John McCarthy

John McCarthy was born in Boston, Massachusetts, on September 4, 1927, the elder of two sons in a politically active immigrant family ^[1]. His father, John Patrick McCarthy, was an Irish Catholic labor organizer who worked variously as a carpenter, fisherman, and union official. His mother, Ida Glatt McCarthy, was a Lithuanian Jewish journalist who had worked for a wire service. Both parents were committed members of the Communist Party during the Depression years, and the household was full of books and debate ^[2].

The family moved frequently in search of work, eventually settling in Los Angeles, partly because young John suffered from a serious respiratory ailment and doctors recommended a warmer climate ^[1]. He recovered his health in California and turned out to be an exceptionally quick student. At Belmont High School in Los Angeles he pushed far ahead of the standard curriculum, teaching himself college mathematics from textbooks used at the California Institute of Technology ^[2].

That self-directed study paid off in an unusual way. When McCarthy entered Caltech in 1944 at age sixteen, he was allowed to skip the first two years of mathematics because he had already mastered the material on his own ^[1]. His time there was interrupted by a stint in the United States Army, where he served as a clerk, but he returned and completed his bachelor's degree in mathematics in 1948 ^[3].

A single lecture changed the direction of McCarthy's career. In September 1948 he attended the Hixon Symposium on Cerebral Mechanisms in Behavior at Caltech, where the mathematician John von Neumann spoke about self-replicating automata. The comparison between computing machines and the brain took hold of McCarthy's imagination, and he began thinking seriously about whether a machine could be made to think ^[2].

He pursued graduate work at Princeton University, earning his PhD in mathematics in 1951 under the supervision of Solomon Lefschetz with a dissertation on differential equations ^[3]. At Princeton he also met Marvin Minsky, a fellow graduate student who shared his interest in machine intelligence and who would become his most important early collaborator. After short appointments at Princeton and Stanford, McCarthy taught at Dartmouth College from 1955 ^[1].

It was at Dartmouth that McCarthy made the move that fixed his place in history. In 1955 he drafted a proposal, together with Minsky, Claude Shannon of Bell Laboratories, and Nathaniel Rochester of IBM, for a summer research workshop on what he called artificial intelligence, a term he chose deliberately to set the new field apart from the existing study of automata and cybernetics ^[4]. The Dartmouth Summer Research Project on Artificial Intelligence, held in 1956, brought together roughly a dozen researchers and is widely treated by historians as the founding event of AI as a discipline ^[4]. Any account of John McCarthy achievements has to begin here: he did not just join a field, he named it and helped call it into being.

In 1956 McCarthy moved to the Massachusetts Institute of Technology, where he and Minsky founded the MIT Artificial Intelligence Project in 1958, one of the first laboratories in the world devoted to the subject ^[3]. That same year he began designing Lisp (short for List Processing), a programming language built around symbolic expressions and recursive functions rather than numeric calculation. He described the language in a landmark 1960 paper, "Recursive Functions of Symbolic Expressions and Their Computation by Machine" ^[5]. Lisp became the dominant language of AI research for decades and introduced ideas, including conditional expressions, automatic garbage collection, and treating programs as data, that later spread through mainstream programming languages ^[4].

McCarthy also transformed how people used computers at all. In the late 1950s, computers ran one job at a time in batch mode, and users waited hours or days for results. McCarthy argued that a machine could switch rapidly among many users, giving each the illusion of a private computer. His advocacy at MIT helped launch the development of time-sharing systems, which made interactive computing practical and prefigured the client-server and cloud computing models of later decades ^[1]. In a 1961 lecture marking MIT's centennial, he suggested that computing power might one day be sold as a utility, much like electricity or telephone service, an idea that anticipated modern cloud services by nearly half a century ^[2].

On the theoretical side, McCarthy insisted that intelligent behavior required formal logical foundations. His 1959 paper "Programs with Common Sense" proposed a hypothetical system called the Advice Taker that would represent knowledge in mathematical logic and draw conclusions from it, a paper often cited as the starting point of the logic-based approach to AI ^[5]. With Patrick Hayes he later introduced the situation calculus for reasoning about action and change, and he identified the frame problem, the difficulty of specifying what does not change when an action occurs, which occupied researchers for decades ^[5]. In the 1980s he developed circumscription, one of the first formal methods of nonmonotonic reasoning, allowing a system to draw sensible default conclusions and retract them when new information arrives ^[3].

