Thank you. I am in the confusing situation of being extremely good at interpreting the nuance in human writing, yet extremely bad at detecting AI slop. Perhaps the problem is that I'm still assuming everything is human-written, so I do my usual thing of figuring out their motivations and limitations as a writer and filing it away as information. For example, when I read this article I mostly got "someone trying really hard to drive home the point that this is a dangerous problem, seems to be over-infatuated with a couple of cheap rhetorical devices and overuses them. They'll probably integrate them into their core writing ability eventually." Not that different from my assessment of a lot of human writing, including my own. (I have a fondness for em-dashes and semicolons as well, so there's that.)

I haven't yet used AI for anything I've ever written. I don't use AI much in general. Perhaps I just need more exposure. But your breakdown makes this particular example very clear, so thank you for that. I could see myself reaching for those literary devices, but not that many times nor as unevenly nor quite as clumsily.

It is very possible that my own writing is too AI-like, which makes it a blind spot for me? I definitely relate to https://marcusolang.substack.com/p/im-kenyan-i-dont-write-li...