In the big picture, I want this podcast to be about everything in the world. I want to talk to

experts in psychology and media and medicine and economics, but there are times when it feels to me

like there are two stories whose significance towers over just about everything else.

Story one is Donald Trump and this administration, and story two is artificial intelligence.

And today's show is about what happens when these two massive objects smash into each other,

and what we can see in the wreckage of that collision.

Recently, contract negotiations broke down between Anthropic, a leading AI company,

and the Department of War, otherwise were previously known as the Department of Defense.

The gist is that after weeks of negotiations, the Pentagon couldn't get Anthropic to agree to the

use of its technology on autonomous weapons and other military applications.

Anthropic claimed that the White House was negotiating in bad faith, forcing a private company to accept contract language that went

against its values. The White House, for its part, felt that Anthropic was trying to play God,

dictating to the military how its technology should be used in an emergency, rather than allowing democratically

elected leaders to decide for itself. I have my biases here. I lean toward anthropic.

But at one level, you could say, this was a typical, boring contract dispute. At a higher

level, however, I think it was a fight over a question with huge implications for

national security, a question that could haunt the next few years or decade of our politics.

That question is, who should control AI?

It's what happened next, however, that was most shocking and infamous.

Soon after negotiations broke down,

Secretary of War, Pete Hegzeth, took the extraordinary step of labeling Anthropic a supply chain

risk, implying that the company could not do business with any firm that holds Pentagon contracts,

...