Sebastian Mallaby

Things Sebastian Says on Podcasts

Writes about finance and technology at the Council on Foreign Relations, where he is the Paul A. Volcker senior fellow for international economics. His books include The Power Law, a history of venture capital, and The Infinity Machine, about Demis Hassabis and DeepMind. He talks about AI adoption inside companies and the economics around it. He hosts The Spillover, a Council on Foreign Relations podcast, and appears as a guest across a wide range of shows, including On with Kara Swisher, Hard Fork, and Prof G Markets, and is a two-time Pulitzer Prize finalist.

Where to Find Them

Sebastian Mallaby hosts The Spillover , writes Ground Truths , writes Unsupervised Learning with Jacob Effron , hosts The President’s Inbox , hosts ChinaTalk , writes World Review with Ivo Daalder and writes Ground Truths (Eric Topol) . They have also been a guest on Bloomberg Surveillance (7 times) , TechStuff (3 times) , Wall Street Week (2 times) , Trumponomics (2 times) , The Exchange (2 times) , Masters in Business (2 times) , Squawk Pod (2 times) , The Blog of Author Tim Ferriss (2 times) , Capital Ideas podcast , Prof G Markets , Software Engineering Daily , Summation with Auren Hoffman , Capital Allocators – Inside the Institutional Investment Industry , Open Book with Anthony Scaramucci , Executive Functions Chat , Prof G Media , Odd Lots , The Information's TITV , Conversations with Tyler , This Week in Startups , Brew Markets , On with Kara Swisher , Yascha Mounk , Enterprising Investor , Hard Fork (NYT) , The Tim Ferriss Show , Barron's Advisor , Future Ready Leadership With Jacob Morgan , Manifold , Twenty-four two , TBPN , Faster, Please! , The Chris Voss Show , The Indicator (Planet Money) and The Foreign Affairs Interview . They also write for Sebastian Mallaby on Substack.

Recently: “Demis Hassabis and the Race to Build AI with Sebastian Mallaby” on Software Engineering Daily (November 2026); “After The Builders: On Dario & Daniela Amodei and Mustafa Suleyman” on TechStuff (October 2026); “What We Got Right (and Wrong) This Year” on The Spillover (October 2026); “Central Banks in a Bind + Could AI Replace the FOMC?” on The Spillover (October 2026); “A White House Accord with AI’s Leaders 9/30/26” on Squawk Pod (September 2026); “Why Won’t Europe Do More For Ukraine? + The Rise of Pro-Russia Populism” on The Spillover (September 2026).

What They Said

“When you look at the really big deals around the data centers and the private credit that is propping up their building, that does look bubbly. Private credit is opaque. With public bonds you can see the price on the screen and whether it's trading at junk levels, so there are warning signs. With private credit, which is a big player at the moment, there are no warning signs other than just gossip. So I have started in the last couple of months to worry about a bubble.” — Sebastian Mallaby, Software Engineering Daily

Discussing whether AI is in a bubble, Mallaby explains why he has become more worried about the financing of AI infrastructure. His concern is that private credit hides distress until it is too late.

Software Engineering Daily · 2026-11-03 Permalink → Listen →
Software Engineering Daily Around 48:28 into the episode
Sebastian Mallaby

…hedge fund, which to say it blew up is a bit unfair because they're still in business, but they did have to sell their public stock portfolio to Citadel because they were down sufficiently that they were getting capital calls because they'd borrowed. And here, again, the leverage is the key point.

Kevin Ball

Exactly. Yeah. Once you're levered, then margin, you don't need that much of a dip to get this self-reinforcing margin call cycle going on.

Sebastian Mallaby

Right, right. They were levered four to one. And so just going down 25% on the equities meant they'd gone down to zero because you have to forex it. So they couldn't hold those positions. But in some sense, who cares about a fund, right? That's fine. Citadel buys it, end a problem. That's not systemic. It's unfortunate for the investors in situational awareness, but it's not a systemic problem. But when you look at some of these really big deals around the data centers and the private credit that is propping up the building of the data centers, that does look bubbly. And unfortunately, also private credit is opaque, right? They don't have to say exactly what their loans are. They may disclose something to their limited partners who are backing the fund. And that might leak out a bit. But basically, it's not like the public bond. Issuance, you can see the price on the screen, you know, that's traded a bit, so you can see what it's worth. You can see if there's distress and if it's trading at junk bond levels. So, at least there are warning signs when you finance it with bonds. But with private credit, which is a big player at the moment, there are no warning signs other than just gossip. So, I do start, I have started in the last couple of months…

