Google DeepMind 2 Lee Se-dol 0

In a landmark victory for artificial intelligence, Google’s AlphaGo has beaten the world champion in the game of Go, Lee Se-dol for a second time, taking a nearly insurmountable lead in the five game series.

In the second day, world champion Lee ended up resigning in the overtime section of the game, which left him just a minute to make every move.

In the post-match press conference, Lee admitted the match had left him “speechless”. “I admit that it was a very clear loss on my part. From the very beginning of the game I did not feel like there was a point that I was leading,” he added.

Demis Hassabis of DeepMind believes that it’s a testament to the skill of Lee Se-dol that his AI baby had to pull out some surprising moves in the second match. “We’re very pleased that AlphaGo played some quite surprising and beautiful moves, according to the commentators, which was amazing to see.”

Perhaps tellingly, while the commentators of the match viewed the game as too close to call at the mideway point, Hasabis says that AlphaGo reported confidence in winning the match at this point – before even the expert commentators could see the tide was turning.

There is a $1 million (£700,000) prize at stake, but there’s a great deal more to consider. If AlphaGo wins the overall set, it will be a significant milestone for AI.

Lee Se-dol appeared to have the upper hand through most of the first match, but in the final 20 minutes, AlphaGo took a decided lead. Eventually, Lee forfeited, conceding victory to his opponent.

“I was very surprised because I did not think that I would lose the game. A mistake I made at the very beginning lasted until the very last,” commented Lee after the match. 

The IBM computer Deep Blue beat world chess champion Garry Kasparov in 1997, and that victory was taken as a leap for the ability of AI. The rules of Go are simpler than of chess, with the aim being to circle your opponents’ stones with your own, but there’s a much wider choice of moves available to players – around 200 compared to 20 in chess. DeepMind’s team has said there are more possible positions in Go than atoms in the universe.

AlphaGo previously beat the European champion of Go, in October 2015. Today’s win is the first in a five-part battle. The games are being streamed live, and replayed via YouTube.