On Sat, 24 Aug 2002, David Dyer-Bennet wrote: #The two are pretty completely unrelated. The inset behavior ones #demonstrate the behavioral complexity possible from a simple net; #ELIZA by contrast is specifically programmed in straight-forward #procedural terms. Neither one has much to do with human cognition #that I can see, but the inset behavior one is starting to explore the #behavior and capabilities of stuff related to the hardware that human #cognition runs on. (Read "insect" for "inset"; ddb's fingers are betraying him.) I agree, their internals are quite different and ELIZA's don't generalize. I was reacting only to what I understood to be a statement about the *behavior* of the NN "insects": making a black box to black box comparison. -- Mark M.