The promises meet the evidence

The small-world problem did not stop the predictions from growing. Within a few years, published forecasts had moved from chess and theorems to machines doing any human job and general intelligence being largely solved.

In 1973, Britain's Science Research Council published a review by .

Lighthill accepted that some narrow work had made progress. His criticism was aimed at the missing bridge between those successes and general intelligence. He also singled out combinatorial explosion as a serious obstacle.

Support for broad AI research in Britain narrowed and scepticism grew.

No single report created the winter. Confidence, money, and attention were already falling.

The period became known as the first .