Serious question for those using #AI to write code comments and code #documentation :

Isn't the kind of documentation it generates trivial?

I mean: if you are just describing what a method or class do, I can read that on the (clean) code myself. I don't need that repeated.

What I am interested in is the parts I can't infer from a quick read of the code. Why this decision was made. Why this method looks weird. Why this class hierarchy. Why this code that seems useless can't be deleted.