The number 1 Go player in the world declares his retirement because AI cannot be defeated

Lee Se-dol, a Korean Go player who has won the world championship 18 times, is considered the greatest Go player in the modern era who announced his retirement.

Lee Se-dol, a Korean Go player who has won the world championship 18 times, is considered the greatest Go player in the modern era who announced his retirement.

"With the participation of AI in Go, I realized that I was not the best player even if I became the best in the world with crazy efforts. Even if I became the best, there was still a forces cannot be defeated, " Lee Se-dol told Yonhap news agency.

Go originated in China with over 2,500 years old is considered one of the most complex games in the world. For many years, the complexity of Go was considered to be "overwhelming" with the most sophisticated computer programs.

Picture 1 of The number 1 Go player in the world declares his retirement because AI cannot be defeated

Lee Se-dol lost to AlphaGo AI with a score of 1-4 in 2016.

However, this changed in 2016 when Lee Se-dol lost 1 to 4 before the AlphaGo AI system developed by Google's AI-owned company DeepMind. The defeat was so shocking that Lee Se-dol apologized to the Korean people: "I failed. I feel sorry that the match ended with such a score. I wanted it to end with the result. better".

According to Google, AlphaGo AI can do this starting by learning, researching and analyzing data of hundreds of thousands of Go games. From there, AlphaGo AI can predict attacks from opponents, also analyze what is the most effective counter attack.

Google said that in 2017, DeepMind upgraded AlphaGo to AlphaGo Zero . And in 100 matches with his predecessor, AlphaGo Zero won. This shows the terrifying development of artificial intelligence.

Picture 2 of The number 1 Go player in the world declares his retirement because AI cannot be defeated

Lee Se-dol retired because he thought that AI could not be defeated in Go.

Lee Se-dol is expected to close his illustrious chess career by confronting HanDol, a Korean intelligence-based development system. Before Lee Se-dol, HanDol defeated the top 5 Go players in the country.

Not only go, artificial intelligence also shows terrifying power in many other games. Last month, AlphaStar, a program of DeepMind for the game "StarCraft II" , reached the grandmaster level when it played better than 99.8% of players on this game.

  • Google's AlphaGo artificial intelligence beats the world's number one Go player
  • AlphaGo Victory: A major event in the evolutionary history of mankind
Update 11 December 2019
« PREV
NEXT »
Category

Technology

Life

Discover science

Medicine - Health

Event

Entertainment