벤처 핸즈업!

A Big Heads Up in Three Weeks : Korean Baduk(Go) Grandmaster vs. AlphaGo 본문

For English Audience/Tech Trends / Events in South Korea

A Big Heads Up in Three Weeks : Korean Baduk(Go) Grandmaster vs. AlphaGo

알 수 없는 사용자 2016. 2. 19. 03:10

Hi all,

In this section of our blog, we will post interesting news and events that are happening in South Korea on topics such as technology and venture startups for the most part.

 

The event that I want to share with you today is the big match of "baduk" between Lee Se-dol and AlphaGo. There will be five matches and the first one starts on March 9th. For those who are unfamiliar with Baduk or Go, you can think of it as a chess game with infinite possibilities of outcomes. To be more specific, it is not "infinite" but it is said that the number of outcomes in the game of baduk is more than the total number of atoms in the entire universe. Sick, right?


 

So what's all the fuss about this event??



Will Lee Se-dol Beat Google's Deeply Learned AlphaGo?

 

In fact, chess matches between a computer program and a chess champion have already taken place a while back in 1990's. 1996 was the first occurrence of the event which Deep Blue of IBM fought the official chess champion Gary Kasparov of Russia. Kasparov won 4-2 in this battle, but IBM came right back with "Deeper Blue" in 1997 to eventually beat Kasparov in the game of chess. Despite the controversies, the improved version of Deep Blue is named the world first computer to beat a human in an official chess tournament.

 

What's different this time is that the complexity of the game has significantly increased from chess, so much that brute force approach to figure out the next move simply doesn't work. AlphaGo is based on the deep learning technology, which makes a computer program to think like a human, in fact a very smart one indeed. Deep learning allows the computer program to learn from the past matches of baduk, and the result has been fabulous. It won 5-0 in the baduk match with an European Go Champion last October.

 

 

 

 

However, Lee Se-dol will not be an easy match even for AlphaGo. He is very popular in South Korea for his creative and aggressive style of play. Also, he is famous for assertive or a little bit pretentious interviews after beating Chinese opponents. In his recent interview with Chosun Daily, as he normally does, he showed off his confidence by saying that it would be difficult to lose to AlphaGo after analyzing its moves in the past matches.

 

So, who will in this baduk battle? Twenty years passed since Deep Blue vs. Kasparov. Regardless of the outcome, it is certain that the AI has improved magnificently, not only in the performance domain but in the methodology of mimicking human thinking.



Related Posts



벤처 핸즈업을 구독하세요.


Comments