The gauntlet was laid down last month, when it was first announced that Google's Artificial Intelligence software AlphaGo will play a one-on-one match with Lee Sedol, the world's No. 1-ranked Go player.
Well, that day has arrived, as Wired is reporting that a Four Seasons Hotel in downtown Seoul, South Korea will be the location where the tech company gets to put its AI to the test in a matchup against Sedol. AI vs. human.
"The interest in this match is huge," Hajin Lee, a professional Go player who helped organize the match, told Wired "It's unprecedented."
Demis Hassabis (pictured above) of Google's DeepMind knows the deck is stacked against the company's AI.
"They give us a less than five percent chance of winning," Hassabis told Wired, referring to the world's elite Go players. "But what they don't realize is how much our system has improved. It's improving while I'm talking with you."
How much it has improved will be discovered today and during the week. Wired reports that, to win this matchup with Sedol, Google's AlphaGo must not only be able to properly emulate the analytical skills of a human, but also be able to catch on to a human's intuition.
Machines have defeated humans in everything from chess, checkers, Scrabble and Jeopardy over the years, but have struggled at Go, a game dating back to ancient China that utilizes black and white stones on a grid — described as a bit more intricate than chess.
Adding to the difficulty of the task of the five-game, seven-day match with $1 million on the line is Sedol's track record, which includes having the second-most international Go championships.
While plenty of eyes will be locked on the match over the next several days, Oren Etzioni, University of Washington computer science professor and executive director of the Allen Institute for AI, says he wishes the technology could be put toward a greater goal.
"The million-dollar question is: 'Who's gonna win?'" Etzioni told Wired about the AI vs. human Go match. "But the billion-dollar question — or maybe even the trillion-dollar question — is: 'How do we build AI systems for fuzzy situations that are less artificial than a board game?'"
