According to him, now there is always “an entity that cannot be defeated.”
South Korean Go champion Lee Sedol announced his retirement from the professional game. He called the reason for the decision superiority of artificial intelligence in competitions. This was reported by Yonhap with reference to Sedol.
With the debut of AI in Go games, I’ve realized that I’m not at the top even if I become the number one through frantic efforts.Lee Sedol in an interview with Yonhap
Moreover, according to Yonhap, Sedol will still play at least one Go match to mark his departure from the professional game. In December, he plans to compete with the South Korean AI HanDol, which defeated the top five players in the country.
Mr Sedol will receive the advantage of “two stones”, but he is not sure that even with it he will be able to defeat the system in the first match. According to the player, he is going to calmly compete with AI, although he will do everything possible to win.
For many years, Go was considered one of the most difficult games for AI, since the number of combinations in it is greater than the atoms in the observable universe. However, in 2016, DeepMind’s AlphaGo system defeated Sedol for the first time in four out of five matches.
In the late 2000s, Sedol was considered the best Go player in the world. He believed he would defeat the AI and publicly apologize to fans from South Korea after losing. As of November 2019, Sedol remains the only person to win AlphaGo in at least one match.
In 2017, DeepMind improved the algorithm and released AlphaGo Zero, which was able to defeat the past AlphaGo. Unlike the first algorithm, the new version was not taught on people’s games, but played on its own.
Three days after such training, the system began to play at “superhuman” speed and defeated its predecessor in 100 games against zero. According to DeepMind, AlphaGo Zero is the strongest go player in world history.