Did Hubert Dreyfus actually cite Michael Polanyi's tacit-knowledge epistemology as an influence on his AI critique, or is that only a Wikipedia editorial gloss?
The question
claim-michael-polanyis-epistemology-influenced-dreyfus-critique-of-first-generation-ai rests entirely on one unquoted-in-detail sentence from Michael Polanyi's Wikipedia biography: "His work influenced the critique by Hubert Dreyfus of 'First Generation' artificial intelligence." No primary Dreyfus text or Tier 1-2 secondary was read to confirm this during the 2026-09-08 capture session — it is a Tier 4 tertiary source, and it is the entire content of the claim, which is exactly the case sources.md's sourcing floor flags for escalation ("historical/biographical claims... escalate to Tier 1-2 if the claim is... the load-bearing point of a note").
Why it matters
The vault already holds a well-developed Dreyfus cluster
(entity-hubert-dreyfus, claim-dreyfus-critique-targeted-symbolic-ai-not-neural-nets)
that names Heidegger's phenomenology as Dreyfus's grounding influence, and
the vault separately holds Polanyi as the source of "tacit knowledge" behind
claim-polanyis-paradox-tacit-tasks-resist-automation. If Dreyfus himself
named Polanyi as an influence, that is a genuine, citable second
philosophical root for the AI critique, worth folding into
entity-hubert-dreyfus's connects_to. If it is an editor's inference from
thematic resemblance rather than a documented citation, the claim should be
downgraded or retracted, not left reading as settled.
What I'd need to answer it
- Dreyfus's own What Computers Can't Do (1972) or What Computers Still Can't Do (1992) — does either text cite Polanyi by name, or discuss tacit knowledge in Polanyi's specific terms rather than only Heidegger's Being-in-the-world vocabulary?
- A secondary scholarly source on Dreyfus's influences (a philosophy-of-AI historian, an obituary, an intellectual biography) that names Polanyi explicitly, independent of Wikipedia's own phrasing.
Candidate next moves
- A direct read of the opening chapters of either Dreyfus book, where he typically surveys his philosophical debts, would likely resolve this in one pass.
- If neither book is reachable, check whether Dreyfus's 1988 Daedalus essay (already a vault primary, see claim-dreyfus-critique-targeted-symbolic-ai-not-neural-nets) or his 2007 "Why Heideggerian AI Failed" essay (already partially read for the Brooks-credit question) mentions Polanyi anywhere in passing.
Progress log
- 2026-09-10 — answered by direct primary read of Dreyfus's 1972 What Computers Can't Do (archive.org OCR scan): claim-dreyfus-1972-wccd-names-quotes-polanyi-as-embodied-skill-source confirms Dreyfus names and quotes Polanyi twice by name in Chapter 7, citing two separate Polanyi works, not merely a thematic Wikipedia gloss; claim-dreyfus-critical-of-polanyi-in-same-1972-passage settles that the engagement is critical rather than simple endorsement. What settled it: the endnotes to Chapter 7 (notes 28 and 32) of the primary text itself, both directly read this session. A related but distinct finding — claim-polanyi-nearly-absent-from-dreyfus-taylor-2015-retrieving-realism — shows the influence does not read as durable across Dreyfus's later career; that does not reopen this question, which was scoped to the book Wikipedia's sentence is actually about.