---
title: "Lighthill's 1973 report singled out 'general problem-solving programs' — the paradigm GPS belonged to — as persistently disappointing"
type: "claim"
status: "seedling"
audit_status: "capture-verified — the batch worker read the Lighthill 1973 report directly from the primary PDF (rodsmith.nz) with the verbatim quote at capture time; independent queen re-check was blocked in this headless promotion run (no web-fetch access). The PDF is open-access, so re-verification is possible and pending — routed at [[question-verify-lighthill-1973-report-quotes]]. | AUDIT 2026-07-09 (claude-opus-4-8, cross-model): primary PDF re-extracted and read this run (sha256 7011ee6f…). Source_quote confirmed VERBATIM against §3 category-B passage ('Some very interesting researches have been carried out to develop general problem-solving programs...has always been disappointing...to be developed in the 20th century.'). The report indeed never names Newell, Simon, or GPS — the note's load-bearing caveat holds. Verified — no defect."
source_url: "https://rodsmith.nz/wp-content/uploads/Lighthill_1973_Report.pdf"
source_author: "Sir James Lighthill (for the UK Science Research Council)"
source_date: 1973
source_quote: "Some very interesting researches have been carried out to develop general problem-solving programs, and such work can be of research interest to psychologists, but the performance of these programs on actual problems has always been disappointing. Students of all this work have generally concluded that it is unrealistic to expect highly generalised systems that can handle a large knowledge base effectively in a learning or self-organising mode to be developed in the 20th century."
source_tier: 1
provenance: "Promotion from 10-inbox/raw/2026-07-09-hop-lighthill-combinatorial-explosion.md, 2026-07-09"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-09-hop-lighthill-combinatorial-explosion.md"
date_created: "2026-07-09T00:00:00.000Z"
tags: ["lighthill-report","ai-winter","symbolic-ai","general-problem-solver","ai-history","history-of-ml"]
drafted_in: ["2026-07-12-good-1952","good-1952","the-wall-they-agreed-on"]
---


Beyond its general diagnosis of the combinatorial explosion
([[claim-lighthill-1973-blamed-combinatorial-explosion]]), Lighthill's *Artificial
Intelligence: A General Survey* (UK Science Research Council, 1973) named one
category of work by its era-standard label:

> "Some very interesting researches have been carried out to develop general
> problem-solving programs, and such work can be of research interest to
> psychologists, but the performance of these programs on actual problems has
> always been disappointing. Students of all this work have generally concluded
> that it is unrealistic to expect highly generalised systems that can handle a
> large knowledge base effectively in a learning or self-organising mode to be
> developed in the 20th century."

**Caveat, load-bearing.** The report never names Allen Newell, [[entity-herbert-simon|Herbert Simon]], or
the General Problem Solver (GPS). It is a UK funding survey whose acknowledged
consultees included [[entity-marvin-minsky|Minsky]] and McCarthy, not Newell or Simon — it is not a critique
of one lab. But "general problem-solving programs" was the period's standard term
for exactly the Newell–Simon research line, whose GPS and Logic Theorist rested on
heuristic search for a "good enough" solution rather than an exhaustive optimum
([[claim-simon-satisficing-and-heuristic-search-same-move]]). The
"learning or self-organising mode" framing shows Lighthill lumping the symbolic
search program with the broader ambition of large-knowledge-base autonomy.

This is the report's paradigm-specific complaint standing next to its general one:
the disappointment Lighthill records is with symbolic problem-solving, the same
"problem solving" faction the Dreyfuses set against the neural-net "learning"
faction ([[claim-dreyfus-critique-targeted-symbolic-ai-not-neural-nets]]). It
sharpens the vault's Perceptrons myth ledger by showing the 1970s AI contraction
had a distinct symbolic-AI half ([[myth-perceptrons-book-killed-connectionism]]).
See [[moc-backpropagation-origins]].