Abstract: Canning Knowledge: Roles of Expert Systems and Knowledge-Based Engineering in Shaping the Knowledge-Based Society

Jo Ann Oravec


Expert system and knowledge-based engineering (KBE) research and development attracted considerable levels of corporate, university, and military support in the 1980s and early 1990s. One of the early exemplars of a "successful" expert system was developed for the U.S. Campbell Soup Company in 1985 to gauge how long to cook its products. One of the experts assigned to this effort (Aldo Cimino) was retiring after forty-four years, so the technological system was reportedly designed to replace him rather than training skilled human apprentices. The artificial intelligence and information technology communities subsequently trained many "knowledge engineers" in certificate and degree programs. This paper addresses the lessons learned from an approach that centered on the notion of individual expertise (through "canning" and "cloning" human experts) rather than collaborative knowledge (such as that obtained through recommendation systems and wikis, made popular with the rise of the Internet). It also analyzes the continuing legacy of the university curricula and business training efforts that were shaped to accommodate KBE approaches. Public discourse generated two decades ago by the prospect of replacing human experts with expert systems holds insights for recent economic and social concerns about the unemployment of many highly skilled and well-educated individuals.