The first round is over – Google AI won the Go-match of world champion – IDG.se

Shortly after half past eight in the morning, it was clear that 33-year-old South Korean Lee Se-dol been beaten. After an exciting game, it is now the second time that a computer has shown that it is capable of hitting a man in the Chinese board game Go. Which, again, raises the question whether computers can be programmed to take over human intellectual abilities.

Judging by the commentators was the atmosphere during the match last hours almost electric. And followers in social media think that the game at times was both exciting and well played.

Read also: The Google AI has beaten the game no computer could cope

game Go has long been seen as the sophistication of a computer to master it in the same way as a human. But researchers at Google Deep Mind did it just the challenge even more enticing. In October, they proved for the first time, they succeeded in producing an artificial intelligence that could not only meet the challenge, but also going to win the game. Then hit AlphaGo for the first time a professional, human players.

The next step was to let the computer take on world champion. And after the first five rounds have AlphaGo line. The round lasted nearly four hours and commentators in to the live match was unsure of the outcome. One of them is self-Go players and compared the game with occasions when amateurs meet a pro. When amateur looks to have the upper hand, but so slowly, in small details, shows pro who actually wins the game.

The winner of the entire match will go home with a million dollars in prize money. If Google wins, they have promised to donate the entire profit to charity. The match was followed by enthusiasts and curious via a live webcast on Youtube. And during #AlphaGo drizzled downfall jokes tight.

Both Lee Se-dol and AlphaGo played aggressively during the round. A strategy that is different from the computer was in October, when the tactic was more cautious. Something has learned since then. What is perhaps not so surprising given that it already managed to play more games than Lee Se-dol could get into throughout their lives.

Read also: Geotags? Nah, now Google can see the pixels where a photo was taken

Google says that their algorithm is more humanlike than that authorized by IBM using the Deep Blue, the chess-playing computer, which really can look through more moves than AlphaGo . They simply use different techniques, Google AI playing the previous games from the memory so as to choose the way forward.

The next match will take place in the night and Lee Se-dol’s revenge craving is no surprise.