You're reading: Posts Tagged: DeepMind

Learning to play Go: computer edition

go circuit board

In a remarkable example of us being psychic (or, what’s also known as ‘a coincidence’), our recently posted introduction to the game of Go has been made more topical by actual Go-related news.

The game of Go has long been considered a difficult game for artificial intelligences to play – much more so than chess, which has plenty of computer players. A Wired article from 2014 describes Go as ‘the ancient game that computers still can’t win’. As well as having a much larger set of possible games ($10^{761}$, as opposed to $10^{120}$ in chess), Go also has highly complicated strategy, compared to its simple rules, and moves made early on in the game can result in important changes to the state of the board further down the line.