An expert system is a computer program that captures the knowledge of a human expert in a narrow field and uses it to give advice, diagnose problems, or make decisions. The knowledge is typically stored as a large collection of “if-then” rules, and a component called the inference engine chains those rules together to reach conclusions, asking the user for facts as needed. Expert systems were the most commercially successful form of AI in the 1970s and 1980s, used for tasks such as medical diagnosis, configuring computer orders, and financial analysis.
The intellectual foundation was laid by Edward Feigenbaum at Stanford, who coined the term “knowledge engineering” for the work of extracting an expert’s knowledge and encoding it in a machine. His Stanford Heuristic Programming Project memo “The Art of Artificial Intelligence: Themes and Case Studies of Knowledge Engineering” (HPP-77-25, August 1977) argued that the power of an intelligent program comes mainly from the knowledge it contains, not from clever reasoning tricks. That insight, that knowledge is the key resource, drove the whole expert-systems movement.
Famous examples include MYCIN, which advised on treating blood infections, and XCON, which configured Digital Equipment Corporation’s computer orders and saved the company millions of dollars. Their successes proved the approach could work, but expert systems also revealed a hard limit: building and maintaining the rule base is slow and costly, and the systems are “brittle,” failing badly when faced with anything outside the cases their designers anticipated.
Why business readers should care: expert systems are the original “AI for the enterprise,” and many of their lessons still apply. They show that capturing expert know-how as rules can deliver real value in narrow, well-defined tasks, but also that such systems are expensive to maintain and do not generalize, which is exactly why modern machine learning, which learns from data, has largely replaced them.