NPR’s Steve Inskeep asks Machine Gods host Casey Newton about the reaction to Anthropic’s preparations to become a publicly traded company.
Transcription
STEVE INSKEEP, HOST:
We will try to understand Anthropic’s business plan. The artificial intelligence company that produces Claude is preparing to go public by selling shares on the stock market. They published a prospectus this week, trying to explain how it might make sense for a company that has made billions of dollars to be worth trillions of dollars. Casey Newton is the co-host of NPR’s new podcast Machine Gods, all about AI. So we called it. Hello, Casey.
CASEY NEWTON, BYLINE: Hi, Steve. Thanks for inviting me.
MAINTENANCE: OK. So when we’re talking about a company that could be worth well over $1 trillion, what does Anthropic say they plan to do to make it worth it?
NEWTON: Yes. So what Anthropic does is sell intelligence. So far, that intelligence mostly goes into writing software to write code, and Anthropic has been able to sell intelligence for those two jobs quite successfully. The question and the type of bet for investors is: can that intelligence be expanded to many other sectors of the economy and perhaps even some entire jobs could begin to be automated? Because if you can do that, then I think there will be a plausible argument that Anthropic would be worth trillions of dollars.
INSKEEP: I want to talk about the mechanics of that, though. As everyone is learning, you have to build data centers to have the computing capacity, the computing power necessary to do more and more with artificial intelligence, and that is costing trillions of dollars. What’s at stake here?
NEWTON: Yes. Well, as you may have noticed, you need data centers to run AI models, and those have become very unpopular politically. At the same time, there are enough that have already been built, and it looks like there is enough under construction that there will be a really healthy amount of compute for the next few years.
And one of the things we learn in Anthropic’s prospectus is that they were able to sell that computer at a really good gross margin. You know, for every dollar they sell at Claude, they keep about 85%. So if that remains the case, they may be able to keep their books balanced and turn a healthy profit.
MAINTENANCE: Oh. I just got some news here because there have been a lot of stories over the last year about how that computing was getting more and more expensive, to the point where people were wondering if it would be worth it for some tasks.
NEWTON: That’s right. And many observers, I read this week, have been surprised at how good the margins Anthropic has been able to achieve. Now, what I would say to worry about is that the reason Anthropic has been able to sell at such a large margin is that it has arguably had the better model for most of the last year or two. There are many competitors in this market and their models are continually improving.
INSKEEP: There are some mind-blowing words in this prospectus, though. I was reading the founders’ letter at the top, and they admit in the letter, they say in the letter that we’re going to spend money. We are going to invest money in such enormous amounts that it will seem, quote, “irresponsibly large at first glance.” Wow.
NEWTON: Yeah. It’s – you know, the really crazy thing about the AI business is that you have to buy computing many years in advance – those data centers that you just mentioned. They are tremendously expensive. They take a long time to build. And you buy them assuming there will be demand, but of course you can’t guarantee that at the time you buy it. So I think it was a signal to shareholders, like, hey, if you want to buy our stock, you have to be on a long journey because we’re going to have to make that calculation.
INSKEEP: Which means if there’s a bubble, if things go wrong, there could be some pain along the way here.
NEWTON: Absolutely. And, you know, Anthropic CEO Dario Amodei talked about that because, believe it or not, the company was criticized earlier in the year for not buying even more computing. And one of the things Darío said was, look, if I had made that bet and been wrong about the amount of demand for what we make, we would be bankrupt right now.
MAINTENANCE: OK. I want to ask about another sentence in this letter from Amodei and others. They say that if everything goes well, they could usher in, quote, “an era of miracles.”
NEWTON: (Laughter) Wouldn’t that be nice?
MAINTENANCE: Yes. But what if it turns out to be a really good business and not an era of miracles? Is everyone going to lose their shirt?
NEWTON: (Laughter) So I think there will be a lot of disappointed shareholders, yes, and people selling their shares, among other things.
INSKEEP: Casey Newton is the co-host of NPR’s Machine Gods podcast. Thank you.
NEWTON: Thanks, Steve.
AND MARTINEZ, GUEST:
And a revelation: Casey Newton is engaged to marry a person who works at Anthropic.
Copyright © 2026 NPR. All rights reserved. Please visit the terms of use and permissions pages of our website at www.npr.org for more information.
The accuracy and availability of transcripts for NPR programming may vary. The transcript text can also be reviewed to correct errors or match updates to the audio. Audio may be edited after its original broadcast or publication. The authorized record of NPR programming is the audio record.
Please note that NPR platforms and products also include certain programming produced by third parties; NPR cannot guarantee transcripts of such content and some transcripts may be produced using automated or other tools.
