitted against Google’s AlphaGo Program in a five game match. With all the odds against him, Lee prevented the much-hyped “comprehensive” loss that was expected and restored some human dignity as he scored an invaluable victory over AlphaGo despite the program registering three wins to clinch the $1 million prize.

Until recently, Go had proved notoriously difficult and more challenging than chess for Artificial Intelligence to master, which in essence made AlphaGo’s win a landmark moment in the history of AI development. Go is a game played by two players who take turns in placing black and white stones on a 19-by-19 grid board and the player who takes control of the most territory on the board is considered to be the winner.

iStock_000079479885_Small

Even with AlphaGo’s strong performance and overall victory, Lee showed that humans could still compete with computer programs and with his single victory that he said he would never replace for anything in the world.

“This single victory is so valuable for me that I will never trade it with anything in the world. Never have I been congratulated so much in my career just because of a single victory,” Lee said.

The 33-year-old Go grand-master from South Korea admitted that he had identified two weaknesses in the AI program. He said that AlphaGo seemed to have problems playing with a black stone and appeared to lack the ability to deal with surprising moves. On their part, DeepMind admitted that Lee was very brilliant and had an amazing fighting spirit that enabled him to pressure AlphaGo into a mistake it couldn’t recover from. The Google team added that they were able to take away some key learnings from the defeat and hope to increase the capabilities of its AI.