In 1955 and 1956, Allen Newell, Herbert Simon, and the programmer John Clifford Shaw, all working at the RAND Corporation in Santa Monica, built the Logic Theorist. It was the first program deliberately engineered to imitate human problem solving, and it is widely described as the first artificial intelligence program. The work was issued as RAND paper P-868, “The Logic Theory Machine: A Complex Information Processing System,” dated June 15, 1956, which is the date used for this entry. The paper was published shortly afterward in the IRE Transactions on Information Theory.
The Logic Theorist worked in the domain of symbolic logic. Given the axioms of Whitehead and Russell’s Principia Mathematica, it searched for proofs of theorems using heuristics, rules of thumb that prune the search, rather than brute-force checking of every possibility. This focus on heuristic search rather than exhaustive calculation was a defining idea that ran through decades of later AI work.
The program proved 38 of the first 52 theorems in Principia Mathematica, and for one theorem it found a proof shorter and more elegant than the one in the book. The result was striking enough that it was presented at the 1956 Dartmouth workshop, the meeting that gave the field its name.
Why business readers should care: the Logic Theorist established the template for symbolic AI, representing knowledge as symbols and reasoning by searching through possibilities. That approach dominated commercial AI through the expert-systems era and stands in contrast to today’s neural-network methods that learn patterns from data.