Owen Gregorian

And is that the type of response you were hoping for from Google?

And do you think that it was, you know, the people at Meta were just...

Do you think they took the right approach for that?

Yeah, and in terms of hallucinations, I would just say I totally agree that I don't think it could be a hallucination when it's so consistent.

I mean, I do have a pretty, I think, decent technical understanding of how AI works and how hallucinations work.

And at least as I've understood the definition of what a hallucination is, is that it's mainly built on the randomness of the algorithm that these AIs depend on, where it...

has probabilities of predicting what's the next word in this response I'm giving.

And it will often give you the most probable one, which would all be based on the training, but sometimes it will give you the second one or the third one or the fourth one.

And so if that were what was happening, where it was just doing a hallucination, then it would only happen one time.

It wouldn't happen over and over and over again.

And it certainly wouldn't give you like a complete fake article with all that.

So to me, that does seem like a ridiculous defense.

Um,

And I don't know what legal standing that would even have because they are still defaming you.

I mean, regardless of whether it's a hallucination or not, they built this thing and I don't think they can stand on something like section two 30.

Maybe Marcella knows more about that than me.

But, um, section two 30 I think is, is where it's user generated content, but this is not user generated content.

Google is publishing all this.

So they're the ones that are defaming you over and over and over again.

Every time someone asks about you knowingly.