Herbert Simon (1916-2001) was an American economist, political scientist, and computer scientist who co-created some of the earliest artificial intelligence programs and won the Nobel Memorial Prize in Economic Sciences in 1978 for his theory of bounded rationality.
Logic Theorist and early AI
Working with Allen Newell and programmer Cliff Shaw at the RAND Corporation and Carnegie Mellon University, Simon co-developed the Logic Theorist in 1955-56, widely regarded as the first program designed to mimic human problem-solving and often called the first true Artificial intelligence program. It proved theorems from Alfred North Whitehead and Bertrand Russell's "Principia Mathematica," in one case finding a shorter proof than the original. Simon and Newell attended the 1956 Dartmouth workshop, where Simon reportedly announced that they had "invented a thinking machine," a characteristically bold claim that captured both the ambitions and controversies of the young field.
Bounded rationality
Simon's economic work challenged the assumption in classical economics that decision-makers are perfectly rational optimizers. He argued instead that real agents, human or artificial, operate under bounded rationality: limited information, limited computational capacity, and limited time force decision-makers to "satisfice," choosing options that are good enough rather than provably optimal. This framework influenced both economics, earning him the 1978 Nobel Memorial Prize in Economic Sciences, and artificial intelligence research into heuristic search and decision-making under constraints, ideas that echo in later work on Reinforcement learning and resource-limited planning in AI agents.
General Problem Solver and cognitive science
Simon and Newell went on to develop the General Problem Solver in 1957, a program intended to use general-purpose heuristics, rather than domain-specific rules, to solve a wide range of problems by working backward from goals, an early precursor to later planning systems. Simon was also a founding figure of cognitive science, arguing that human cognition itself could be understood as a form of information processing comparable to computation, a stance central to the broader project of Symbolic AI.
Legacy
Simon and Newell shared the Turing Award in 1975 for their contributions to artificial intelligence and cognitive psychology, making Simon one of the few researchers to hold both a Nobel Prize and a Turing Award. His interdisciplinary career, spanning economics, psychology, computer science, and public administration, exemplified the early conviction that intelligence, whether human or machine, could be studied as a unified phenomenon of symbol manipulation and search, a foundational premise of general approaches to AI that later gave way to narrower, data-driven Machine learning methods and, eventually, debates about Artificial general intelligence.