CONCEPT
The 'Good Enough' Threshold
The quality level at which AI output becomes economically preferable to human expertise—not because machines surpass humans, but because eighty-percent quality at forty-percent cost is, for most purposes, superior economics.
The useless-class mechanism operates not through sudden displacement but through gradual compression of the expertise premium. Before AI, senior and junior knowledge workers occupied different economic categories because they produced different-quality outputs—a gap justifying two-to-three-times salary differentials. AI raises the floor: the AI-assisted junior now produces eighty-percent of the senior's quality. The market recalculates. Does the remaining twenty percent justify the remaining salary gap? For most commercial purposes: no. The senior's depth remains real, genuinely valuable in situations where it makes a difference. But the situations where it makes a difference shrink as AI expands the range where surface competence suffices. Market value migrates from depth to adequacy.
In The You On AI Field Guide
The threshold is not fixed but domain-specific and moving. In software development, the good-enough threshold for routine tasks crossed in 2025; for architectural decisions, it remains distant. In legal work, contract drafting crossed; trial strategy has not. In medical diagnosis, pattern recognition crossed; bedside manner