> they'll make fewer assumptions and ask clarifying questions before going ahead and writing code.
Which model are you talking about here? Because with ChatGPT, I struggle with getting it to ask any clarifying questions before just dumping code filled with placeholders I don't want, even when I explicitly prompt it to ask for clarification.