- Skilled action mostly isn’t rule-following — it’s “know-how” that resists being put into words, and consciously naming the rules mid-task (a driver suddenly thinking about which gear they’re in) degrades performance rather than improving it [1].
- The five-stage model — novice, advanced beginner, competent, proficient, expert — traces skill acquisition as a change in how the task is perceived and acted on at each stage; at the top, the expert doesn’t solve problems or make decisions but simply does what normally works [1]. See the Dreyfus Model (the sixth rung MC adds, mastery, is not in this book).
- The book’s “moral”: there is more to good judgment than calculative rationality [1]. Competent performers judge by consciously combining rules and facts; proficient and expert performers draw on concrete past experience in a way that defies that kind of decomposition — a mode the authors call “arational,” not irrational [1].
- The case against 1980s expert-system AI is that it was built backward: engineers tried to extract an expert’s rules by interviewing them, but experts past the competent stage aren’t following extractable rules — so a system built from what an expert can articulate reproduces, at best, competent performance, not expert performance [1].
- A manager who changes companies loses the concrete, company-specific experience their judgment was built on, and regresses toward competent-level, rule-based decision-making in the new job until they’ve put in enough time to rebuild it [1].
Mind Over Machine
Hubert and Stuart Dreyfus's case that human expertise is intuitive rather than rule-following — the book-length elaboration of the five-stage skill-acquisition model behind MC's Dreyfus Model. First published by Free Press in 1986; the revised 1988 paperback is the edition to cite.
- Published
- 1986 (revised paperback 1988)
- Edition cited
- Free Press revised paperback, 1988
- Purchase
- AmazonBookshop.org
When you’re coaching someone who’s reached the point of trusting their gut on a call: this book’s core claim is that stopping to deliberate at that stage often makes the decision worse, not more careful. Don’t mistake a fast, confident call from someone at that level for carelessness — ask about the outcome, not for a justification they may not have.
When you’re designing training for a competency and defaulting to more rules and checklists as the answer: past the early rungs of the Dreyfus Model, the book’s argument is that rules stop helping and start getting in the way. The fix for someone stuck at “competent” isn’t a longer rulebook — it’s more reps with feedback.
When you’re hiring a senior executive or star performer away from another company: their track record was built on concrete, company-specific experience that doesn’t move with them. Expect a real dip while they rebuild that judgment in your context — plan for it in their first months rather than reading it as a bad hire.