How the 1956 Dartmouth Workshop Founded AI as a Science

This article examines how the 1956 Dartmouth Summer Research Project on Artificial Intelligence transformed scattered computational theories into a formal academic discipline. By bringing together pioneering researchers at Dartmouth College, the workshop established the official name of the field, defined its core research goals, introduced the first working artificial intelligence program, and built a dedicated network of scientists that guided computer science for decades.

Before 1956, research into thinking machines was fragmented across various fields, including cybernetics, automata theory, and complex information processing. John McCarthy, a young mathematics professor at Dartmouth College, recognized the need to unify these efforts under a single banner. In 1955, alongside Marvin Minsky, Nathaniel Rochester, and Claude Shannon, McCarthy authored a proposal for a summer workshop. In this proposal, McCarthy coined the term "artificial intelligence" to distinguish the new discipline from cybernetics, shifting the focus away from analog feedback loops toward symbolic computation and human-like reasoning.

The workshop established a foundational hypothesis that gave the new academic discipline its clear, defining methodology. The proposal stated that "every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it." This assertion provided researchers with a concrete scientific objective: break down human cognitive functions—such as language use, abstraction, and problem-solving—into precise mathematical and logical rules that a digital computer could execute.

The event also provided immediate empirical proof that this new field was viable. During the workshop, Allen Newell and Herbert A. Simon, along with programmer Cliff Shaw, debuted the "Logic Theorist." This computer program was capable of proving mathematical theorems from Principia Mathematica by using symbolic logic rather than just crunching numbers. In one instance, the program even found a more elegant proof for a theorem than the original human authors had devised. The Logic Theorist served as the first working AI program, proving to the academic world that machines could perform tasks requiring human-like intelligence.

Rather than functioning as a traditional conference with rigid paper presentations, the Dartmouth Workshop operated as an extended six-to-eight-week brainstorming session. The core attendees—including McCarthy, Minsky, Newell, Simon, Arthur Samuel, Oliver Selfridge, and Ray Solomonoff—debated methodologies and shared early successes. They outlined the specific sub-fields that still define AI curricula today, including natural language processing, neural networks, machine learning (demonstrated through Samuel's checkers-playing program), and algorithmic search.

Ultimately, the 1956 Dartmouth Workshop formalized artificial intelligence by creating an institutional framework for future research. The core attendees left Dartmouth and established the world's primary AI research laboratories at MIT, Stanford, Carnegie Mellon, and IBM. For the next several decades, these specific scientists and their students received the vast majority of government funding for AI, wrote the foundational textbooks, and established the university departments that turned a summer brainstorming session into a permanent pillar of modern science.