Kevin Ball

So, I realize we're coming close to the end of our time here, and we've wandered over a wide range from the individual story of Demis to the current sort of layout and even the financial situation. Are there any things we haven't talked about that you think would be important to cover before we…

Sebastian Mallaby

You know, I'm very interested in this debate about open and closed models, right? I'm always most interested in the ones where I'm at least certain that I'm correct, because I do think this is a finely balanced argument. But maybe I could tell you my perspective, and you could tell me yours, and…

Kevin Ball

Yeah, it's interesting. Well, and it's interesting also when we talk about models, because often what we talk, the equivalent that we talk about there is open weights, where the weights are published. Weights are not legible in the way that code is, in the sense that open weights means I can take…

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.
“If you can 2x a function, say every coder has a coding assistant and produces twice as much code, individually they are twice as productive. But the company's IT output goes up 1.1x. Why? Before you ship the code you need the product team to look at it, you need legal to make sure it's compliant. These bottlenecks mean there's a big delta between individual productivity gain and institutional productivity gain. And we're in that phase right now.” — Sebastian Mallaby, Software Engineering Daily

Asked what institutions are needed to cope with AI-driven change, Mallaby draws on the history of electricity and on executives he talks to who find AI productivity gains hard to realise outside coding and call centers.

Software Engineering Daily · 2026-11-03 Permalink → Listen →
Software Engineering Daily Around 44:26 into the episode
Sebastian Mallaby

Yeah, so first of all, your comment makes me think the ideal investment portfolio in this world is like some GPU investments together with a portfolio of premium restaurants or something like that, right? Or maybe gyms, or

Kevin Ball

other in-person experiences, right? Yeah, exactly. Things that happen together. Like you talk about obsessions, I'm obsessed with in-person experience and immersive theater and improv and all these different things that just get you together in a place with people.

Sebastian Mallaby

Yeah, right, exactly. So, totally, that's right. You want to invest in the pure silicon and the pure experience. But you're saying, how do institutions adapt? So, I think this is again super exciting because it's so early and we're going to see how this plays out. But what we know is that typically in the first phases of invention through economic history, people do small things with the new tech. So, electricity comes along and you can reconfigure certain tasks that are more efficient, but the sort of design of the factory where maybe there had been a kind of coal-fired steam engine or something in the middle powering your loom, and then the workers were kind of organized around this shape. You know, you kind of stick with that shape, you just have electricity running the loom instead of the steam power, for example. And it takes quite a long time between the advent of electricity and the first automobile production line where you have a straight line production line where you rearrange how the workers are laid out. Because actually, oh, you know, we don't need to put the power center in the middle because it's kind of a localized steam thing. It's coming along wires on electricity, and you can make it go to lots of different…

Kevin Ball

absolutely. Yeah, the theory of constraints and bottlenecks has absolutely seen a resurgence with this stuff. Something you were talking about, you know, in making the electricity comparison reminded me of Carlota Perez's book about technological revolutions and financial capital. And in that book…

Sebastian Mallaby

Yeah, I agree with your framing very much. And back in, let's say, January of this year, 2026, I had no trouble saying flat out there is no general AI bubble. I really was comfortable in stating that very confidently because the demand for GPUs was just going straight up and data centers were being…

Kevin Ball

Exactly. Yeah. Once you're levered, then margin, you don't need that much of a dip to get this self-reinforcing margin call cycle going on.

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.
“These guys left OpenAI, the seven founders of Anthropic. They left in 2020. They start the company in 2021. And at the time, they're way behind. They don't have a big source of funding like Google DeepMind did because of Google's enormous balance sheet. They didn't have the head start that OpenAI had … And yet it turns out that by 2026, these guys are at the forefront scientifically. They are the biggest revenue, fastest revenue ramp, I think, in the history of capitalism.” — Sebastian Mallaby, TechStuff

The host asked Mallaby what his biggest question is as he starts a book about Dario and Daniela Amodei. He frames Anthropic's rise as improbable given its late start, thin funding and early setbacks.

