John McCarthy

John McCarthy: The Man Who Named Artificial Intelligence

In the summer of 1956, a young Dartmouth mathematics professor gathered a small group of researchers for a two-month workshop under a proposal he had co-written the previous year, and used, for the first time in print, a term he had coined for the occasion: "artificial intelligence." The Dartmouth Summer Research Project is now generally credited with starting the field as a distinct discipline, and John McCarthy spent the next five decades trying to make good on the term's promise.

A Depression-Era Childhood, an Accelerated Education

Born on 4 September 1927 in Boston to an Irish immigrant father, John Patrick McCarthy, and a Lithuanian Jewish mother, Ida Glatt McCarthy, both Communist Party members during the 1930s, McCarthy's family moved repeatedly during the Great Depression as his father searched for work, eventually settling in Los Angeles where his father organized for the Amalgamated Clothing Workers. McCarthy graduated from Belmont High School two years early, having taught himself college-level mathematics from Caltech textbooks before he ever enrolled there. His formal education was interrupted when Caltech suspended him for missing physical education classes; he served in the U.S. Army before returning to finish his BS in Mathematics in 1948. A Caltech lecture by John von Neumann proved formative, and McCarthy went on to earn his PhD in mathematics from Princeton in 1951 under Donald C. Spencer, with a dissertation on partial differential equations.

Naming and Founding a Field

As an assistant professor at Dartmouth in 1955, McCarthy co-authored, with Marvin Minsky, Nathaniel Rochester, and Claude Shannon, the proposal for what became the 1956 Dartmouth Summer Research Project — the meeting historians of computing now treat as artificial intelligence's founding event, and the occasion for the field's name. McCarthy is counted, alongside Minsky, Allen Newell, and Herbert Simon, among AI's founding fathers. He moved to MIT as a research fellow in 1956, helped establish its first AI research project in 1958, and in 1962 took a full professorship at Stanford, where he built and led the Stanford Artificial Intelligence Laboratory until his retirement in 2000, turning it into a research center that rivaled MIT's own.

LISP and the Machinery of Symbolic Computing

McCarthy's most consequential technical invention was LISP, the programming language he developed in the late 1950s after recognizing that the mathematics of primitive recursive functions could be extended to manipulate symbolic expressions rather than just numbers. Published in a seminal 1960 paper, LISP became, for decades, the default language of AI research, prized for treating code and data as interchangeable and for supporting the kind of symbolic reasoning early AI work depended on. Solving a technical problem LISP created — managing memory for programs that generated data structures on the fly — McCarthy also invented automatic garbage collection around 1959, a technique now built into most modern programming languages far beyond AI.

Time-Sharing and the Roots of Networked Computing

Believing computing power should be accessible the way electricity or water was, McCarthy proposed in 1961 that computation could be sold as a utility, and he was instrumental in building some of the first time-sharing systems, including the Compatible Time-Sharing System, the BBN Time-Sharing System, and the Dartmouth Time-Sharing System, which let multiple users share a single computer's processing power interactively rather than queuing single jobs. His colleague Lester Earnest later put the stakes plainly: "The Internet would not have happened nearly as soon as it did except for the fact that John initiated the development of time-sharing systems." He also served on the ACM committee that designed ALGOL 60, proposing recursion and conditional expressions that became standard features of modern programming languages.

Common Sense, Logic, and a Skeptic's Debates

Beyond infrastructure, McCarthy pursued what he saw as AI's deeper problem: giving machines something like human common sense. His 1958 "Advice Taker" proposal outlined a program that could reason about the world using logic, and he later developed circumscription, a method for non-monotonic reasoning that let systems draw sensible default conclusions from incomplete information. He was, by his own description, an AI optimist who believed "every aspect of human intelligence could be formalized precisely enough that it could be programmed into a machine," a conviction that put him in a long-running public debate with philosopher Hubert Dreyfus over whether human reasoning could ever be fully reduced to logic.

Personal Life and Politics

McCarthy married three times; his second wife, mountaineer and computer scientist Vera Watson, died in 1978 attempting to climb Annapurna I Central, and he later married computer scientist Carolyn Talcott. Fluent in Russian and a frequent visitor to the Soviet Union, McCarthy grew disillusioned after witnessing the 1968 Soviet invasion of Czechoslovakia and became a self-described conservative Republican, a sharp turn from his parents' Communist Party membership. He declared himself an atheist in a speech at Stanford's Memorial Church and remained an active, opinionated presence on early internet forums, discussing world affairs and sustainability alongside computer science.

Why John Is Called a Genius

The Kyoto Prize, one of the highest honors he received, named McCarthy outright "the Father of Artificial Intelligence," crediting him with having "challenged the basic problems" of the field since its inception and "consistently presented new theories" across five decades — not a single invention but a sustained record of naming, founding, and re-founding an entire discipline's technical vocabulary. That is as concrete a case for genius as intellectual achievement allows: McCarthy did not merely apply existing tools to a new problem, he built the tools, coined the field's name, wrote the language much of its early research ran on, and proposed the systems architecture that let computers be shared at all. The honest complication is that McCarthy's central optimism, that logic-based, symbolic AI could scale to full human-level reasoning, was largely superseded decades later by statistical and neural-network approaches he was skeptical of, meaning the specific research program he spent his career championing did not itself deliver the breakthroughs later associated with the field he named. His genius was in framing the right questions and building durable infrastructure — LISP, garbage collection, time-sharing — more than in the particular theory of intelligence he believed would answer them.

Legacy

McCarthy died at his home in Stanford on 24 October 2011, at eighty-four, having received the Turing Award in 1971, the Kyoto Prize in 1988, and the National Medal of Science in 1990, among other honors, and having advised some thirty PhD students over his career. He left behind a phrase as much a manifesto as a name — "artificial intelligence" — and an infrastructure, from LISP to time-sharing, that computing quietly still runs on.

Achievements

• Turing Award — 1971
• National Medal of Science — 1990
• Kyoto Prize in Advanced Technology — 1988
• Notable work: artificial intelligence
• Affiliated with Princeton University, Massachusetts Institute of Technology and Stanford University
• Educated at California Institute of Technology, Princeton University and Belmont High School
• Worked as mathematician, computer scientist and engineer