Stanford’s model helicopters teach themselves to fly
Now this sounds to me like >narrow< ai, of a sort, but it is certainly generalizing, learning type behaviour. The helicopter monitors the activity of an expert helicopter pilot, and then, compensating for environmental differences (wind etc), performs the same maneuvers itself.
Stanford computer scientists have developed an artificial intelligence system that enables robotic helicopters to teach themselves to fly difficult stunts by watching other helicopters perform the same maneuvers.
Read the rest of this entry »