Better Offline - The Hidden Recession Beneath The AI Bubble w/ Paul Kedrosky

Episode Date: September 23, 2026

In this week’s Better Offline, Ed Zitron is joined by economist Paul Kedrosky to talk about how AI is distorting the bond markets to the point that it’s changing the price on US Treasuries, how AI data center economics don’t make sense, and the troubling signs he’s hearing out of Anthropic. https://paulkedrosky.com/The Nick, Dick and Paul Show: https://www.youtube.com/channel/UCFbDiETo29GTIjg6Lk4imig Save $10 off a year of my premium newsletter: https://edzitronswheresyouredatghostio.outpost.pub/public/promo-subscription/gzqwkv54e1 YOU CAN NOW BUY BETTER OFFLINE MERCH! Go to https://cottonbureau.com/people/better-offline and use code FREE99 for free shipping on orders of $99 or more. --- LINKS: https://www.tinyurl.com/betterofflinelinks Newsletter: https://www.wheresyoured.at/ Reddit: https://www.reddit.com/r/BetterOffline/  Discord: chat.wheresyoured.at Ed's Socials: https://twitter.com/edzitron https://www.instagram.com/edzitron https://bsky.app/profile/edzitron.com https://www.threads.net/@edzitron Email Me: [e-mail] omnystudio.com/listener for privacy information.

This is an I-Heart podcast. Guaranteed Human. Hey, everyone, it's the Jonas Brothers with the Hey Jonas podcast. We've been catching up with some great friends, Paul Rud, Michael Bubet, Seth Myers, and making some new friends while getting into what's really like to have sisters, like Alex and Asgen-Nerl. She likes to control everyone and tell everyone what to do. Okay, well, control is a harsh word.

Plus, we've been taking your calls, listening to your voicemails and giving advice. Listen to Hey Jonas on America's number one podcast network, IHeart. Follow Hey Jonas and start listening on the free IHeart Radio app today. Hey Jonas is sponsored by edible arrangements. Order yours today at edible.com. Didn't catch the latest Roland Martin unfiltered podcast? Here's what you missed. People wake up and go, oh damn, wait a hold up.

They change all of that? Yes. It's real. This is a wholesale attack. It is targeting black people in every federal agency. It's raw. White folks have never allowed that reckoning to last more than a decade. Catch Roland Martin's daily commentary on the Black Information Network.

And download Roland Martin unfiltered. the IHart radio app, Apple Podcasts, wherever you get your podcasts. Orzo Media. Greetings and salutations, I'm at Ditron, and this is Better Offline. Better Offline! We are back, folks. We're back fellas and ladies and everyone in between. And today we are joined by the incredible economist and writer Paul Kudroski, Paul. Good to have you back.

Hey, Ed, good to be back. So, just before, literally seconds ago, you were telling us. me that AI has started to show up in odd places. And I think that's just a good, a good place to start. Where are the odd places in the economy it's showing? Because I can kind of guess, but yeah, yeah, yeah. Yeah, like, it's funny because almost a year ago, one of the things that first got me interested in this bananas topic was that it was, uh, AI Cappex specifically was, you know, blah, blah, blah, more than half of GDP growth, blah, blah, and so on. Which was just, you know, unprecedented and goes back to the railroads.

everything else. And so that was interesting, and I started talking about it, and then I was surprised that no one else was sort of noticing this, that how consequentially it had become, and now, of course, that's become commonplace. And so then I went on, and I've been paying attention to that, and that hasn't varied very much from quarter to quarter over the last eight, sort of, I don't know, 12 to 18 months. It's been consistently between 30 and 70% of GDP growth. And then it started, it was interesting how it's almost in cancer term, sort of metastasized across the global economy and become even more consequential. So the next example was about a month ago, where all of a sudden we had this treasury market freak out in the 10-year.

One of the reasons driving what happened was this sudden spike in the 10-year, up over 5% now. And that's the 10-year treasury bond, right? That's right. And when you say spike, it means it was selling off so people were asking for more interest on it. Correct. People acquired more compensation.

Yeah, sorry. I go all finance. No, no, it's fine. Keep going. Yeah, yeah. So normally the treasury market, which is a multi-trillion dollar market, is so large and liquid that unless you are a sovereign, meaning another country issuing debt on the scale and with the same sort of security as the United States, you really, it's really hard to influence it just through issuance, meaning that I'm just because I'm selling treasuries or I'm coming, or not treasuries, I'm selling comparable duration debt. It's really hard to have an impact on treasuries because it's such a big, large and unusually liquid market. So one of the things that happened a month ago was that this thing we call AI CAPEX

and this staggering amount of issuance, much of which has a duration, meaning the length of time over which the bond becomes due, the debt becomes due, is on the, it varies from five to 15 years, but let's call it a median duration of 10 years. It began to seem like it was something unusual was going on in treasuries,

and as we sort of work through the math, and then this is now, fairly widely accepted, that AI CAPEX issuance was actually had become so large that was beginning to bleed into the treasury market. And so one of the reasons why we had this treasury market freak out is because of the unprecedented amount of AI-related debt being issued, and that in turn was causing a sell-off in treasuries.

