I think we should label devs overreliant on AI as "Engineers who dismiss themselves"
You download a tool written by a human, you can reasonably expect that it does what the author claims it does. And more, you can reasonably expect that if it fails it will fail in the same way in the same conditions.
I wrote some Turing Machine programs back in my Philosophy of Computer Science class during the 80's, but since then my Turing Machine programming skills have atrophied, and I let LLMs write them for me now.