It doesn’t read like AI to me,

i agree at first glance, but being confidently incorrect (especially getting the source material correct but drawing a dead wrong conclusion) is sort of a hallmark of the model.

a couple years ago i was pretty good at spotting AI work but it does get harder as time goes on.