The guiding questions are helping. Would I stake my professional reputation on this output?
Not always.
I've mainly been using AI for administrative projects like building app scripts to help me manage communications with students - or to help match tutors and students. Is the code AI is generating for those scripts good? I have no idea.
When something doesn't work properly-or as expected- I've usually been able to explain the problem well enough to AI for it to fix it- and I've learned a few things about how coding works- but I'm a novice. I wouldn't be able to perform those tasks in that way without AI, period.
And that's generally been my ethical line for using AI in a broader sense. Because it's hard on the environment and the subject of larger ethical questions on the societal scale, I've tried to use it only when it's helping me with a project that I can't do myself.