Human beings have won AlphaGo, Google's AI, for the first time

Today was the first victory for a human against a Google AI specialized in Go.

AlphaGo, created by DeepMind, was initially presented with skepticism. Few experts believe that was able to face the best players in the world of go, also known (wrongly) as Chinese chess.

Although AlphaGo managed to beat the European champion, most people, including their own creators, were aware that the biggest challenge was going to be facing the world champions of the modality, and that was what the project did, challenging Lee Sedol, considered the second best player in history.

The story of a human's victory against a Google go-to AI

To everyone's surprise, including Lee Sedol himself, AlphaGo has not only managed to keep its face against such an adversary, but even has completely crushed him.

AlphaGo had three wins in a row against Lee Sedol, three games that left the legendary champion speechless for the strange movements that the AI ​​used and that left him completely defenseless.

The second encounter was especially devastating for Sedol, who had no choice but to surrender when he saw that the machine had completely broken its flow of play in just a few movements.

With the third consecutive loss, it got to the point that Lee Sedol had to apologize publicly for underestimating the AI before the fight, ensuring that he had never felt such pressure before, despite his vast experience.

Commentators themselves did not know what to say about the AI ​​movements, which initially Seems nonsense until the master shot is revealed that leaves his opponent without options, a style of play that has surprised its own creators.

Although the winner of the matchup has already been decided with that third victory, there was still the consolation of the two remaining games, and Sedol set out to win them. And so he has, beating in fourth instance to leave the confrontation at 3-1 for the last game.

The human being wins the battle, but not the war

The key was in an error made by the AI ​​in turn 79, which left him confused. AlphaGo decides the style of play depending on how the game is progressing, switching to an aggressive or defensive mode depending on the circumstances. But that mistake left AlphaGo unsure what style to adopt, and Sedol took advantage of this indecision.

In this way, Lee Sedol becomes the first professional go player to win a game at AlphaGo, since the European champion lost 5-0. So in these moments humanity is losing 8-1.