I like this pice. Version 5 of ChatGPT has actually meaningfully addressing the AI psychosis issue. I don’t think its all the way out yet, but its definitely going in the right direction.

Here’s a fun linguisticism…

Let’s eat, kids.

Let’s eat kids?

No, Let’s eat, kids.

Sorry I thought you said let’s eat kids.

That’s insane, do you know how insane that sounds?

Not really, its just how you said let’s eat, kids. Sounded a lot like let’s eat kids. But that would be insane, right?

Right…

Then again it does kinda sound… “right”… right now.