ABSTRACT
Metaphors strongly influence scientific inquiry, biasing our models and the questions we ask. Historically, metaphors for the human brain have ranged from “mechanical like a clock” to telephone switchboards and digital computer systems. Recently, some computer scientists have suggested that we study human-computer interaction in terms of how the “wet computer” (the human brain) is related to the “dry computer” (programs and hardware). Although such talk may have a place, it is scientifically a questionable guide for either appropriately using today’s computer programs or inventing more intelligent machines. In particular, to move beyond present-day programs to replicate human capability in artificial intelligence, we must begin by acknowledging how people and today’s computers are different. We must ground and direct our inquiry by the frank realization that today’s programs are incapable of conceiving ideas-not just new ideas-they cannot conceive at all. The nature of conceptualization is perhaps the most poorly grasped of all cognitive phenomena we seek to explain and replicate in computer systems.
