Textbook of AI

1936–, Computer scientist; "father of expert systems"

Also known as: Edward A. Feigenbaum

Edward Albert Feigenbaum is an American computer scientist sometimes called the "father of expert systems". A student of Herbert Simon at Carnegie Tech, he moved to Stanford in 1965 and led the development of two foundational expert systems: DENDRAL (with Joshua Lederberg and Bruce Buchanan, from 1965), which inferred molecular structure from mass-spectrometry data, and contributed to MYCIN (with Edward Shortliffe and Buchanan, from 1972), which diagnosed bacterial infections.

Feigenbaum's central methodological insight was that knowledge, not algorithms, is the primary source of an AI system's competence. DENDRAL's chemistry rules and MYCIN's clinical rules came from extensive interviews with human experts; the resulting programs sometimes outperformed their human teachers on benchmark cases. The methodology became the dominant approach to applied AI through the 1980s expert-systems boom.

Feigenbaum co-founded the Stanford AI Lab's Heuristic Programming Project, served as Chief Scientist of the US Air Force (1994–1997), and won the Turing Award (jointly with Raj Reddy) in 1994. His 1983 book The Fifth Generation (with Pamela McCorduck) brought expert-systems thinking to a broader audience.