But this is missing the fact that the vast majority of starting jobs for artists/writers would be in the former category. Similar to how AI coding or automation hurts junior hiring more than it does senior.

I found myself thinking about this issue when I was experimenting with an MCP server to handle tuning some precision parameters for scientific simulations. Claude did a much better job than I used to do when I was a fresh PhD student, yet being given tasks like that was how I learned, so it almost felt like pulling the ladder up after myself.

In the sciences, I think this is less of a problem because the PhD to scientist pipeline is pretty normalized, labs are used to the idea of having to let younger people take longer on problems that experienced people could solve much faster. But this doesn't seem to be as normalized elsewhere.