The word "noise" is doing important work here. This particular job can be automated because, in many advertising transactions, nobody really cares what marketi

The word "noise" is doing important work here. This particular job can be automated because, in many advertising transactions, nobody really cares what marketing copy says. The people writing it don't care; the people reading it certainly don't care. It just needs to give a certain gestalt impression, the same way that AI images can give a certain gestalt impression. Which is to say, this isn't just something that AI could automate; this is a job that could have been outsourced to much-lower-paid human workers in (English-speaking) developing nations decades ago. There is no reason anybody should have been making $80/hr doing this job in the first place; there wasn't $80/hr of labor going on, nor was there a moat to doing the job that qualified only a rarified number of people that could then demand $80/hr. Instead, the wage being paid was a path-dependent coincidence of how most companies see marketing as a profit-center, and therefore want to offer "incentive pay" for marketing jobs; and how these same companies happened to consider copywriting — really just marketing-adjacent — as a "marketing job." Coupled with the fact that any given company doesn't usually need very many copywriters, and so doesn't focus on what sort of per-headcount cost-optimizations would be needed to "scale" their copywriting.