Go world champion Lee Sedol has finally put humanity on the scoreboard after claiming his first victory over Google DeepMind's artificial intelligence program AlphaGo in the fourth game of a five-game series between man and machine played in Seoul, South Korea. The series is now 4-1 in favour of AlphaGo with Sedol deftly avoiding a clean sweep for the AI in the historic series.
Google's software has already claimed the $1m (£700,000) in prize money, which the company says will be donated to charity.
The 33-year-old South Korean used up all of his time and won by resignation after 180 moves and two periods of "byō-yomi" (which means "second reading" in Japanese), a type of overtime, on 13 March. After AlphaGo played some unexpected moves, commentators questioned if they were mistakes or just unusual strategies that would lead to another win for Google.
According to a post-match brief, Google said, "AlphaGo held a strong position for the first half of the game, but commentators noted that Lee Sedol played a brilliant move 78, followed by a mistake by AlphaGo at move 79."
DeepMind founder Demis Hassabis said the loss was a valuable learning tool and would help identify weaknesses in the AI program that need to be addressed moving forward.
"It's a real testament to Mr Lee's incredible fighting spirit and he was able to play so brilliantly today after three defeats," said Hassabis.
Sedol, who predicted a 5-0 or 4-1 win over the AI prior to the series, said this one win feels more valuable after losing three straight games against the program. "This win is invaluable and I would not trade it for anything else in the world," said Sedol.
The ancient East Asian game of Go is considered much more complex than chess because it relies on a certain level of human "intuition" and has more possible positions than the number of atoms in the universe. In 1997, IBM's Deep Blue famously beat former world chess champion Garry Kasparov.
Although AlphaGo did beat European Go champion Fan Hui in October 2015, its win over world champion Sedol in the ongoing series marked a milestone in the progress and history of artificial intelligence.
Based on deep neural networks and machine learning, AlphaGo learns from experience rather than specific programming by studying old professional matches and simulated games to improve itself and "learn from its mistakes."