Room 3 · The Winters · 1973
The Lighthill Report
The Lighthill Report (James Lighthill · United Kingdom, 1973): The report that opened the first AI winter.
The British government asked the mathematician James Lighthill to evaluate AI research. His 1973 report concluded that the field had not delivered on its big promises and recommended cutting support. Funding fell in the United Kingdom and became more demanding in the United States: the first AI winter began.
In 1972 the British Science Research Council asked Sir James Lighthill, a distinguished applied mathematician and an expert in fluid dynamics, to review the state of artificial intelligence research. His report, "Artificial Intelligence: A General Survey," was published in 1973 together with responses from researchers in the field.
Lighthill divided the field into three parts. Category A covered advanced automation, such as industrial robots and practical applications, which he considered reasonable engineering. Category C covered computer-based research on the brain and the mind, which he considered legitimate science. Category B, the bridge between the two, was building robots and general intelligent systems, and that was where he saw the least progress for the money spent.
His central technical criticism was the combinatorial explosion. Methods that worked on small, tidy problems, such as games or toy worlds, became hopelessly slow when the number of possibilities grew to the size of real-world problems. In his view, the field's early promises had not taken this into account.
The report had an immediate effect. Funding for AI in British universities was sharply reduced, with a few exceptions such as Edinburgh. In the United States, research agencies were already asking for more practical results, and support became harder to get. Historians count these years as the start of the first AI winter.
In 1973 the BBC broadcast a debate on the report at the Royal Institution in London, in which Lighthill faced the AI researchers Donald Michie, John McCarthy and Richard Gregory. The recording survives and is still worth watching.
Ask your AI: what is a combinatorial explosion? Ask for an example with a game, and whether modern AI has found ways around it.
Source: www.chilton-computing.org.uk/inf/literature/reports/lighthill_report/c
See it in the museum →← MYCIN, the Expert System That Diagnosed · XCON, the Expert System That Made Money →
Cite this: Nazareno, I. (2026). The Lighthill Report (1973). New Mind Museum. Retrieved November 17, 2026, from https://newmindmuseum.com/museum/the-lighthill-report-1973/