Professional go player who won against AlphaGo stops because of too strong computers

Lee Se-dol, the go player who lost to DeepMind’s AlphaGo computer in 2016 but won a single game, stops playing the board game. He believes that even if he is the best human in the world in the game, there is an entity that is better than him.

Se-dol cites the strengthened computers as the main reason for stopping in an interview with Yonhap, a South Korean news agency. “Even if I become number one, there is an entity that is unbeatable,” said Se-dol. He plays his last game as a professional against HanDol, a South Korean computer that plays go. Se-dol falls behind at the start of the game and is not sure that he will win.

The go player played against AlphaGo in 2016 and lost 4-1. According to the player, the fact that he did win that one game was due to a bug. “Move 78 I made was one that shouldn’t have been played right away.” AlphaGo did, allowing Se-dol to win the game. The rest of the matches AlphaGo won, which was awarded the rank ‘9 dan’.

The relatively simple basic rules make it possible for a computer to learn go. With the help of artificial intelligence, the software can get better and better and so it was possible to beat a professional player in the end. A documentary about AlphaGo’s victory was released in 2017. Since then, AlphaGo Zero has appeared, an improved version of the software. AlphaGo comes from DeepMind, a company owned by Google.