AlphaGo beats world champ
If you’re an avid reader of The Week Junior, you might remember an article we published last year about a clever computer program named AlphaGo beating a human at Go – one of the world’s most complex board games. Well, AlphaGo has done it again – but this time it beat the world champion.
Invented more than 2,500 years ago, Go is considered one of the most difficult games in the world. It is played with black and white stones on a board marked with a grid made up of 19 lines across and 19 up and down. The players take it in turns to place one of their stones on the board. If a player surrounds one of the other player’s stones with their stones, they can take that stone away. A player wins when he or she has covered the board with more than half of their stones.
There are more than 200 different types of moves in Go. This means that there are a huge number of combinations: there are said to be more possible positions than there are atoms in the universe.
AlphaGo is unlike any other computer program built before. Instead of being programmed to play Go, it learnt how to play the game all by itself, like a human does. It did this by studying 30 million moves by human competitors. It then practised by playing against itself millions of times to work out how to win.
On 25 May, AlphaGo beat Ke Jie – the world champion Go player. This was the second out of three games that AlphaGo won, meaning that the program won overall.
DeepMind, the company that created AlphaGo, also hopes to use the intelligent technology to help solve important and complicated problems and address issues such as climate change.