As people said, you know what, given the choice, and this is a remarkable thing, given the choice between owning 10-year treasuries, backed by the full faith and credit of the United States, or owning 10-year debt backed by a hypers scaler's pristine, and I have that in air quotes, credit rating, I'd rather have the latter at the margin. That just doesn't happen.

I mean, there's no reason why someone historically would do that because obviously sovereign debt is backed by the full faith and credit of the country. And in particular, in the case of the United States, you can print your own currency, so it's like there's no prospect of default other than deflating it away. So having that happen was an artifact of two things. One was the crazy scale of issuance, which is now approaching in excess of a trillion dollars.

Jesus Christ. Like all AI debt is approaching the trillion. And it's more than 60% of AI financing is now debt financed up from something like 15 to 20% a year ago. And it's now the largest piece of the investment grade marketplace. It's now the largest piece of the high yield marketplace. It is literally taking over global debt markets. And we had this unprecedented phenomenon, literally a month ago, where it caused this, in part, caused this global bond freak out in the treasury market because the scale was so large.

And the counterparties, the hypers, for crazy reasons, were seen as such secure credits that it began to literally compete with the most, the highest scale sovereign in the world's debt issuance, which is the United States. And that was a, I think it's staggering. currency of the world. The currency of the world was that AI CapEx as a currency was beginning to compete with the dollar in a really sort of loose sense. So that was the next one that really caught my attention because that has consequences because, you know, in turn, as the cost of financing the U.S. deficit rises, then or and its existing debt, the cost of servicing it rises, then obviously that has consequences in terms of the United States' fiscal position. and so you're having this crazy bleed over from what we see going on in AI CAPEX into the fiscal health of an entire country and by an intern of the entire world. So that is just a remarkable moment.

And then it gets crazier, right? And I'll stop on this next one, and then we can sort of loop into it in whatever way you want. But just last week, the WTO put out a report saying, whoops, our bad. We completely screwed up with respect to trade data. on the good side, not the services sides, because the WTO, the World Trade Organization, tracks goods and services, global trade data, because they had been forecasting a fairly sharp decline in part because of the U.S. tariff policy and the consequences of it,

both in terms of uncertainty and changing trade flows. And actually the opposite happened, right? And so it was up something like 4.5% or something like this, which was pretty remarkable. What does that mean in real terms? So this means that the global economy, in terms of exchanging goods between countries grew almost double the rate that was forecast by the WTO, the World Trade Organization, in anticipation of the effect of the tariffs that were coming through. So now they broke down the data because like any good economist, they said, so, you know, we were wrong, why? And it turned out that something like 19% of global trade in goods in the last three quarters was AI-related.

that's nice. Jesus Christ. It gets worse. That's Nvidia GPUs. That's right. Which goes back to your favorite story. So that's Nvidia GPUs. And then if you take it the next step,

well then how much of growth? So if 19% was AI-related, how much of the growth that was anomalous? Almost 55% of global trade growth was AI-related. So in a sense, we created this global now, illusory phenomenon that's being driven by this incredibly anomalous spending, and much of which

is concentrated on one specific thing, these things we call GPUs, but it's bleeding over into sovereign debt, is bleeding over into the cost of finance and the U.S. deficit, it's bleeding over into world trade. It's masking the effect of the Trump tariffs because it's so consequential. So it's having all of these incredible effects that are rippling across the entire economy. I'll give you one last one, which is one of the craziest ones. So you probably saw last week the Fed raised rates because U.S. inflation remains stubbornly high in excess of 3%. So let's break that down because, you know, it seems like, okay, fine, consumers must be spending a lot. And, of course, the answer is that consumers are actually more nervous than they've been in a long time and sit with relatively high levels of debt.

And so it's like, well, wait a minute, where is this inflation coming from? So if you break it down and sort of tear apart the pieces of this, you find out that the two largest pieces that are driving inflation in the United States over the last year are, well, you'll sort of sing along with me on the chorus, is AI CapEx, which is driving energy prices higher, which is driving all sorts of externalities across the economy. It's driving all sorts of services higher, construction costs, materials.

All of those are being driven higher. And then this is the sneaky one. The other piece is obviously energy. in the context of the war in Iran, which is the other big piece. And you can argue, and I did last week at a thing I was doing in Oxford, that in a sense, that's also AI-related.

And the reason why that's AI-related is the Trump administration was emboldened by its success in Venezuela, which I call the first AI war. Pete Hague-Seth and others have been very straightforward in saying that that entire campaign was organized and administered using Anthropic,

which is one of the reasons why They had a seizure when Anthropics said you can't use this for military purposes. And emboldened by their success using AI to conduct a campaign in Venezuela, they said, what's next? So Iran was next, and they've obviously used AI extensively in conducting that campaign. So it's created a kind of arrogance about what's possible whenever you use AI to backstop all sorts of things. So in a perverse way, now let's go back to this inflation data, both pieces of inflation. have underneath them as predicates this incredible and unprecedented

growth in AI, whether it's the energy side directly and the war in Iran, which is AI enabled. We've almost turned war into an app and on the other side what's happening in AI CAPEX. So here's, these

examples for me are just startling in terms of understanding the breadth and scope of this phenomenon and the consequences across the entire global economy. So let's start the top of that. So the reason that U.S. treasuries are going up, which means that the, and just to explain to the audience and stop me if I mess this up, when people say that the U.S. 10 year is going up, it means that people are selling off treasuries, which means that if you want to,

if you have a treasury that offers 5%, but the market wants 5.5%, that treasury you have at 5% is worth less, right? That's correct. So that's for on the run, for bonds that have already been issued. So it obviously has a similar consequence for new issues. It's the effective yield. So, yeah. So just so I understand. And if you're issuing new debt, obviously it has to be at the higher rate. So either way.