Recognition followed. McCarthy received the Turing Award, computing's highest honor, in 1971 for his contributions to artificial intelligence ^[3]. He was later awarded the Kyoto Prize in 1988, the National Medal of Science in 1990, and the Benjamin Franklin Medal in Computer and Cognitive Science in 2003 ^[6].

In 1962 McCarthy left MIT for Stanford University, where he would spend the rest of his career. Three years later he founded the Stanford Artificial Intelligence Laboratory, known as SAIL, and served as its director until 1980 ^[3]. Housed for years in a semicircular building in the foothills above campus, SAIL developed a freewheeling culture and produced early work on robotics, computer vision, speech recognition, and expert systems, alongside practical inventions by its members that ranged from early display terminals to contributions toward laser printing and computer typography ^[2].

McCarthy's own interests at Stanford stayed close to the logical core of AI, but he also enjoyed provocative demonstrations. In 1966 he arranged a series of computer chess matches played by telegraph against a group at the Institute for Theoretical and Experimental Physics in Moscow, one of the first international computer versus computer contests; the Soviet program won the match ^[2]. He had been thinking about chess as a testbed for machine reasoning since the 1950s, and he is credited with naming the alpha-beta pruning technique used in game-playing programs ^[4].

He formally retired from Stanford in 2000 as a professor emeritus but kept writing, lecturing, and arguing. His personal website filled up with essays on the sustainability of human progress, in which he marshaled numbers to argue that material progress could continue for centuries, along with a page of science fiction and commentary on whatever public question had caught his attention ^[1].

McCarthy married three times. His first marriage, to Martha Coyote, ended in divorce; the couple had two daughters, Susan and Sarah ^[1]. His second wife, Vera Watson, was a programmer at IBM and an accomplished mountaineer who had made a solo ascent of Aconcagua in South America. She died in 1978 while attempting Annapurna in the Himalayas as a member of the first American women's expedition to the peak, a loss that friends said affected McCarthy deeply ^[2]. He later married Carolyn Talcott, a computer scientist at Stanford and afterward at SRI International, and they had a son, Timothy ^[1].

Colleagues remembered him as brilliant, blunt, and sometimes prickly, a man who might answer a question days after it was asked, picking up the conversation as if no time had passed ^[2]. He was a lifelong lover of argument. Raised by Communist parents, he moved steadily rightward in adulthood and by the 1970s described himself as a conservative in the Republican tradition, delighting in challenging campus orthodoxies on Usenet forums and mailing lists ^[1]. Among the John McCarthy facts that people find most surprising is that this champion of machine reasoning was also a devoted reader of science fiction who wrote his own story, "The Robot and the Baby," poking fun at both AI hype and AI panic.

Even in his final decade McCarthy kept a public presence unusual for a scientist of his generation. He continued to publish papers on formalizing common sense and context in logic, topics he had pursued since the 1950s, and he maintained that human-level AI was possible in principle even though he thought the field had underestimated the difficulty of common-sense reasoning ^[5]. He was skeptical of approaches that abandoned logic entirely, and he never stopped defending the research program he had laid out half a century earlier.

Honors continued to arrive, including election to the National Academy of Sciences and the National Academy of Engineering, and in 2011 he was inducted into IEEE Intelligent Systems' AI Hall of Fame ^[6]. McCarthy died of heart failure at his home in Stanford, California, on October 24, 2011, at the age of 84 ^[1]. His death came in the same month as those of Steve Jobs and Dennis Ritchie, prompting widespread reflection on the passing of a founding generation of computing ^[2].

Few scientists get to name a field; fewer still supply its first working tools. McCarthy did both. The phrase artificial intelligence, his 1955 coinage, now labels one of the largest research enterprises and industries in the world, and Lisp, in dialects such as Common Lisp, Scheme, and Clojure, is still written and taught more than sixty years after he designed it ^[4]. The interactive style of computing that time-sharing made possible became simply the way computers work, and his utility computing prediction reads today like an early sketch of the cloud ^[2].

His intellectual legacy inside AI is more contested but no less real. The logic-based tradition he championed shaped knowledge representation, automated planning, and the semantic technologies behind modern databases and reasoning engines, and problems he identified, above all the frame problem and the formalization of common sense, remain live research questions ^[5]. Stanford's computer science department, the laboratory he founded, and generations of students carry his influence forward. Any John McCarthy biography ends with a paradox he would have appreciated: the machines around us grow more capable every year, yet the common-sense reasoning he considered the heart of intelligence is still not solved, exactly as he warned it would not be easy ^[3].