I think the copyright/ownership topic is still an open one when it comes to AI generated content/code?

This person isn't a coder, but I get your point. I'll admit I'm a skeptic but do see benefits and use what I'm comfortable with, which has been very helpful in specific cases and different ways of viewing a problem.

but yes, AI evaluating AI is a slippery slope. There's a lot of talk (again around code generation) on how the heck we validate code when an AI can create thousands of lines of it. One strategy is having a different AI writing tests. I come back to 'who do you trust'?