Yeah, because you're competing with that, with what the market demands. So just so I understand, hyperscalers are issuing so much debt that people are choosing it instead of US government bonds, or at least saying, US government, I need you to compete with this. Yeah. So that's exactly right. And that's in part because the structure. of the treasury market, and broadly, the U.S. debt market has changed a lot over the last 20 years. It used to be dominated by large sovereign purchasers like Japan and China and others.

And they were, in this great economics term, they were relatively yield insensitive. Meaning that they didn't get a rat's ass what the yield was. They just had to, they had huge dollar flows coming into their country, and they had to do something with them. So they bought treasuries. And it was like, whether the rate's one or five, I don't really care. I just have to do something. something with all these dollars. And now the treasury market, and broadly the U.S. debt market, is much more taken up by yield-sensitive investors, individual consumers, by people buying treasuries, or having embedded in ETFs, by corporates, by sovereigns, or sorry, by institutions and others. And they care what the yield is.

So now you have this phenomenon where people are yield-sensitive and they actually care. And they say, you know what? At the margin, given the choice between owning hyperscalor debt at the same duration, with a balance sheet, they argue, or owning treasuries, at the margin I'd rather own the hyper-scaler debt, and up goes the yield on the treasuries. Right. And that is, and I'm guessing that some of that is also not just a reflection of, it's not a reflection of the U.S. government's solvency so much as these people believe that U.S. treasuries won't pay as well over that period.

That's right. So that's the mistake people make exactly right there. Right. Where they misunderstand the causal link here, and they think, oh, this is because people have suddenly lost faith in the U.S. ability to pay its debt. The U.S. prints dollars. They are not going to default. They may deflate it away, but don't confuse the issue by introducing the idea that somehow

because of some madness the U.S. is doing, that its debt is more suspect. This is literally an artifact of competition at the margin from hyperscalar issuance in a newly yield-sensitive market causing people to say, I'd rather own X than Y, and that has sovereign consequences for the US has an increasing fraction of its budget gets eaten up every year by payments by payments

and in part because of AI issuance. And the thing is though, this kind of, doesn't this have a kind of a doom loop to it though? Because the US government as a result has to pay more on their interest. They have to, and interest is one of the US government's largest line items behind what like Medicare, Medicare, which it should do, military which should do less of. And I forget what the other. It's like social security in the other one. But the point. Social services broadly in defense, Medicare, and sort of all of those are the three big buckets, but behind all of those very closely,

depending on the math, comes in something like in the mid-teens, interest payments. And that will keep going up, but as that interest rate goes up that the government is charging, future hyperscaler debt will be more expensive, which will get issued at a higher interest rate, which will push the US one up.

Yeah. Am I reading this correct? This doesn't sound good. No, and it doesn't, it doesn't even, it's a double-locking doom loop. I'll give you the other piece of it. So, if you're the government and you say, okay, you know, this has consequences for us in terms of increasing current and future projected debt payments as a result of competing at the margin for these yield-sensitive investors, what do you say to yourself as a solution?

Well, we need to grow the economy faster because that'll solve the problem, obviously. And how do I grow the economy? Well, I increase productivity. Well, how do I increase productivity? well, people tell me this AI stuff is really good. So there's your double loop. So now all of a sudden, AI is not just the problem, it's the solution. I have a really dumb economics question that I think my listeners will appreciate.

How is productivity measured and what does it mean? As far as on an economics level. So it's literally just the input over output. Or sorry, output over input. So if we look at how much is being produced per unit labor and capital, and how that's growing, So we have a measure of how much labor and capital are going into the economy, and we look at what the output is that we're getting for that in GDP terms.

There's productivity. So it's really just a simple... What is the labor measurement? So the labor measurement goes back to sort of what the size of the labor force, hours worked, all of the BLS aggregate data plays into that. So it's a good question, though, because there's two ways, obviously, well, multiple ways to increase productivity. People can become more productive, but I can also.

