Back in 1997 I took a course in AI Programming at UCLA. In the intro to the book it described "agents" as processes which took in information about their environments through sensors and acted on their environment through actuators. It then went on to classify the variety of types of environments in which an agent could be intended to function, and listed some of the implications of these environments.
Since then, I lost the textbook. It's probably on a friend's or relative's bookshelf, but I know not where. Thus I was thrilled to discover a similar discussion on Wikipedia which I will quote (in its current incarnation) and discuss in the 'read more' part of this article.
