Support Provided ByLearn More
Tech + EngineeringTech & Engineering

Watch Live As a Computer Challenges the Reigning World 'Go' Champion

ByAllison EckNOVA NextNOVA Next

Receive emails about upcoming NOVA programs and related content, as well as featured reporting about current events through a science lens.

The number of board configurations in the game of ‘Go’ is greater than the number of atoms in the universe.

And yet, Google engineers have devised a computer that can master the maneuvers required to beat a human expert in this 2,500-year-old Chinese game.

Support Provided ByLearn More

In October, the artificial intelligence system (named ‘AlphaGo’ by Google’s DeepMind team) defeated three-time European champion Fan Hui 5-0. Now, AlphaGo is about to face a more formidable opponent: the world champion of Go. There will be five matches in total between AlphaGo and 32-year-old Lee Se-dol of South Korea, beginning at 11:00pm Eastern time tonight (1:00pm March 9 in Seoul, where the event is being held). If Se-dol wins the tournament, he will receive a $1 million dollar prize. But if AlphaGo wins, those proceeds will go to charity—though Se-dol will still walk away with a hefty amount of cash merely for showing up.

We’ve embedded the live stream of the tournament at the bottom of this page.

game-of-go_1024x576
AlphaGo will adapt its gaming style as it learns more about how Se-dol plays.

Because the game of Go is too complex for even the most sophisticated algorithms, the AlphaGo system uses an approach that combines neural networks with reinforcement learning, as I reported back in January:

David Silver and Demis Hassabis of Google DeepMind in London have now created such a program, reported in the journal Nature today. Called “AlphaGo,” it builds on the concept of deep learning, which approximates neural networks of the human brain as a means of absorbing information and acquiring new skills. In other words, Silver and Hassabis trained AlphaGo by repeatedly exposing it to human-against-human games. They also let AlphaGo play against itself to reinforce those lessons. As a result, AlphaGo achieved a 99.8% winning rate against other Go algorithms and defeated the human European Go champion in a recent tournament five separate times.

Go requires more intensive strategizing than most games—so AlphaGo’s victory shows that it can do long-term planning. For that reason, it is considered to be “the pinnacle of game AI research for more than 20 years,” according to Hassabis. “For us, it was an irresistible challenge.”

Watch the tournament live on YouTube below. Each match is expected to last more than four hours—so if you want to catch the whole thing, make some popcorn (or coffee) and watch through the night.