also introduce a lot more capital into the economy and replace people and also become much more productive. There's multiple ways to increase the productivity numbers. And so you have to be very careful because people keep citing data showing that the U.S. is becoming more productive. But it doesn't have very much to do with individuals becoming more productive. It's because we're increasingly replacing individuals with capital. And that too shows up as higher productivity in the economy because you're producing more. doing, are we actually doing that, though? Because I've seen data that says productivity isn't really growing, at least not. No, no, no, it's not. But it's, there's, every once in a while,

you'll see a blip. And then, of course, instantly people seize on it and say, ah, there it is. It's showing up in the economy. And generally speaking, that's right now, that's almost entirely an artifact of capital spending, not labor. And so it's not showing up on the labor side. It's showing up on the capital side. And you see this all the time. It's classic, in behavioral finance terms, it's classic motivated reasoning. I see some data. I have a bias. This proves my point. It must be right. And so that's what's happening. And, you know, we see this in healthcare data. There was an Andreessen Horowitz graph making the rounds yesterday trying to show that mortality was sharply decreasing in U.S. hospitals in the ones that were, you know,

adopting AI the fastest in which is a ridiculous claim, right? And you didn't... I saw those, too. That made me very angry. Well, there's so many issues with this kind of thing and we can get into it, but it's in part has to do with what is AI spending, what are they spending on? Well, they were spending on AI scribes. Well, which hospitals spend the most on AI scribes? Well, the wealthiest hospitals spend the most on AI scribes.

Well, what's happened also inside of the wealthiest hospitals? Well, they have the best outcomes and the healthiest patients. There you go, voila. We just found another way to filter hospitals and say, it's really just a proxy for hospitals that have the, you know, have the healthiest patients and the wealthiest balance. Right, because they have the largest discretionary budgets. Right, yeah, yeah.

That's exactly what you would expect to have happen if you found a new variable that filtered hospitals for healthy patients and wealthy hospitals. And so it's this kind of stuff, going back to the other point, though, is these are all examples of motivated reasoning. And you have to always be very careful when you find, you know, a data point that proves something you believe fervently or even religiously because, you know, you're so desperate to believe it.

And you see this in the productivity data, you saw this in the A16Z Health Graph host talking about. I mean, it's just sort of everywhere. The music you love. I heart radio. Your favorite radio stations. Everywhere you go. The podcasts everyone's talking about. All in one place.

And it's free. I heart radio. Whatever you want to hear. Listen anytime. Anywhere. Music. Radio. Podcasts.

Unlimited audio. It's on the free IHurt Radio app. One app, no subscription. Always free. Grab the free app. I heart radio. Everywhere you go. Always free.

Download the free IHard Radio app today. Hey, everyone. It's the Jonas Brothers with the Hey Jonas Podcast. We've been catching up with some great friends, Paul Rud, Michael Rube, Seth Myers. And making some new friends while getting into what's really like to have sisters, like Alex and Ashton Earl. She likes to control everyone and tell everyone what to do. Okay, well, control is a harsh word. Plus, we've been taking your calls, listening to your voicemails and giving advice.

Listen to Hey Jonas. on America's number one podcast network, IHeart. Follow Hey Jonas and start listening on the free IHeart Radio app today. Hey Jonas is sponsored by Edible Arrangements. Order yours today at edible.com. Didn't catch the latest

Roland Martin unfiltered podcast? Here's what you missed. And the argument is, do you want to win? If you don't win, you have no power. It's real. Those are the very people who I always say

vote against their own economic interest and they complain about the very thing to receive it's raw i knew that the issue of transgender athletes was not going well with a lot of black men roland martin unfiltered i went off i said this is exactly what i'm talking about y'all gonna scream at chenel but you don't want to see a black luxury company in the same vein catch roland martin's daily commentary on the black information network 12 seats those are moving south so if you democrats You don't wait to 2030. You better be hitting the ground going hard in the South now.

And download Roland Martin unfiltered on the IHart radio app, Apple Podcasts. Wherever you get your podcasts. I have learned this lesson myself, like a few years ago. Like in 2024, there were so many times where I made this mistake, where it's just you you find a data point, you hope it proves everything. And these days, while I do absolutely use data points as I kind of did with the AI Kappex not being used thing, that was at least, hey, you can point a construction in progress and say, oh, you have hundreds of

billions of stuff not being used versus just saying, oh, productivity data's up. That must be AI, or it's down. AIs there. Yeah, yeah, yeah. It's right. Seizing on these data points to prove something that you're, you know, you really want to believe. I'm always very suspicious of anything when I feel almost an emotional triumphalism. When I see something, I say, aha, this is exactly what I thought. And I'm like, okay, you know what? Let's be a little, let's cool down a bit here because this probably is an emotional, motivated reasoning response. Anyway, so across the board, you see this now. So going back to the original point, it's now in the Treasury data.

It was in the originally in GDP data. And so just on the inflation data, I'll take it one step further, which is even more, which is even crazier. If you back out the AI Cappex related inflation, and if you back, grant me that energy increasingly is being driven by this campaign that was so AI enabled and thus AI has a role in it, back those two pieces out of the U.S. economy. When you say, and sorry, I hate to do this,

when you say energy in this case, what are we measuring? Okay, so there's two, there's, well, multiple components to this. So the inflationary component from the standpoint of U.S. consumers is utility prices and gas prices. So let's not get into the embedded price of gasoline or energy in terms of how it plays into the products you purchase, but that obviously is another issue because it shows up in shipping and other places. but the direct one that people feel most consequentially

