Expert system
Rule-based AI program that encodes domain knowledge as if-then rules and an inference engine; commercial flagship of 1980s symbolic AI.
An expert system is a narrow symbolic AI application: a knowledge base of if-then rules plus an inference engine that chains them to mimic a human specialist (diagnosis, configuration, credit checks). Edward Feigenbaum and colleagues at Stanford popularised the idea in the 1970s; commercial hits like MYCIN and DEC’s XCON fuelled a 1980s industry boom.
Strengths were transparency and deployability in fixed domains. Weaknesses included the knowledge acquisition bottleneck (rules had to be hand-captured from experts), brittleness at the edges, and high maintenance as rule sets grew. When the bubble burst into the second AI winter, many firms pivoted toward machine learning that learns patterns from data instead of manual rule coding.