July 13, 1956

Dartmouth workshop

Dartmouth Summer Research Project on Artificial Intelligence begins

The founding meeting supposedly held on 13 July has no documented 13 July session. That date instead matches the opening of a 2006 anniversary conference.

No reliable source puts anything on 13 July

The earliest visiting participants in the Dartmouth Summer Research Project on Artificial Intelligence began arriving in Hanover, New Hampshire, around 18 June 1956 to join the organizer John McCarthy, who already lived there.

No primary or scholarly source consulted supports 13 July as a unique opening or assembly date. Ray Solomonoff's notes, the strongest evidence for the project's timing, give the start only as "about June 18"; his surviving Dartmouth notes begin on 22 June.

The July date may come from confusion with Dartmouth's fiftieth-anniversary AI@50 conference, which ran 13–15 July 2006.

Eight weeks, three to eight people a day

The proposal had asked for a two-month, ten-man study. What happened was a rotating informal workshop of roughly eight weeks, about 18 June to 17 August, with typical daily attendance of three to eight people.

Participants met on the top floor of Dartmouth's mathematics department for discussions of machine learning, language, abstraction, neural networks, calculation, self-improvement and creativity. They arrived and left at different times; there was no plenary at which all four organizers demonstrably sat together.

The summer produced no formal collective report.

The term was already a year old

McCarthy of Dartmouth, Marvin Minsky, Nathaniel Rochester of IBM and Claude Shannon of Bell Telephone Laboratories had defined the project in a proposal dated 31 August 1955. The phrase "artificial intelligence" appears there — it was not coined in a July meeting.

The proposal also lists Minsky's affiliation as Harvard University, not MIT.

Participants did not converge on a method either: brain-inspired models and symbolic problem-solving were already pulling the young field in different directions.

Sources

Researched 24 Aug 2026 6 sources not yet audited Date disputed

How this was checked

Researched from the web into a fact sheet and rewritten from that sheet. This article has not yet had an independent model audit. How the pipeline works →

Date disputed. Sources disagree; the research places it at ~1956-06-18.

What the sources leave uncertain

No reliable primary or scholarly source consulted supports July 13, 1956 as a unique opening or assembly date.
Solomonoff’s evidence gives the start only as “about June 18”; his surviving Dartmouth notes begin June 22.
The July 13 date may result from confusion with Dartmouth’s fiftieth-anniversary AI@50 conference, which ran July 13–15, 2006.
Sources disagree over total duration, attendance and who was present at particular moments because participants came and went. The approximately eight-week reconstruction is the best documented.