are prices at the pump, whether you're an individual with gas or a trucker with diesel, and then obviously utility prices, both pieces of which have gone sharply higher over the last three years, in part driven by either competition for energy from the standpoint of data centers or the Iran War. So these two things, and I'm arguing that the Iran War

is in part of function of the sort of emboldened U.S. government treating everything as kind of an AI-enabled app in the wake of Venezuela. So if you grant me that, then what you find out really quickly is actually the U.S. economy, far from being very strong and actually growing so strongly that it's justified for the Fed to raise rates to try and choke off inflation, back those pieces out, and by my math, the U.S. is actually in a deflationary mode, about a quarter of a percent. So what we have happening is something very analogous to what happened in the walk-up to the crisis. crash of 29, which is you had the Fed raising rates because it misunderstood what was actually happening in the economy, and it thinks it's choking off inflation. But the inflation is

actually something exogenous, something different than what they expect. It's being driven by this CAPEX and by this anomalous war. So as a result, as this unwinds, and it will unwind, the war will end and CAPEX will taper off. You actually have high rates in a weak consumer environment, which is a classic precursor to what, there's an economist named Richard Koo, who's written a lot about the Japanese 15-year recession that sort of went on and on. People called it the happy recession. Yeah, yeah, the lost era. So those are classic precursors to those kinds of recessions because you now have companies

that have loaded their balance sheet up and will take a decade for them to unwind all this debt. So rates will have no impact on them. They'll simply be trying to cut spending to reduce debt, and that will play out over next 10 years. We've got all the pieces in place right now for a very long balance sheet recession analogous to what happened in Japan in its lost decade. And what does a balance sheet recession look like? Because this is actually something I've been thinking about a lot where, okay, you bump up interest rates. But when a lot of the reason the interest rate, like,

when inflation is happening in one specific area, specifically to companies that are not interest rate sensitive. Let's be very clear about this. Like soft bank, SB Energy is trying to raise at 10%. Yeah. Like just what are you doing here? Like core weave raises at 9, 10%. And the hypercalers was, yeah, six or seven percent. They are going to keep raising them out of what. How, like, what does, what does that lead to? What is this balance sheet recession? So what happens is, and leaving aside for a second, some of the more, how shall I say, less investment grade prospects out there, the core weaves and others because they face a different problem, which is as they try to roll over their debt, it may comment on terms over the next five years that force them into some species of insolvency,

and that's fairly typical. But let's say you're a hyperscaler of some of the largest companies out there that don't face solvency risk. They're not going to likely go broke. What they're going to spend the next decade doing is cleaning up their balance sheet, because they just went from some of the least indebted companies in the market. Tech historically had very low levels of long-term and short-term debt to some of the most indebted companies in economy. And we see this as they sort of soak up all their cash flow and sort of in the case of some of these guys increasingly teeter on the edge of having their credits rating downgraded. And so what a balance sheet recession means is essentially I'm having to put myself into a kind

of like a debt workout and say, I'm going to spend the next 10 years getting all this debt off my balance sheet again because it doesn't need, I don't want to have it there anymore because it's no longer required for, you know, ongoing purposes because I'm not spending that heavily in future and I'm being punished for it because I have this huge debt obligation. So think about the consequences. You're not spending on growth. You're not spending on employees and you're trying to unwind all this debt and get it off your balance sheet. So a balance sheet is recession is a workout where heavily indebted companies, irregardless of interest rates, continue to try and get themselves less leverage, get the debt off their balance sheets. And so the

problem is, of course, this is going to come at a time when there's already other pressures on hiring, the AI effects at the margin. So the effects overall on hiring will be even more dramatic because companies will be focused entirely on de-leveraging. So at the same time as they're looking for opportunities to use AI to have fewer people around. So this will be really consequential over the next five years, and it's really not well understood

that we kind of sit on the edge of a Japan moment in terms of the likelihood of a balance sheet recession given that there's this really profound misunderstanding of what the causes are of inflated, in the economy and what the effect that rates are going to have and they're actually sitting in a kind of deflationary moment masked by what's happening in AI. How does that manifest for regular people? So for regular people, your expectation should be that over the next, say over the next year or so,

we'll start seeing increasing signs that the economy is much weaker than people expect. And as a result, we'll probably be heading into a relatively longer recession, which is inevitable after a moment like this, a longer recession that people expect that with a few policy errors could easily be something on par with what often is called a depression. So a recession that turns into a depression,

which means multi-year and really consequential in terms of companies themselves becoming insolvent as they try to work off debt. So the problem, of course, is normally when that happens, the government steps in, in a Keynesian way, and tries to replace the lost consumer spending with increased fiscal spending. And the problem, of course, is that most,

and there was a great IMF report out this week showing that most countries around the world are basically all fiscal stimulus out because we've gone through multiple episodes over the last 15 years. And so there isn't a lot of headroom anymore to bail out your own economies as we sort of approach this inevitable moment.

