Google Hive Mind
Ars Technica has a pretty good introductory article on how robots think. One amusing bit, near the bottom of page 4:
One great demonstration of this process is a contest held every year called the Semantic Robot Vision Challenge. In this contest, robots are given a list (in words) of objects they've never seen before, and they have to drive around a contest area trying to find those objects using vision. So how do they figure out what those objects look like? Same way we answer pretty much all questions these days: Google.