Humanity could really use a win right now, and the latest test comes out of China, where a teenager named Ke Jie—the world’s best player of the ancient game Go—is taking on Google’s ultra powerful AlphaGo computer program. Unfortunately for us humans, it’s not looking great so far.
On Tuesday, 19-year-old Ke Jie was defeated in the opening match of his three-game series against AlphaGo, leaving him “deeply impressed” and “quite shocked” by the robot’s ability. If he wins the remaining two matches, he will get to take home the $1.5 million prize and retain his status as the best Go player in the world. If he loses—which experts apparently expect him to—it means world domination for AlphaGo, and yet another knock on humankind.
Here’s a rundown of the game from the New York Times:
In the first game, Mr. Ke made several moves that commentators said were reminiscent of AlphaGo’s own style. Wearing a blue tie and thick-framed black glasses, the boyish Mr. Ke kept things close in the early going. By AlphaGo’s own assessment, it did not have a big statistical advantage until after the 50th move, according to a DeepMind co-founder, Demis Hassabis.
Mr. Ke, who smiled and shook his head as AlphaGo finished out the game, said afterwards that his was a “bitter smile.”
AlphaGo—which was created by DeepMind, Google’s AI division—is apparently “getting better,” according to one commentator, which says a lot, considering it was already pretty damn good. Last year, the program defeated South Korean World Champion Lee Sedol. The software was so brilliant, according the New York Times, that top players changed the way they made moves in the game. Experts were reportedly impressed with the software’s ability to make unorthodox moves and challenge assumptions made by top players.
The match against Ke Jie also serves as a symbolic test of machine’s ability to perform better than humans in complex tasks, and shows just how well Google’s new machine learning programs can mimic the way the human brain functions. Broadly speaking, AlphaGo’s success will impact the way computers are able to search for a search for a sequence of actions.
The game of Go has long been considered an exceedingly challenging game for an AI to learn. It dates back 2,500 years and is played on a 19 x 19 grid with black and white stones. The goal of the game is to control at least 50 per cent of the board by capturing your opponent’s pieces. Computers have had difficulty defeating grandmasters in the game because there’s an estimated 10 to the power of 700 possible variations. By comparison, chess only has 10 to the power of 60 possible scenarios.
Now all that’s left for AlphaGo is to finish crushing the 19-year-old grandmaster Ke Jie, or for the teen to miraculously pull out a win. To that, we say, good luck. Do it for humanity. We desperately need this. [New York Times]