Google’s AlphaGo beats go world champion in first match

Google’s deep learning system AlphaGo has won the first of five matches against South Korean world champion Lee Sedol. The human top player gave up after three and a half hours. It is the first time that a deep learning system has beaten a world champion in go.

Lee had to give up against AlphaGo with 28 minutes and 28 seconds left on the clock. AlphaGo played with white and played very strongly according to experts, with the computer program often changing its style. However, Lee Sedol held out for a long time.

It is not the first time that AlphaGo beats a strong player: at the end of 2015, European champion Fan Hui was defeated. World champion Lee Sedol, however, has a 9p ranking, making him a significantly stronger player than Fan Hui at 2p.

Lee Sedol still has every chance to win the match series against AlphaGo: there are still four matches to come. The competition will take place at the Four Seasons Hotel in Seoul, South Korea. Go is an ancient board game that is much more complex than chess. For years, artificial intelligence scientists have barely made any progress in developing a high-performing go algorithm, but things have been moving fast in recent years.

In addition to DeepMind, the name of the company that developed AlphaGo before Google took over, Facebook is also working on a deep learning system that should be able to beat go players.