Menu Content
Go Top

Science

'Lee's 78th Move Confuses AlphaGo During 4th Match'

Written: 2016-03-14 16:46:37Updated: 2016-03-14 19:30:47

'Lee's 78th Move Confuses AlphaGo During 4th Match'

Google DeepMind, the creator of the artificial intelligence program AlphaGo, says that world Go champion Lee Se-dol’s unexpected move in the fourth match confused AlphaGo.
 
Google DeepMind CEO Demis Hassabis said on Monday in a news conference that Lee’s 78th move during Sunday's game confused AlphaGo, which didn’t realize its mistake until ten moves later.
 
Hassabis said that Lee’s creative move rendered AlphaGo’s prediction meaningless, and confused the program.
 
The DeepMind CEO also said Lee was accurate in his assessment that AlphaGo shows weaknesses when playing black. Hassabis credited Lee for figuring out AlphaGo’s weaknesses after only a few matches. 
 

Editor's Pick

Close

This website uses cookies and other technology to enhance quality of service. Continuous usage of the website will be considered as giving consent to the application of such technology and the policy of KBS. For further details >