Sarah Kreps

And one of the reasons why it was, I think, a real sort of

Negotiating point with Anthropic is that line between fully autonomous and human in the loop can be very ambiguous.

And Anthropic said, we do not want our AI systems to be used without human oversight.

And so there's no doubt that everyone's using AI in a lot of different ways.

But at the moment, there is still very much a kind of this is feeding information to decision makers.

And what the Pentagon was saying last week, and I think that this is still the case, is that they are keeping, quote, the human in the loop that receives this information and then uses it to make a decision rather than the algorithm making a decision.

Yeah, I mean, it's hard to know.

I think there's been a lot of discussion about escalation.

And I don't see this

playing a big role given the way these technologies are being used now.

What I think is more likely to happen is what we're seeing in Russia and Ukraine, which is without the same human cost of war, these wars can just go on and on.

And there isn't a set of constraints about, hey, is this juice still worth the squeeze?

Because there's not as much visible cost.

And

There has, of course, been cost in Russia and Ukraine.

But I think in this case, in the Middle East, it might just be the case if we're using this technology, none of our American people are at risk.

This might just continue without the same set of constraints.

And so I don't know that that affects escalation as much as sort of protractiveness of a conflict.

Yeah, and I think we should take a step back and acknowledge why they're at the tip, no pun intended, the tip of the spear here with the Pentagon, which is that a year or so ago, they seemed to go in the direction of focusing on enterprise.

And with that, more than a year ago, because their first contract with Palantir was in 2024.