Just a quick one today.

‘In-game’ AI’s have come a long way since the ‘uniform blocks of baddies twitching back and forth’ of space Invaders.  Some of the more sophisticated ‘bots’ now use basic weak-AI principles to allow them to learn from player behaviour – resulting in them becoming continually more efficient at killing the gamer.  Furthermore some of these bots are now so sophisticated that they passed a Turing test in 2012.

Here’s a link to a (very readable) thesis that J.M.P. van Waveren wrote on the subject back in 2001.

Anyway – check out this terrific (and somewhat creepy) little story about in-game AI’s (apparently) learning the most important lesson of all; that ‘the only way to win the game is not to play’ – as predicted by the 1983 film ‘War Games’.  What’s more these clever and ostensibly peaceful bots then gang up on and eliminate pesky humans who threaten their security and stability.

[youtube=http://www.youtube.com/watch?v=NHWjlCaIrQo&w=420&h=315]

Awesome.

I think…