I don't get the skepticism. Sure it needs more data to be certain, but apparent phenomenon is easily observed - just look at AI generated images, of any subject or type or size.

It drives you crazy. Triggers primal fear for shape mimics and corpses. If you weren't the one looking, it gets people visibly agitated just looking at one. Stick them into fMRI and double blind them with images. It'll probably show up in the data.

It won't take much imagination to suspect same could happen with text.

Is that a transitional issue or fundamental flaw? Is this end of the world? idk. That'll be the open question.

Why does it have to be "it's because they're clinically insane and/or backwards thinking" ? Just because you can't tell apart doesn't mean nobody can't - are those people who are keen to spot AI generated data even a minority, not the majority?