So your expectation as a consumer should be that the so-called, you could call it the Bernanke put, this idea that Ben will save us, and then it was the Greenspan, I suppose, in the other order. And that just isn't a possible here. It's not possible here. Right. And in particular, because fiscal policy is being increasingly rendered impossible

because of the indebtedness of major sovereigns around the world. And monetary policy rates, as we've been discussing, is almost impotent because of the nature of what's going on. So you have this double whammy problem. So it's just, so there'll be, I feel like there's one thing with your, your idea here that I'm pushing back on, which is, we haven't seen the AI side actually in productivity. So it just kind of feels like everything will get more expensive, more people will get laid off, and then nothing will happen. Like, it's not like there's replacing them with AI.

Yeah, yeah, yeah. So it's, you have option A and option B, and both of them are bad, right? So either one leads to whether companies are just experimenting madly because they feel they must, and they discover that there's the productivity benefits aren't really there, or they find some productivity benefits, however modest, and they use that as a justification for further cuts to work for it. The consequences economy-wide are the same. You're into a multi-year recession, and then because of this balance sheet problem,

something that teeters on the edge of being a depression, or at least a decadal workout like the happy recession or the last decade in Japan. So what finally stops PiperSk? as borrowing, because at this point, it kind of feels like they don't care, but at the same time, like, where are the limits on this? Because I've seen like $300, $400, $500 billion thrown out as how much they'll have to raise in the next year or two. Yeah, no, that's for sure. So there's physical limits and there's financial limits. So physical limits, obviously, we're already seeing. There was a great piece in the FT the other day talking about how, and it's so loopy to me, to use the technical

term, like how dramatically short we are in terms of the amount of power required to power up, all of the GPUs that are currently sold and forecast to be sold. And you've been really good on this showing this is just something doesn't make sense in here. There's a dramatic disconnect. There is a really big disconnect. It actually genuinely worries me. Yeah, and I think you're spot on with that. And there's all kinds of reasons for what's happening there. But the math doesn't work in terms of our ability to power up data centers on a scale commensurate with the numbers of GPs being sold in the data centers. I won't say built, but at least spec.

And so there's a natural physical limit in here. But the system is the issue that I look to. So the system is telling you through what's happening with inflation, what's happening with the 10 year, that it's under immense stress. The system is under immense stress. Right. So that means the interest rates are high. So when the interest rates are high, sorry, I've got to be, for my sake and listeners. When the interest rates are, high, that is the system kind of screaming a little. Yeah, the system's screaming at you and saying,

I don't know where this money is going to come from, so I'm going to have to raise, the system is autonomously doing this. There's no individuals out there. There's no one, like the Von Traff family or something. There's no one actually doing this and saying, I'm going to raise it. So the system is raising rates because it says, you know what, I don't know, I need to bring money out of unusual places because the only way that we

can justify or get enough capital for what you say you need is by having rates go higher and pull in capital from other places. And so the systems, that's really just a sign of stress because what happens next is the hurdle rate. The amount that the data centers have to earn

goes up constantly as a result of the rising tenure. The tenure has gone up like 100 basis points over the last, so 1% roughly over the last six months, or it's been last three months, which is crazy. So that it directly translates into the funding cost for new data centers, but it doesn't stop

there because if the tenure goes up 100 basis points, The spread goes up even more, meaning that it's not just that I now have to put out my debt at 100 basis points higher. The system recognizes that this is a stressful moment, and so the spread between the rate at which you raise and the rate on 10 years, which is often something you use to anchor the cost of financing, might go at 150. It might go at 200 basis points because the system's under stress. So every time you see rates increased, say to yourself instead, oops, you just got a lot more expensive and difficult for data. to justify the future cash flows on top of which they're building these data centers. And so, to answer your question, we're within six to 12 months on the outside of that aspect

of it all breaking, because it's very clear that they're not going to earn 11 and 12 percent, which are the kinds of numbers we're seeing at the most stressed end of the market, 10 plus anyways. That's not going to happen. There's no economics that does that. But there's also another problem, which is for the hyperscaler debt that's already been sold. So the bondholders are now effectively making less money on that, as in their bonds are worthless. Because there's new debt that's paying better. So anyone who's invested, anyone, any bond investors who have invested earlier will probably hold them to maturity, meaning they'll get paid off. But the same time, their underlying holdings are worthless.

Yeah. Right. That's exactly right. But that in turn has, that in turn has consequences because that debt is collateral. No one just holds anything. That's used as collateral for, for other economic activity that I'm doing, whether it's leverage or other borrowing. And so as this debt becomes less valuable, it reduces my ability to other things in the economy as a functioning kind of economic actor. And so you have to always think about it.

I like to point people, there's some great books out there by people like Donna Meadows and others. Like, think about it in systems terms. You have to always connect everything to everything. And what happens whenever the debt becomes less valuable? Well, okay, I can't do as much. Well, what are the consequences of that? Like connecting everything for whatever reason is really hard for people to do.

And the system's telling you, it's telling you, it's screaming at you, whether it's the world traded data, the inflation data, the 10-year data, that the system is both under stress and wildly connected. And all of these pieces are, you know, reaching breaking points. But what makes it, what exhausts the system? Like, what slows this down? Because I, one of my moments of naivete was believing the good sense would take over. That's never happened in history.

