Grant Harvey

Or reinforce it?

Yeah.

There was the other one recently too, right, Corey?

There was the one where OpenAI basically went in and said, look, we figured out why hallucinations happen.

It's systemic based on the reward function.

Just for people who haven't read it, essentially they're like, we basically incentivize it to BS, like what Corey said, because we don't give it points for saying, I don't know, or we don't give it partial points for trying, but then realizing it's wrong.

Yeah.

Well, that actually reminds me, Corey, if you don't mind, like there was the article that I just read recently that was from 12 Grams of Carbon talking about the problem, which is the very scientific term of inshedification, which is basically how over time our platforms get worse and worse and worse.

And he basically wrote it down where it's like there's competitive reasons for that, you know, like and both sides, both the platforms and the users force that mechanism.

But then there's the alternate take to that, which was Max Levchin, who basically said, hey, we believe that over time, AI will make people make better decisions because it basically looks at all the options and it's able to parse out and pick, you know, like, let's say our credit card or our... They do...

What's it called?

Buy now, pay later.

So basically, it's like point of sale loans.

We think point of sale loans are actually better for the consumer than credit cards.

So over time, the AIs will point you to the thing that's actually better for you.

So there's this dichotomy, which is basically, is AI going to make everything worse with slop on slot?

Or is it going to make us actually choose, like you said, Mark, the best option?

I don't know if you guys can both share your thoughts on that, but I wanted to point that out.

Yes.

You remember that one, Corey?