The Dartmouth Summer Research Project on Artificial Intelligence (1956)

The Dartmouth Summer Research Project on Artificial Intelligence is widely considered the founding event of AI as a named discipline. It has been called “the Constitutional Convention of AI.” The workshop took place during the summer of 1956 at Dartmouth College in Hanover, New Hampshire, and ran for roughly six to eight weeks, from mid-June to mid-August.

The proposal

On August 31, 1955, four researchers submitted a proposal to the Rockefeller Foundation:

• John McCarthy (Dartmouth College): who coined the term “artificial intelligence” in the proposal itself,
• Marvin Minsky (Harvard): who had built a neural-net learning simulator,
• Nathaniel Rochester (IBM): co-designer of the IBM 701 computer,
• Claude Shannon (Bell Labs): the founder of information theory.

Their opening sentence set the ambition: “We propose that a 2 month, 10 man study of artificial intelligence be carried out during the summer of 1956 … on the basis of the conjecture 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.”

The name “artificial intelligence” was chosen deliberately. McCarthy later explained he wanted to “nail the flag to the mast”, to separate the work from the broader, less focused “automata studies” community.

What happened at Dartmouth

The workshop was not a tightly organised conference. It was an extended brainstorming session where researchers came and went. Only three people stayed the full duration: Ray Solomonoff, Marvin Minsky, and John McCarthy. On a typical day, between three and eight participants attended. Notable attendees included Allen Newell, Herbert Simon, Arthur Samuel, Oliver Selfridge, and Trenchard More.

The most celebrated result presented at the workshop was the Logic Theorist by Allen Newell, Herbert Simon, and Cliff Shaw. Built using the list-processing language IPL, the Logic Theorist could prove mathematical theorems from Whitehead and Russell’s Principia Mathematica. It proved 38 of the first 52 theorems, and for one of them it even found a proof more elegant than the original. McCarthy later acknowledged that “Newell and Simon were the stars.”

Other contributions included:

• Minsky’s ideas for a geometry theorem prover that only attempted to prove statements true in a diagram. Herbert Gelernter later implemented a prototype at IBM.
• Ray Solomonoff’s early work on algorithmic complexity, which would eventually influence the theory of inference and learning.
• Alex Bernstein’s chess program, and McCarthy’s own alpha-beta pruning heuristic for game-tree search.
• Arthur Samuel’s checkers program, one of the earliest examples of machine learning.

Why it matters

The Dartmouth workshop did not produce a single breakthrough paper or a unified research programme. Its lasting importance is symbolic and institutional: it brought together the people who would define the field for the next decades, it gave them a shared name, artificial intelligence, and it legitimised machine intelligence as a serious area of scientific inquiry.

The themes outlined in the 1955 proposal, language use, abstraction, problem-solving, and self-improvement, remain central to AI research seventy years later. From symbolic AI and expert systems to today’s large language models and neural networks, the questions asked at Dartmouth are still the questions being answered.