In my opinion, this is not the same as people who use AI comment bots.
Of course, other cheaters had other reasons, but the few I was friends with had similar motivations. (Actually, a few just created tools because they enjoyed creating tools, not so much actually using them.)
Suppose you check something with ChatGPT, get nine hundred words, edit that down to three or even one and send that. You won't get AI;DR for "I checked, we're good."
Maybe. The issue here is we know most people are simply posting 900 words wholesale, maybe editing down 20-30 words to try and sweep some dust under the rug (meanwhile the trash is still fully featured on top of the carpet).
For your style: I'm not a writer but I question if it's even more efficient to edit 900 AI words down to 100. I can see some people finding it easier to change other's words than generate your own ideas, but I don't think it's a majority. I certainly know it's more annoying to edit 900 lines of AI code into a succinct 100 line function/class.
I'm sure that's the intention, but the alarming trend, at least here on HN, is crying wolf at the first em-dash.
Please note the key item is without checking. I'm not denying that many AI-generated documents are likely to be garbage. But refusing to read them just because of it will result in a lot of throwing the baby out with the bathwater.