Remember the IBM Deep Blue-Gary Kasparov chess match? It ended somewhat controversially in 1997 when Deep Blue – after a significant upgrade – rebounded from a deficit to escape with a 3½-2 ½ win. The match took nearly a year and one half. Yesterday, Google AlphaGo platform ended its Go match with a decisive win over Lee Sedol of South Korea and one of the world’s top Go players.
Go is the ancient Chinese strategy board whose number of possible moves is vast – 10761 compared to the 10120 possible in chess – making it an extremely complex game despite its relatively simple rules. The final score in the AlphaGo – Sedol match was 4-1. The match last roughly a week.
The AlphaGo technology, developed by Google acquisition Deep Mind, is “a largely self-taught Go-playing AI” as described by the IEEE Spectrum. The $1 million dollar prize offered by Google will be donated to charity. There are interesting accounts of the match in IEEE Spectrum and Wired (links below).
“The victory is notable because the technologies at the heart of AlphaGo are the future. They’re already changing Google and Facebook and Microsoft and Twitter, and they’re poised to reinvent everything from robotics to scientific research. This is scary for some. The worry is that artificially intelligent machines will take our jobs and maybe even break free from our control—and on some level, those worries are healthy,” offered Wired (In Two Moves, AlphaGo and Lee Sedol Redefined the Future).
Losing was a shock to Sedol. “Yesterday, I was surprised,” he said through an interpreter, referring to his loss in Game One. “But today I am speechless. If you look at the way the game was played, I admit, it was a very clear loss on my part. From the very beginning of the game, there was not a moment in time when I felt that I was leading.”
AlphaGo uses two neural networks, a policy network that was trained on millions of master games with the goal of imitating their play, and a value network, that tries to assign a winning probability to each given position. That way, the machine can focus its efforts on the most promising continuations. Then comes the tree-searching part, which tries to look many moves ahead.
Links to articles
IEEE Spectrum article (AlphaGo Wins Final Game In Match Against Champion Go Player, http://spectrum.ieee.org/tech-talk/computing/networks/alphago-wins-match-against-top-go-player
Wired: (In Two Moves, AlphaGo and Lee Sedol Redefined the Future, http://www.wired.com/2016/03/two-moves-alphago-lee-sedol-redefined-future/