People who made use of LLM responsibly to create high quality output doesn't look like they're using AI.

For example, using AI as an editor. It doesn't write anything for you and you try and avoid suggestions unless you're stuck.