From my experience people who heavily rely on LLMs are allergic to learning anything new (with the exception of learning new and improved ways to generate slop). They just 'want to get stuff done', even if it means staying in a local maximum forever.