Claim
AI has made competent, consensus-matching output available on demand. Distinctiveness is no longer a style choice. It is an economic obligation: the only output that retains value is the one that could not have been generated by averaging the training distribution.
Mechanism
When any model produces passable work in any domain, the marginal value of average output converges toward zero. What retains value is the output shaped by genuine perspective, specific experience, or a worldview the training data cannot reconstruct. The mechanism is supply: AI flooded the zone with adequate content, so adequate content stopped being scarce.
Conditions
Holds when: the output domain is one where AI produces passable or good average-quality work (most writing, most analysis, most visual content).
Fails when: the task requires verified real-world credentials, physical action, or a relationship-specific trust layer that no model can replicate.
Evidence
McCormick in "Riding the Leopard":
distinctiveness is "not a competitive tactic but an obligation"
He quotes Nietzsche on the stakes and frames the post-AI condition as one where conformity is economically fatal, not merely aesthetically dull.
Signals
- Brand audits find copy indistinguishable from a generic AI prompt output
- Engagement falls on content covering the same ground as ten competitors
- Buyers cannot distinguish vendor positioning without reading fine print
Counter-evidence
Many B2B procurement committees still reward familiarity over distinctiveness. Unusual positioning can read as risk. Some categories genuinely reward consensus: compliance, legal, and technical documentation.
Cross-references
- ins_brendan-hufford-four-content-failure-modes
- ins_swyx-scaling-without-slop