In 2016, Go grandmaster Lee Sedol, known as the world’s best human Go player at the time, was defeated in a series of games by AlphaGo, an artificial intelligence (AI) system developed by Google’s AI unit, DeepMind. This historic encounter marked the first-ever loss of a professional human Go player to an AI system. Now, looking back eight years after the event, Lee Sedol contemplates what he might have done differently had he been aware of AlphaGo’s capabilities beforehand.
Go is a strategic board game, originated from China over 2,500 years ago. The game centered on capturing and occupying more space on the board than the opponent. AlphaGo, developed in 2014, is one of the AI products created by DeepMind.
Lee Sedol was the indisputable master of the Go game, with multiple international titles won from 2002 to 2011, and glory of the world’s best Go player in 2007 and 2008. Google requested a game with AlphaGo in 2016. “I thought I would definitely win,” Sedol mentioned. He agreed to play, considering it a mere experiment. However, the news that followed proved Sedol wrong.
The alien sensation of playing against an artificial intelligence system was strange for Sedol, like playing tennis with a wall. AlphaGo won the first three rounds, Sedol rallied in the fourth, but the AI won the fifth and the final round. This loss along with the growing AI dominance led Sedol to retire professionally from Go in November 2019. He admitted that he would have considered a different career path if he’d known about AI’s potential.
Sedol indicated that AI’s advent fundamentally altered Go, especially how records of the games were used for learning. Traditional game records now only hold historical value as AI records provide an efficient tool for current players with superior learning content. Go enthusiasts grieve over how AI has changed the player’s encounter with the game. They believe a player’s playing style adds a personal and meaningful signature, which is disrupted by AI.
Sedol has since decided to develop his new board game. He intends to combine his board game design with Gemini, Google’s AI assistant, bringing a fun touch to game creation. He has already started experimenting with Gemini.
