Google’s DeepMind AlphaGo AI went up against Lee Se-Dol, world champion at Go, a Chinese strategy game.
In a best of five games, the AI defeated Mr. Lee four times for a score of 4-1. This has been a major breakthrough in the field of Artificial Intelligence.
Go is a Chinese board game and is one of the oldest known to man. The square board has gridlines. A player has either black or white stones. The two players take turns placing their stones on the grid. To capture an opponent’s pieces, a player must surround the opponent’s stone or group of stones on all sides. At the end of the game, a player’s score is calculated based on the number of pieces he has on the board and the number of pieces he has captured.
The game is generally played on a 19x19 board. While the rules are simple, there are a number of complications that come out of the numerous possibilities of placing the pieces.
It is a highly intuitive game and requires a lot of strategy and observation. There are believed to be more combinations than atoms in the universe! In fact, it was believed that computers would never be able to beat humans at this.
Here, the British company, DeepMind, comes in. The company, now owned by Google, specializes in AI – Artificial Intelligence.
AI is an advanced form of a computer program that can work independently without any human intervention after setting up the system. The programming follows complex algorithms to recreate how we think.
Instead of simple yes or no questions, human beings use emotions, expression, and intuition to make decisions. Suppose you like both pizza and pasta but you can only eat one now, which would you choose and how would you choose it? Human beings can make decisions like that but not a computer. The brain is a very complex part of our body and we have not understood everything about it to make a machine work like it.
No Match For Mortals!
Mr. Lee is a 9-dan level grandmaster at Go. Each level of skill called a dan and measures the player’s ability. In October 2015, AlphaGo had defeated a 2-dan player 5-0. The match against Mr. Lee is the hardest match for the system.
Lee forfeited the first three matches. In the fourth game where the computer played first as Black and Lee played second as white, the computer was found to struggle more playing first than playing second. The game revealed a flaw in the program that the developers could work on. For the fifth game, Mr. Lee asked to start with black in order to test his worth against the system and prove he could win with the disadvantage of starting.
The fifth match started with a mistake by the computer from which it successfully recovered to win the match. This is considered a landmark point in the development of AI due to the complexity of the decisions to be taken for the game.
How do you think this might change the future? Where might you think AI can be used? Do you think robots might successfully replace human beings or might be better than us?