Treating an AI agent like a person stops you checking its work, and that would still be a problem even if these systems turn out to have an inner life - which I get at by way of our vacuum robot, a colleague's AI intern, and an engineer who says you should never test your agents.
The argument about whether AI photography counts is old and boring; the question worth having is where a photograph gets its meaning - and whether a machine that can make meaning too quietly writes the photographer out of the conversation. The same tool points two ways, and what settles which is whether you had something to say before you opened the box.
I ran a full comparison of nine different models - from remote giants to my local setup - and found that how finished your draft is matters way more than which model you pick, especially when privacy and speed are non-negotiable.
A small toolkit of habits for using AI honestly - tagging what it touched, owning what you ship, and setting the tool up to poke holes in your thinking instead of flattering it.
A look at why I built a minimalist local-first static site generator powered by simple Python scripts, AI, and Markdown.