Christian Catalini

I'm your host, Jim Pethokoukis.

My guest this episode is Christian Catalini.

He's a research scientist and founder of the MIT Crypto Economics Lab.

Recently, he and some co-authors put out a really great paper called Some Simple Economics of AGI.

We didn't, but on the other panel that I did, which was kind of focused on the economics of open weights models, and then really we covered a lot of stuff.

I just basically asked Christian Catalini, who I like a lot, and this guy, John Komkov, who's a really, really fascinating thinker, a lot of thoughts around autonomous agents.

Both of them are very interested in, I guess... that question, how to like, um, sort of like sign and validate, I guess, traces as well.

Um, so yeah, I mean, I think it would be worth going even deeper with them on that, but

Like Open, OpenClau, one of the reason it probably went crazy is also like people being willing to run the local agents.

And, uh, Zooko, you have a recent tweet, uh, like, uh, tweet like, uh, Christian Catalini, who was a guest in the episode itself few, a few, a few episodes back.

He mentioned about how OpenClau is one of the big breakthrough is that it can traverse, it can, it can break through the interoper- like non-operable internet or like the closed internet, right? And makes it much more interoperable.

And you m- And I want to read the tweet you mentioned, is you said that AI will bring-

And many people quickly honed in on this being one of the real areas to focus in the wake of the incident.

MIT's Christian Catalini retweeted Ryan and said, "We flagged this exact failure mode as the most urgent to solve.

The gap between what agents do and what we can measure and verify is widening.

We're flying blind.

One, open source models will be the highest utilization and consumption over closed source.

Two, Frontier will still realize most of the economics because premium intelligence commands an economic premium." MIT's Christian Catalini thinks that it will actually split in three different ways.

The first spend category is cheap generalist, which is the commodity open weight models.

On the other end of the spectrum is the state-of-the-art generalist, i.e., the tokens from the closed labs.