Former Go Champion Loses to DeepMind, Quits Because AI "Cannot Be Defeated"
South Korean Go champion Lee Se-dol has retired after being beat by DeepMind's AlphaGo software. He told Yonhap news agency his decision was influenced by the fact that artificial intelligence (AI) "cannot be defeated."
An entity that cannot be defeated
“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,” Se-dol told Yonhap. “Even if I become the number one, there is an entity that cannot be defeated.”
RELATED: 11 TIMES AI BEAT HUMANS AT GAMES, ART, LAW AND EVERYTHING IN BETWEEN
Se-dol is the only human to have managed to beat AlphaGo at Go. Still, he claims his one win was due to a bug in the system.
Back in March 2016, AlphaGo played against Se-dol and won four out of five matches. However, in Game 4 Se-dol made an unexpected move at white 78 that turned the game in his favor.
He developed a white wedge between blacks at the center. AlphaGo then made a mistake on move 79 only to then declare its surrender.
A brilliant divine move
Although Se-dol's white 78 has been described as "brilliant, divine" move, the go master claims he only won due to a bug in AlphaGo that responds poorly to "tricky" moves.
"My white 78 was not a move that should be countered straightforwardly. Such a bug still occurs in Fine Art (a Chinese Go-playing computer program). Fine Art can hardly be defeated even after accepting two stone handicaps against humans. But when it loses, it loses in a strange way. It's due to a bug," Se-dol said.
Se-dol started playing Go at the age of five. He turned pro at age 10. He is considered one of the greatest Go minds in the world.
AlphaGo was developed by Deepmind. Deepmind is owned by Google's parent company Alphabet.