TechStuff · 2026-10-30 Permalink → Listen →
TechStuff Around 10:20 into the episode
Nitasha Tiku

…Think of it as being, you know, part of this momentous period in time. They talk about how each of their employees is worth $ 6. 6 million. Like that's how talent dense this organization is. And they also go through some of the existential issues risks, which they touch on in the interview as well.

Oz Woloshyn

Sebastian, I mean, I know it's just been announced in Publishers Weekly, and it's sort of probably a quiet period, so to speak, on the book. But what's your biggest question about Dariam Ali going to be?

Sebastian Mallaby

…you know, enormous balance sheet. They didn't have the head start that OpenAI had. OpenAI then goes on to do the chat GPT, goes viral in 2022, and Anthropic seems even more behind. They scrounge together some money miraculously from the effective altruist community with Sam Backman-Fried at the center of that. And then he goes to prison. I mean, it's like the setback after setback after setback. And yet it turns out that by 2026, these guys are at the forefront scientifically. They are the biggest revenue, fastest revenue ramp, I think, in the history of capitalism. I would go maybe one notch further than Natasha did. I think it is the biggest, fastest ever. And then also in policy terms, you know, they stick their necks out. They take principal stands. They take on the Trump administration. And they make the weather on the kind of policy and safety front. So I think there's a kind of triple here, right? The policy stuff, the business thing, and the…

Oz Woloshyn

It's so crazy how the tech news cycle is nowadays because Microsoft announced they've offered buyouts to 60,000 employees, more than 25% of their entire workforce. It's probably like the fifth biggest tech story of the week. But I mention it just because obviously the other person we're going to be…

Sebastian Mallaby

Yeah, I mean, I haven't followed that specifically, but it's part of a big thing where, you know, when you're trying to onboard a lot of artificial intelligence. And I use that word deliberately. When you onboard a human, you have to kind of figure out how to use that person, train them and so…

Oz Woloshyn

Um, and so they can't write the essays, but they can write their perspectives.

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.
“in an organization, there are bottlenecks and the whole organization moves forward at the speed of the slowest piece of it.” — Sebastian Mallaby, The Spillover

Sebastian Mallaby explains why AI coding tools can double an individual programmer's output without doubling a company's overall productivity. Approvals, sign-offs, and other slow steps downstream of the code cap the gain the whole organization actually captures.

The Spillover · 2026-08-18 Permalink → Listen →
The Spillover Around 25:50 into the episode
Sebastian Mallaby

…good side about the US story is that when you have a big technological shift, you have to do radical things within the organization in order to kind of reap the productivity gains. And trying to kind of band-aid the whole thing just may slow you down and you wouldn't get the productivity benefits.

Rebecca Patterson

…been a couple courts that have presented guidelines, and there have also been a few government statements suggesting companies have to balance any automation with protecting jobs. So they're clearly aware of it, but I'm not sure where they fall in that spectrum kind of between the US and Europe.

Sebastian Mallaby

…describe the way that individuals within companies, like a coder, for example, has probably increased productivity thanks to AI a lot because you have a coding assistant and now one coder produces twice as much code. So that's a 2X on that individual. But organizations don't increase their productivity anything like 2X. It's more like, for the sake of argument, 1.1. Why is that? It's because in an organization, there are bottlenecks and the whole organization moves forward at the speed of the slowest piece of it. So let's say you have lots of coders and they're producing twice as much code and it's really great. It's really fast. But in order to actually ship the code, turn that into a product, let's say it's a bank and it's some kind of consumer-facing thing on the website of the bank, which the retail customers will use, then you're going to have a few managers, a few product managers, a few sort of…

Rebecca Patterson

Yeah, I agree with you 100%. And I'm hearing that anecdotally from different industries. And in addition to all the points you made, I would add the risk management layer, right? Even if your coders are twice as productive, churning out great things that could revolutionize your company and…

Sebastian Mallaby

Yeah, yeah. And there are some companies which literally their role in life is to sort of be in a liability sponge, right? If you're a law firm advising the investment bank and doing kind of the detail of the contract on some IPO or something, you have to diligence every sentence in the document…

Rebecca Patterson

Yeah.

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.

Collections They Appear In