When was AI invented? From Turing’s 1950 question to the 1956 Dartmouth workshop
AI has no single invention date. Most histories start the field in 1956, at a summer workshop at Dartmouth College whose 1955 proposal coined the name "artificial intelligence", six years after Alan Turing asked whether machines can think.
Quick answer
- 1956 is the year usually given. That summer, the Dartmouth Summer Research Project on Artificial Intelligence met at Dartmouth College in Hanover, New Hampshire. Dartmouth describes 1956 as the birth of the field of research.
- The name came a year earlier. “Artificial intelligence” appears in the workshop’s proposal, dated 31 August 1955 and written by John McCarthy (Dartmouth College), Marvin Minsky (Harvard University), Nathaniel Rochester (IBM) and Claude Shannon (Bell Telephone Laboratories). The Computer History Museum credits the term to that proposal.
- The question came first. Alan Turing’s paper “Computing Machinery and Intelligence” (Mind, October 1950) opens: “I propose to consider the question, ‘Can machines think?'”
- Some ideas are older still. The simplified model of brain cells behind today’s neural networks was published by Warren McCulloch and Walter Pitts in 1943.
AI’s starting points at a glance
| Year | What happened | Source |
|---|---|---|
| 1943 | McCulloch and Pitts publish “A Logical Calculus of the Ideas Immanent in Nervous Activity”, a simplified model of how neurons work that later neural-network research built on. | Computer History Museum |
| October 1950 | Turing’s “Computing Machinery and Intelligence” proposes the imitation game, now known as the Turing test. | Mind, vol. 59, no. 236 |
| 31 August 1955 | McCarthy, Minsky, Rochester and Shannon propose a “study of artificial intelligence” for summer 1956. | The proposal |
| 1955–1956 | Allen Newell, Herbert Simon and J.C. Shaw begin Logic Theorist, a program that went on to prove 38 theorems from Whitehead and Russell’s Principia Mathematica. | Computer History Museum |
| Summer 1956 | The Dartmouth workshop meets: the founding event of AI as a field of research. | Dartmouth College; Computer History Museum |
| 1958 | McCarthy creates the LISP (“List Processing”) programming language at MIT, in which data and programs are both lists. | Computer History Museum |
What Turing asked in 1950
Turing thought “Can machines think?” was too vague to answer, so he replaced it with a game. In his version, an interrogator in another room exchanges typed messages with two hidden players and has to work out which is which; voice can’t give anything away. He then asks what happens if a machine takes one player’s place. If the interrogator is wrong about as often as with two people, the machine has done well at the game.
Turing didn’t use the words “artificial intelligence”; that name came five years later. But his paper set out the question the field has argued about ever since: whether behaving intelligently is enough to count as intelligence.
What the Dartmouth proposal actually said
The proposal, dated 31 August 1955, asks for “a 2 month, 10 man study of artificial intelligence” during the summer of 1956 at Dartmouth College. It rests on a bold conjecture: that every aspect of learning or intelligence can in principle be described precisely enough for a machine to simulate it.
It names seven problems to work on:
- Automatic computers
- How a computer can be programmed to use a language
- Neuron nets
- The theory of the size of a calculation
- Self-improvement
- Abstractions
- Randomness and creativity
Several of these are still central today. Language is what chatbots do, neuron nets became neural networks, and self-improvement is close to what we now call machine learning (that link is our reading, not the proposal’s). The authors were optimistic about pace: they wrote that “a significant advance can be made” on one or more of the problems in a single summer. Most of these problems took decades.
Why there’s no single invention date
AI wasn’t invented the way a single device is. Different things arrived at different times:
- The ideas: mathematical models of neurons (1943) and the question of machine thinking (1950).
- The name: the August 1955 proposal.
- The field: the 1956 workshop, where researchers met under that name with a shared agenda.
- Working programs: early reasoning programs such as Logic Theorist, begun in 1955.
When someone asks “when was AI invented?”, 1956 is the most widely accepted answer, because that is when AI became a named field of research. Answers of 1943, 1950 or 1955 are also defensible, as long as you say which milestone you mean.
What it means for you
- Today’s AI tools rest on old ideas. The chatbots and writing tools small businesses use now grew out of questions posed about 70 years ago. What changed is the computing power and data available, not the basic ambition.
- Be sceptical of timelines. The field’s founders expected real progress in one summer, and many of their problems took decades. Vendors’ claims about what AI will do next deserve the same caution.
- “AI” covers a lot. From its first proposal, AI meant language, learning, reasoning and creativity together. When a product says it “uses AI”, ask which of these it actually does.
What we don’t know yet
- Dartmouth’s own page gives only the year of the workshop, not its exact dates, and accounts of who attended and for how long vary.
- Who suggested the words “artificial intelligence” in particular is usually attributed to McCarthy, but the sources used here credit the proposal as a whole.
- Whether passing Turing’s imitation game would show real intelligence is still debated; Turing’s paper itself discusses objections to it.
Common questions
Who invented AI?
No single person. Alan Turing posed the question in 1950. John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon named the field in their 1955 proposal, and the 1956 Dartmouth workshop they organised founded it as a field of research.
Who came up with the term “artificial intelligence”?
The term appears in the Dartmouth workshop proposal dated 31 August 1955. The Computer History Museum credits the coining of the term to that proposal. McCarthy, its lead author, is often named as the person who chose it.
What was the first AI program?
There’s no agreed single answer. Logic Theorist, begun in 1955 by Allen Newell, Herbert Simon and J.C. Shaw, is one of the best-known early programs; it proved 38 theorems from Principia Mathematica.
How old is AI?
About 70 years, counting from the 1956 Dartmouth workshop. Counting from Turing’s paper, it’s 76; counting from McCulloch and Pitts’ neuron model, 83.
What is the Turing test?
It’s the popular name for the imitation game in Turing’s 1950 paper: an interrogator exchanges typed messages with a hidden person and a hidden machine and tries to tell which is which.
Sources
- A. M. Turing, “Computing Machinery and Intelligence”, Mind, vol. 59, no. 236, pp. 433–460, October 1950. doi.org/10.1093/mind/LIX.236.433. Read 21 October 2026.
- J. McCarthy, M. L. Minsky, N. Rochester and C. E. Shannon, “A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence”, 31 August 1955. Copy on John McCarthy’s Stanford site. Read 21 October 2026.
- Dartmouth College, “Artificial Intelligence (AI) Coined at Dartmouth”. home.dartmouth.edu. Read 21 October 2026.
- Computer History Museum, “Timeline of Computer History: AI & Robotics” (entries for 1943, 1955, 1958 and 2006). computerhistory.org. Read 21 October 2026.