AlphaGo is a Go playing computer program created by Google DeepMind. It's the first computer program to win against a professional human player, in an even game, in an official match.

Background Edit

AlphaGo was created by the London based DeepMind to both research and showcase the potential of the so called "deep learning" algorithms. Before AlphaGo was created, DeepMind experimented with programs that played old Atari games. Having achieved positive results, they moved on to the game of Go which is considered to be a significant challenge for AI researchers.

Match against Fan Hui Edit

During its development, AlphaGo played a 10-game match against Fan Hui[1] (2p), a former professional player from China. Half of the games were considered informal and not counted in the official score. The formal games had longer time controls. AlphaGo won the match 5-0.

Games against Fan Hui
Date Black White Category Result
5/10/2015 Fan Hui AlphaGo Formal W+2.5
5/10/2015 Fan Hui AlphaGo Informal B+R
6/10/2015 AlphaGo Fan Hui Formal B+R
6/10/2015 AlphaGo Fan Hui Informal B+R
7/10/2015 Fan Hui AlphaGo Formal W+R
7/10/2015 Fan Hui AlphaGo Informal W+R
8/10/2015 AlphaGo Fan Hui Formal B+R
8/10/2015 AlphaGo Fan Hui Informal B+R
9/10/2015 Fan Hui AlphaGo Formal W+R
9/10/2015 AlphaGo Fan Hui Informal W+R

Nature paper Edit

On January 28th 2016, the DeepMind team published a scientific paper in Nature, describing their work on AlphaGo and the algorithms used therein.[2]

Match against Lee Sedol Edit

After winning against Fan Hui and continued improvement of AlphaGo, DeepMind issued a challenge to Lee Sedol 9p[3], a Korean pro considered to be one of the strongest players in the world. The match took place in March of 2016, and was streamed live via YouTube.

Before the match Lee Sedol was reported to say that he would win 5-0 or 4-1.

The match ended with a score of 4-1 for AlphaGo.

Games against Lee Sedol
Date Black White Result Kifu
9/03/2016 Lee Sedol AlphaGo W+R [1]
10/03/2016 AlphaGo Lee Sedol B+R [2]
12/03/2016 Lee Sedol AlphaGo W+R [3]
13/03/2016 AlphaGo Lee Sedol W+R [4]
15/03/2016 Lee Sedol AlphaGo W+R [5]

References Edit