Carl

Then he starts talking about hackathons and the things that they were doing and the coding competitions between the United States, China, Russia.

It was really cool about that.

Then Scaling Dead, the AGI timeline, and really the big focus for them is automating research within two years, hence the research intern for next year.

He was talking about the future of AI hardware.

He gets a little...

I think he gave, I don't think he gave it away, but he was like, it's a good, I would let you, I would have you listen.

And then he was just talking about deep seek, like the recent deep seek stuff and then schemes and, uh, like preventing scheming models and things like that.

So I think it's like a fantastic listen hour and a half.

It's not all over the plates, but it's worth listening to.

Cause it was like, okay.

You kind of get the long-term.

I don't know how true that reporting is.

Maybe it's true, maybe it's not, but I'm like, this is a better listen if you're looking more at the long-term of AI rather than, hey, what's the next model release?

Yeah.

I don't know.

The information said there'll be a reasoning model next week being released.

That's better.

But it's interesting that it coincides with Mark saying, yeah, we've got models who are way better than that or beat that and better.

And it's just like, we'll release them.

So I always go back to...