PERSON
Edward Feigenbaum
The Stanford computer scientist who coined “knowledge engineering” and built the first expert systems, establishing that machine competence lives in accumulated knowledge rather than clever reasoning—a thesis vindicated by large language models through a method he spent his career rejecting.
Edward Feigenbaum is the thinker whose deepest conviction came true and whose central technique was abandoned in the same stroke. Born in 1936 in Weehawken, New Jersey, and educated under Herbert Simon at Carnegie Mellon, he arrived at Stanford in 1965 with a heresy: that intelligence is not clever reasoning but accumulated knowledge, that a machine grows smart by knowing a great deal about a narrow domain rather than by reasoning brilliantly about everything. He called this the
knowledge principle, and he spent decades proving it on real problems with real chemists and real physicians looking over his shoulder. The systems he built—DENDRAL for molecular chemistry, MYCIN for infectious disease—were the first expert systems, and the discipline he founded, knowledge engineering, briefly seemed like the only road artificial intelligence would ever need. What he did not foresee was that the road would be abandoned for another: instead of carving knowledge by hand from human experts,
large