Maybe LLM tooling can help people demystify some of that. I think that would be a good thing.
In my experience, LLM tools are not infallible and I've seen them introduce changes in one place that causes defects in two or three other places. They write code that I find difficult to read and that can make diagnosing these bugs harder than it needs to be. People have brought me code that LLM tools have written because they are surprised that it doesn't work and are wondering if somehow they, the human, just doesn't understand something.
Maybe this is just another bump and these tools will continue to improve. Until these things aren't a problem. As it stands today, I think there's still a place for a people in commercial software development.