Hubert Dreyfus's foundational 1965 RAND Paper critiquing AI was commissioned through his brother Stuart Dreyfus's RAND employment, to evaluate Newell and Simon's "cognitive simulation" program

rand-corporationhubert-dreyfusstuart-dreyfusallen-newellherbert-simoncognitive-simulationhistory-of-aicold-war-science

In a 2005 interview, Hubert Dreyfus narrates his own route into AI criticism: his brother, then employed at the RAND Corporation, "got me invited to the Rand Corporation as a consultant to study the work of Allen Newel[l] and Simon which was the work called 'cognitive simulation' using computers." The interview says only "my brother" throughout this passage; the identification of that brother as Stuart Dreyfus is an inference, resting on Stuart being named as Hubert's brother elsewhere in the same interview and on Stuart's RAND employment being documented separately in this vault — it is not a quoted fact. Dreyfus names both the family connection that produced the invitation and the specific research program he was sent to evaluate — Allen Newell and Herbert Simon's "cognitive simulation" project, the symbolic-AI line the vault already tracks through Lighthill's 1973 critique (claim-lighthill-1973-named-general-problem-solving-disappointing). The resulting document is independently confirmed by library and museum catalogue records: "Alchemy and Artificial Intelligence," RAND Corporation Paper P-3244, Santa Monica, Calif., December 1965, viii + 90 pages (Penn Libraries catalog; Computer History Museum catalog). Those two records describe one document and corroborate only its bibliographic identity, not the commissioning. The commissioning itself rests on a single root: Dreyfus's own recollection, given in 2005 about events of roughly 1964–65. Per sources.md's oral-history rule a transcript is Tier 1 for the speaker's testimony but Tier 2–3 for the events the testimony is about, and no contemporaneous record of the invitation — a RAND consulting roster, a personnel file, an acknowledgments page — has been located by any vault session: [unverified — needs contemporaneous record].

This 1965 paper became Dreyfus's career-defining critique of AI and is commonly described in the secondary literature as the direct ancestor of What Computers Can't Do (1972) — one of the two Dreyfus works Rodney Brooks cites by name in his 1986 memo (claim-brooks-1986-memo-cites-dreyfus-as-rebuttal-target). That specific textual lineage — RAND paper expanded into book — has not itself been confirmed from the book's own preface by any vault session to date, so it is recorded here as context rather than as settled: [unverified-mechanism — needs primary]. No note in this vault previously recorded Dreyfus's RAND Paper, its number, or his brother's role in bringing it about.