IT History Journal
March 9

March 9, 2016 — AlphaGo Defeats a Human Go Champion

March 9, 2016

March 9, 2016 — AlphaGo Defeats a Human Go Champion

On March 9, 2016, the artificial intelligence system AlphaGo, developed by DeepMind (a company later acquired by Google), defeated Lee Sedol, one of the strongest Go players in the world.

The match took place in Seoul, South Korea, and became one of the most famous milestones in the history of artificial intelligence.

For decades, the game of Go had been considered one of the hardest challenges for computers. Unlike chess, where brute-force calculation works reasonably well, Go has an enormous number of possible positions. Because of this complexity, many experts believed that a machine beating a top human player was still many years away.

AlphaGo proved them wrong.

Why Go was so difficult for computers

The board in Go is 19×19, which creates an enormous number of possible moves.

Researchers often compare it like this:

  • Chess has about 10^47 possible positions
  • Go has roughly 10^170 possible positions

This number is larger than the number of atoms in the observable universe.

Because of this complexity, classical search algorithms used in chess engines were not enough. Go requires intuition, pattern recognition, and long-term strategy — abilities traditionally associated with human thinking.

How AlphaGo worked

AlphaGo combined several technologies:

  • Deep neural networks
  • Reinforcement learning
  • Monte Carlo Tree Search

The system was trained using thousands of human games and then improved by playing millions of games against itself.

Key researchers behind the project included Demis Hassabis and David Silver. Their work represented one of the first large-scale successes of deep learning applied to strategic decision-making.

The match against Lee Sedol

The match consisted of five games. AlphaGo won the series 4–1.

One particular moment became legendary: during Game 2, AlphaGo made a move (move 37) that professional players initially considered a mistake — but later realized was brilliant. The move demonstrated a creative style of play that surprised even Go experts.

Lee Sedol managed to win Game 4, showing that human creativity still mattered. His winning move in that game became famous among Go players.

The earlier battle in chess

This victory inevitably reminded many people of another famous event in computing history.

In 1997, the chess supercomputer Deep Blue, developed by IBM, defeated world chess champion Garry Kasparov. That was the first time a machine defeated a reigning world champion in chess.

But the AlphaGo victory felt even more surprising, because Go had long been considered far more difficult for computers than chess.

Why this event matters

The AlphaGo victory demonstrated that machine learning systems could achieve intuition-like strategic thinking in complex domains. It also showed the power of deep neural networks combined with self-learning.

After AlphaGo, DeepMind continued the research:

  • AlphaGo Zero learned to play Go without human data
  • AlphaZero generalized the approach to chess and shogi

These developments influenced many areas of artificial intelligence, from robotics to scientific discovery.

The match on March 9, 2016 became a global media event watched by millions. It symbolized a moment when AI systems began moving beyond narrow calculation and into domains that require pattern recognition, strategy, and creativity.

Just like Deep Blue vs Kasparov marked the end of one era in computer chess, AlphaGo vs Lee Sedol marked the beginning of a new era in artificial intelligence.