This week, human and machine will go head-to-head in a battle to end battles… of the board game Go.
Entering the ring for humanity is Lee Se-dol, a 33-year-old South Korean who is the current world champion of the ancient Asian board game. In the corner for machines is AlphaGo, an artificial intelligence program designed by Google subsidiary DeepMind.
There will be five matches in total, all of which Deepmind will live stream on YouTube straight from Seoul, South Korea. The matches will be broadcasted at 1 p.m. local time in Seoul (4 a.m. GMT, 11 p.m. EST the day before, 8 p.m. PST) on March 9th, 10th, 12th, 13th and 15th. Each game can take around four to five hours to complete.
The live stream is above.
While DeepMind may seem confident by broadcasting the match-up, Se-dol appears to be equally self-assured. In a press conference, he said: “I don’t think it will be a very close match. I believe it will be 5-0 [to me], or maybe 4-1. So the critical point for me will be to not lose one match,” The Guardian reports.
In October last year, AlphaGo managed to defeat the three-time European champion of Go, Fan Hui, at the game 5-0. This breakthrough, which made it onto the cover of the scientific journal Nature, was the first instance of a computer program defeating a professional Go player, heralding a new era of artificial intelligence.
Go has been played in Asia for thousands of years, and even today has over 40 million players worldwide. The game takes place on a 19-by-19 grid, and two players move black or white stones on the board in an attempt to conquer as much territory as possible by surrounding their opponent’s stones.
But what’s the big deal about a computer beating a human at a game – especially since your Windows 95 could give you a run for your money at chess? For years, Go has been considered a “grand challenge” for artificial intelligence to beat. The hurdle for AI used to be the game of chess, until IBM's computer Deep Blue beat Grandmaster Garry Kasparov in February 1996. However, compared to chess, Go contains a ludicrous amount of possible positions and patterns. In fact, the number of possible configerations on a Go board is more than the amount of atoms in the known universe. As such, it’s extremely difficult to play with a methodical chess Grandmaster-style technique, and instead requires some sense of "human intuition."
For more information on the rules of Go, head over to DeepMind's website.