anything put out by chat gpt is pretty formulaic.

Yes, strangely it still feels that way even if you try to get it to change it's style. That's why I use the term 'averaging'. It feels like it takes a bunch of personalities and distills them down into one generic person who is doing impersonations of other personalities. Like a bad actor or something.

That isn't always going to be the case though

God, I know. When AGI arrives we are in trouble.

I have started down voting the comments where it is obvious they used AI.

This one really bugs me the most. It is easiest to spot because it is usually a summary of your post spewed back at you in some poetic fashion. And then you go to that Hive user's blog and they clearly cannot form a proper sentence in English. Of course I don't have anything against someone not speaking English properly, in fact I would rather get a broken English comment from a real person than that garbage.