Google AI Wins Final Go Match
It’s a good day for the programmers on Google’s DeepMind team.
In the final matchup pitting their artificial intelligence against a real, live world-champion South Korean Go player, the machine emerged victorious over human, taking the series 4-1, in a major feat for the field of AI.
South Korea’s Lee Se-dol cinched just one win in the 5-game series against the Google (GOOGL) bot.
But this final matchup was no easy task for the machine. Lee, 33, fought back hard, going into overtime against the program before ultimately resigning the match, The Verge reported. The machine, called AlphaGo, also made a major mistake early on in the final game that had some worried it could cost it the match.
The win for machine learning is significant. In the past, computers beat humans at games like chess or Jeopardy by simply calculating all possible moves on a board or rote learning.
But Go, a game with more possible moves than atoms in the universe, was a different kind of challenge. Go masters say the game relies on particularly human skill; requiring creativity and intuition, the ability to think ahead and reason, a task that’s proved near impossible for computers.
DeepMind researchers said that kind of computing power has real-world promise outside of gaming, especially in health care and science. DeepMind developer Demis Hassabis, who spoke after the final game, said in the future the new kind of artificial intelligence could help enhance how we understand diseases and conduct scientific research.
Get Data Sheet, Fortune’s technology newsletter.
“We think of AI as a powerful human tool to help experts achieve more,” Hassabis said. “We’re at the stage now that we can build systems that perhaps can do something very helpful and beneficial for society.”
The Google team is taking home the $1 million prize they offered for the series. The team says they’ll be giving the money away to charity.