Early ML & symbolic AI
What were expert systems and how did they work?
An expert system captured a specialist's knowledge as if-then rules and used an inference engine to apply them. In the 1980s they were the first big business success for AI.
By the 1970s, researchers had a practical version of the symbolic idea. An expert system, in Wikipedia’s words, is “a computer system emulating the decision-making ability of a human expert.”[1] Instead of writing ordinary procedural code, you reason “through bodies of knowledge, represented mainly as if-then rules.”[1]
Every such system has two parts: “a knowledge base, which represents facts and rules; and an inference engine, which applies the rules to the known facts to deduce new facts.”[1] So if the knowledge base holds “if X then Y” and “if Y then Z,” the engine can work out that X leads to Z. The knowledge came from interviewing human experts and writing down what they knew as rules.
The early ones came out of Stanford’s Heuristic Programming Project, led by Edward Feigenbaum, who is sometimes called the “father of expert systems.”[1] Dendral identified unknown organic molecules; MYCIN diagnosed infectious diseases from a patient’s symptoms.[1] A later system, XCON (also called R1), configured DEC VAX computers from a set of constraints.[1] In the 1980s the idea spread fast: expert systems were “created in the 1970s and then proliferated in the 1980s,” with two-thirds of Fortune 500 companies applying the technology.[1] XCON alone “was estimated to have saved the company 40 million dollars over just six years of operation.”[2]
References
- Expert system — Wikipedia
- AI winter — Wikipedia