AI labor-market exposure follows an inverted-U pattern rather than a universal replacement curve.
AI is not replacing workers evenly. The strongest early pattern is an inverted U: the top is protected by tacit knowledge and the bottom by low replacement economics, while the middle layer absorbs the sharpest compression.1
The important distinction is not optimism versus panic. It is whether a role is being augmented, compressed, or restructured by systems that can absorb codified cognitive tasks.2
That makes the practical question sharper: watch hiring gates, job-description requirements, workflow consolidation, and the time gap between AI investment announcements and headcount actions.
Model context
This article maps to ACDP because it updates the model's assumptions about compression and augmentation in white-collar work.
Inspect ACDP