REVIEW AND PROSPECTUS

Any discussions of artificial intelligence must either be conducted at a very technical level, or must come to grips with the question of how the study of thinking machines should or could influence one's philosophical and psychological concepts. It is generally agreed that intelligent behavior can only be produced by a system of information processing elements; feature detectors, theorem provers, and decision makers. A body of scientists and philosophers, known loosely as general systems theorists, has asserted that there are nontrivial laws about how such systems must be organized. If the assertion is true, then systems theory ought to be applicable to the organization of intelligent machines as well as the organization of intelligent biological, social, and physical systems. A more intuitive but perhaps more useful application of general systems theory has been suggested by Simon. He points out that all complex systems that work have certain features: the most conspicuous being hierarchical organization of components, decomposability of the total system into separate subsystems, and repeated use of similar components.