> The model produces better work. Some of my peers believe that large language models produce strictly better writing than they could produce on their own. Anecdotally, this phenomenon seems more common among English-as-a-second-language speakers. I also see it a lot with first-time programmers, for whom programming is a set of mysterious incantations to be memorized and recited.

AI usage is a lot higher in my work experience among people who no longer code and are now in business/management roles or engineers who are very new and didn't study engineering. My manager and skip level both use it for all sorts of things that seem pointless and the bootcamp/nontraditional engineers use it heavily. Our college hires we have who went through a CS program don't use it because they are better and faster than it for most tasks. I haven't found it to be useful without an enormous prompt at which point I'd rather just implement the feature myself.