IBM's Watson is far from the first computer program to defeat at human at a game: A PhD candidate in 1952 developed software that won tic-tac-toe. Checkers was bested by technology in 1994, and Deep Blue conquered chess three years later. In 2014, meanwhile, Google's algorithms learned to play Atari games from raw pixel inputs.

Still, one game alluded artificial intelligence: the ancient sport of Go. Until now.

Invented in China more than 2,500 years ago and played by more than 40 million people worldwide, Go requires players to place black or white stones on a board and capture the opponent's pieces or surround empty spaces to build territories.

Despite its apparent simplicity, the game features more possible moves than there are atoms in the universe, eliminating traditional "brute force" AI methods, which simply search for all conceivable sequences of moves.

"We saw this as an irresistible challenge," David Silver and Demis Hassabis of Google's DeepMind AI division, wrote in blog post. "We started building a system, AlphaGo … that would overcome these barriers."

The key was reducing the nearly infinite search space to something more manageable. To accomplish this, the team combined a state-of-the-art tree search with two deep neural networks—one to suggest intelligent moves, and another to evaluate each option. AlphaGo then chose what it deemed the most successful maneuver.

When pitted against top artificial intelligence Go programs, Google's software won all but one of its 500 games. Even reigning three-time European Go champion Fan Hui was no match for the machine, which triumphed five games to zero, marking the first time a computer program beat a professional Go player.

AlphaGo's next challenge is to play (and presumably outdo) the top player in the world, Lee Sedol, which will happen in March in South Korea.

"The most significant aspect of all this for us is that AlphaGo isn't just an 'expert' system built with hand-crafted rules, but instead uses general machine learning techniques to allow it to improve itself, just by watching and playing games," the DeepMind team said.

"While games are the perfect platform for developing and testing AI algorithms quickly and efficiently, ultimately we want to apply these techniques to important real-world problems," Silver and Hassabis added.

After all, IBM already has.

This article originally appeared on PCMag.com.