Allen Newell (1927-1992) was an American computer scientist and cognitive psychologist who, together with Herbert Simon, co-created some of the earliest artificial intelligence programs and helped found the field of cognitive science at Carnegie Mellon University.
Logic Theorist and RAND
Newell began his career at the RAND Corporation, where he worked on complex systems and became interested in how humans solve problems. Collaborating with Herbert Simon and programmer Cliff Shaw, he helped design the Logic Theorist, first demonstrated in 1955-56, a program that proved mathematical theorems using heuristic search rather than brute-force calculation, often cited as the first genuine Artificial intelligence program. The team presented it at the 1956 Dartmouth workshop, the meeting conventionally treated as the founding event of AI as an academic field.
General Problem Solver
In 1957 Newell, Simon, and Shaw developed the General Problem Solver (GPS), which attempted to separate general-purpose reasoning strategies, particularly means-ends analysis, from domain-specific knowledge, so that the same reasoning engine could in principle be applied to many different problems. Although GPS's ambitions of true generality were not realized, given its heavy dependence on how a problem was represented, it was an influential early attempt at what would later be recognized as a step toward Artificial general intelligence, and it shaped decades of subsequent work on search and planning within Symbolic AI.
Cognitive architectures and Soar
Newell spent most of his career at Carnegie Mellon University, where he pursued the idea of a "unified theory of cognition," a single computational architecture capable of accounting for the full range of human cognitive behavior. This effort culminated in Soar, a cognitive architecture developed with John Laird and Paul Rosenbloom in the 1980s that modeled reasoning, learning, and memory within one integrated system. Soar and related cognitive architectures influenced later research on integrated, goal-directed systems, a lineage that connects to modern discussions of AI agents that combine planning, memory, and tool use.
Legacy
Newell and Simon jointly received the Turing Award in 1975 for their foundational contributions to artificial intelligence and the psychology of human cognition. Newell delivered the prestigious William James Lectures at Harvard in 1987, later published as "Unified Theories of Cognition" (1990), which remains a touchstone for researchers who argue that intelligence should be studied as an integrated architecture rather than a collection of separate, narrow capabilities, a view now often contrasted with the scale-driven success of modern Deep learning and large language models.