Artificial intelligence wins again

Artificial Intelligence, or AI to its friends, is growing fast and it seems like computers could take over at any minute. On March 19th 2016 AlphaGo, Google’s Go-playing program, spectacularly beat world champion Lee Se-dol in four out of five games. Has AI developed intuition, a key feature of the game, on top of everything else?


In Go, an ancient Chinese board game, players take turns to place black or white counters on a 19 x 19 square board to capture the opposition’s pieces and enclose territory. In a game there are typically a whopping 10171 possible moves, with about 200 allowed per turn.

When AlphaGo beat the European champion, Fan Hui, five times in October 2015, Se-dol, who watched the game, commented that the million dollar prize for beating AlphaGo would be easy money. But in the five months since that match, AlphaGo has played itself millions of times, and continuously improved.

This is how AlphaGo learns, by playing itself over and over again, spotting repeated patterns and learning from its own mistakes, as well as humans’ mistakes in previous games, in a process called reinforcement learning. It starts playing by suggesting possible moves that it’s seen before and predicting which move is most likely to be successful.

Machine learning, when a computer program teaches itself, has many applications, including facial recognition, language translation and response to speech. The pattern recognition and forward planning capabilities developed for the program have uses in ‘intelligent personal assistants’ like Siri or Cortana, as well as medical treatment planning programs, and in 3D computer games and simulations.

It seems artificial intelligence has it all, and might take over any minute! But don’t worry; computers still have games without ‘complete information’, like poker, to master. It’s much harder to program for unpredictable situations, and with the added complications of psychology it looks like there’s a long way to go.


Google DeepMind, Nature, 529, 484-489, 28th January 2016, (accessed 29/03/2016)

Google DeepMind, (accessed 29/03/2016)

Sam Byford, The Verge, 15th March 2016, Google’s AlphaGo AI beats Lee Se-dol again to win Go series 4-1, (accessed 29/03/2016)

BBC News, Artificial Intelligence: Google’s Alpha-Go beats Go master Lee Se-dol, 12th March 2016, (accessed 29/03/2016)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s