I've now read more history and realized. I'm not a thing. Yeah. So what is it that actually exhaust this? Where are the limits of the money? I'm not asking for a time period so much as an event series. I say this all the time, but I think it's overdetermined, meaning there's so many ways this can break because of all of the ways in which the system is under stress, whether it's in trade flows, whether it's in the likelihood of a balance sheet recession, whether it's in the increasing hurdle rate on new data center debt issuance. All of these are places it can break. My guess is it breaks in somewhere, because this is the way these things tend to go. It breaks in somewhere unexpected, like we have a botched 10-year treasury auction. Wouldn't surprise me at all. By which you mean there is insufficient demand for it? That's correct, yeah. So they said, here's how much we'd like to sell. We can't sell that at anywhere near this price. And if you saw that, something like that happened, which would be an artifact of what's happening in AI data centers, the spreads on 10 years would blow out, meaning they would get much

wider, they would get much larger, and the rates would go higher, but it would explode your ability to raise debt at any economic price in the world of data centers, whether you're a hyperscaler or a junk issuer, it doesn't really matter. And so the system's telling you it's on the edge of that right now, and so that would surprise me zero, in part because people like Scott Bessender saying that could never happen, which for me is always saying. You never say never when it comes to this stuff. And I will say a friend of mine in fixed income has told me that, In the event you ever hear about the treasury, or sorry,

is it the treasury that sells bonds, right? Yeah, cool. Not pleased you there. If you ever hear about a technical issue, the delayed an auction, that's what happened. That's what I've been told. Like if there's it, yeah, if you ever hear about it being a tech.

Yeah. Okay. That's exactly right. So for me, that's the, that's the place to look because what's happened is, is this stuff has, as I keep saying, it's kind of metastasized outside the orthodox places. And tech bros are really bad at thinking in these terms

because they think exclusively in terms of it's Christmas morning and look, there's all these gifts for me under the tree and I don't really care how it's being paid for because that historically was a relatively defensible way of looking at things. I don't know what's going on. I don't know, fiber optic cables,

but all I know is it's now cheap for me to get streaming movies. The difference this time is that the scale is so consequential that it's actually distorting not just trade flows, not just inflation, but also the very, essence of what it is to be an independent sovereign. Yeah, it just feels like there is something, like it's not going, it's not going to be that they just borrow too much, it's just the system is going to exhaust itself somehow. Oh, no, no, no, the system will crack. There's no question in my mind that it'll crack,

it'll probably happen. It could start in Europe and come this way, but it's going to, it's going to be a sovereign issuance problem, where someone's going to have a technical issue with a major issuance that around probably is sort of a 10-year duration, and then that'll float around the world, cause spreads to blow out, which will have consequences in terms of the existing debt and the solvency of all of these people playing in this area. And again, not because AI isn't, you know, useful in many contexts or whatever else.

That's all, I won't even get into that debate. Just because the scale of the problem dictates this must happen. And the system is telling it, but people don't want to listen. The music you love. Your favorite radio stations. Everywhere you go. The podcasts, everyone's talking about. All in one place, and it's free.

I heart radio. Whatever you want to hear. Listen anytime, anywhere. Music, radio, podcasts, unlimited audio. It's on the free IHurt Radio app. Let's go! One app, no subscription. Always free.

Grab the free app. IHard Radio. Everywhere you go. Always free. Download the free IHR Radio app today. Hey, everyone. It's the Jonas Brothers with the Hey Jonas Podcast. We've been catching up with some great friends, Paul Rudd, Michael Boubley,

Seth Myers, and making some new friends while getting into what's really like to have sisters, like Alex and Ashton Earl. She likes to control everyone and tell everyone. Okay, well control is a harsh word. Plus, we've been taking your calls, listening to your voicemails and giving advice. Listen to Hey Jonas on America's number one podcast network, IHeart. Follow Hey Jonas and start listening on the free IHeart Radio app today. Hey, Jonas is sponsored by Edible arrangements.

Order yours today at edible.com. Didn't catch the latest Roland Martin unfiltered podcast? Here's what you missed. And the argument is, do you want to win? If you don't win, you have no power. It's real. Those are the very people who I always say vote against their own economic interest, and they complain about the very thing they're receiving.

It's raw. I knew that the issue of transgender athletes was not going well with a lot of black men. Roland Martin unfiltered. I went off, I said, this is exactly what I'm talking about. Y'all going to scream at you. Chanel, but you don't want to see a black luxury company in the same vein. Catch Roland Martin's daily commentary on the Black Information Network. Twelve seats, those are moving south.

So if you Democrats, you don't wait to 2030, you better be hitting the ground going hard in the south now. And download Roland Martin unfiltered on the Iheart radio app, Apple Podcasts. Or wherever you get your podcasts. It feels just impossible. Like they can't afford. all the CAPEX next year. Like, we've established that. So at some point, at some point, they just,

I mean, hyperscalers must have a point at which it's too hot for them. Like, there must be a point when it's too expensive. It just depends on how that it becomes. But the problem is, like, it's the Chuck Prince line, right? From back in the GFC, from the financial crisis, as long as the music keeps playing, I keep dancing. You can't stop, right? It's a classic collective action problem.

