reh@wam.umd.edu (Richard E. Huddleston) (03/21/91)
It's starting to look at though the phenomenon of emergent behavior can explain the generally mixed results from traditional AI. Perhaps 'intelligence' isn't a 'thing' -- it's more a side effect that occurs within sufficiently complex systems. It seems, also, that the systems themselves don't have to be all that complex. Getting intelligent behavior out of computer systems might start becoming a little more like designing and building the behavior you want by properly defining the agents who behave in that desired way completely by 'accident'. (1) Does anyone know of any papers/etc. describing efforts at building models of intelligence from observed emergent behaviors? (2) How about any systems built around _The Society of Mind_ concepts? email to reh@epsl.umd.edu or post it Thanks in advance.