Claim
Specific knowledge is the kind that cannot be mass-trained. If society can train someone to do what you do, society will — and your wage / margin will collapse to the cost of the training. Defensible careers and businesses are built on the intersection of curiosities that no school teaches as a single curriculum.
Mechanism
Markets price labour and services according to substitutability. A skill that can be acquired through a standard curriculum has near-infinite supply at training cost; the wage approaches that floor over time. Specific knowledge — the rare combination of talents, deep domain immersion, and accumulated context that emerges from genuine curiosity — has a thin supply curve. The operator who possesses it captures the surplus because the buyer cannot substitute another operator at lower cost. The defence does not come from gatekeeping; it comes from the fact that the path to the knowledge is not legible to schools, bootcamps, or LLMs trained on existing curricula.
Conditions
Holds when:
- The operator is positioning their work around an unusual intersection (PMM × dev tooling × AI agents; growth × narrative design; finance × biology).
- The market is willing to pay a premium for sophistication and is not price-anchored to commodity supply.
- The operator continues to compound the specific knowledge over years; one-time intersections decay if not maintained.
Fails when:
- The "specific knowledge" is actually a fashion that resolves into a bootcamp curriculum (e.g., generic prompt engineering circa 2023 collapsing into LLM-augmented developer tooling by 2026).
- The buyer cannot perceive the value of the specific intersection (selling specific knowledge to procurement-driven enterprise where everyone is interchangeable on RFP).
- The operator monetises by hourly rate rather than by leverage (selling specific knowledge as a billable hour caps it at labour rates).
Evidence
"knowledge that cannot be mass-trained: if society can train someone to replace you, it will."
— see raw/expert-content/experts/naval-ravikant.md line 18.
Signals
- Competitors / peers cannot find someone to copy your offer at lower price within 12 months.
- Buyers describe your work in terms that reference a combination of skills, not a single one.
- Pricing power that increases (not decreases) as the AI tooling for adjacent commodity skills improves.
Counter-evidence
The "specific knowledge cannot be mass-trained" claim assumes societal training capacity is the bottleneck. AI-assisted curriculum creation increasingly compresses the time from "rare" to "trainable" — from years to months for tasks like code review, copywriting, and basic positioning. Naval's heuristic still holds in spirit but the half-life of any given specific-knowledge edge is shrinking. The corollary: the knowledge has to be a moving intersection, not a fixed one.
Cross-references
- Wealth = Specific Knowledge × Leverage × Judgment, compounding over time — specific knowledge × leverage × judgment; this card unpacks the first term.
- Knowing what you don't know beats being brilliant — the discipline is the boundary, not the expansion — Munger's circle is the boundary inside which specific knowledge actually lives; outside it, the "specific" part fails.