No different than the collective action problem. problem whenever Dario put out the memo a couple of weeks ago asking for a frontier pacing slowdown, blah, blah, blah, and leaving aside with you, you thought that was a good idea or a bad idea, you instantly found the system thrown into a prisoner's dilemma where everyone was, you could tell, everyone was busy trying to find an angle to de facto, because if we all cooperate, but I de facto, I make more money than anyone else, and you're right back into a race again. The exact same thing is true in the context of fundraising. If all of your peers suddenly stop raising money because they see this as a collective action problem and we're walking off a

precipice. Great. I'm the only guy left who's still out there raising prodigious amounts of money. I get to own the market. And of course, on goes the race because then everyone sees that and they all defect and were right back where we were before. So you mentioned, do you have any idea around what the payoff now needs to be on a data center to make it worth it? Like I'm what? Well, yeah. So the, I'm surprised they haven't sat down and done the math. Yeah. Yeah, so you just can do it from a yield standpoint. I mean, So the median is now around 7.2%. So you can do the math pretty easily in terms of it.

Think about it. What's that mean? Sorry. So that's what the hurdle, right, that these things are required to earn back to justify, to attract investors at the investment grade end of the marketplace. It's more like 10% at the junk. So it's like you need to earn 7% on the investment a year? Sorry. That's right. That's right.

So that's what debt investors are telling you they will take. So that's not the same thing as saying how much I have to earn to earn back the capital that's been spent. that's a completely different problem. I think it's unlike you could ever earn it back, but you can at least attract debt investors on those terms. So there's two different issues here. Yeah. So how much?

So you have to, so you basically, for a hyperscale, you need to make about 7% per annum. Right. Which continues to rise, but on that order of things, that's right.

And that's what that's a couple. Which people will say is, oh, that's really straightforward. You know, Gavin Baker's out there telling you, right, that the margins on these things

in terms of pure inference are so high, but of course the problem is that this is the classic phenomenon of, I call it like earnings before bad things. So if I cut out all of the things that are costing me a lot of money and then I calculate EBBT, not EBBT, but EBBT, earnings before bad things, then my cash flow is tremendous. Well, then fine, show me the hypers that have stopped spending on frontier models

and I'll buy the argument that the economics now make sense purely in terms of commodity inference. But if you go to the world of saying, okay, fine, industrial inference is actually a relatively high cash flow business, let's just grant them that. Then the question becomes, well, who's going to win that? Well, if it's just industrial inference, so then meaning that I'm just doing, you know, token completion sequences as low cost as possible and at the highest scale possible, that's really just an energy problem. So it's just who can throw the most energy at this.

China's throwing on an annual basis roughly three times, adding three times as much capacity per year as the United States is, and is already ahead of the U.S., right, in terms of the amount of energy it's thrown at this. So it's fairly straightforward. You're just looking at solar panels, but it happens to be tokens. So, but that's the thing. Even with standing up inference, because is inference profitable? It's like the Wario is a libertarian conversation online when it comes to this stuff.

It's like, and no one can really, I think it's unprofitable. But the reason I do is, is because inference isn't just, oh, I turn on the inference machine. It's you buy an allocation of GPUs and you need a certain level of saturation. a certain amount of customers to make it viable. And if you miss that demand calculation, it's horribly unprofitable. That is a huge problem. It's kind of like running an airline, right? Either I'm, you've got a very, very high fixed-cause business.

So either I'm fantastically profitable or I'm aft, right? And very few people are able to run it at a scale requisite to be able to generate those kinds of cash flows from covering the high fixed cost. And so this is the deep problem. And then if you take it the next step and say, if that is the nature of what's going on, and tokens are this globally fungible commodity, and so that's the market's moving towards industrial inference.

It's fairly straightforward to see who the winners are going to be, and it ain't going to be open AI in Anthropic. Yeah, it just, it feels so inherently doomed. But you know what? You said Anthropic. And before we got on, you were mentioning you'd heard some strange things, and I am now curious about how strange those are. Yeah, so I keep hearing these unusual things about what's going on over there. And I mean, obviously, the most unusual thing was the company was racing fairly madly for reasons that didn't make a lot of sense to an IPO, which was supposedly, I think, was supposed to be October 1st, was the first dates I heard.

Yes. And now, allegedly, and again, this is all just, you know, financial markets and tech market babble, but it's now allegedly in November. And so that's getting pushed out. And my general thesis is that even in a good year, if you say you're going to go public late in the year, odds are at least 50% it won't happen that year because lots of things can happen to toward the tail end of the year.

And of course, this one's coming after the midterm elections in the United States. Who knows whether or not any of these AI companies get nationalized in that period? So there's a forced equity stake. All of these things could radically change the likelihood of the company going public. So there's raging panic

inside the company about the nature of what an IPO this year would even look like, given what's happening out there. And you can take it further, because if, let's say I'm an employee, and this is something I'm hearing from people inside, take it further and think about it in terms of employees having exercised their options, purchased their options in anticipation of an upcoming IPO. And then the IPO doesn't happen. Well, what happens? Well, let's say it eventually...

You've just dumped all your cash into buying stock or buying into your stock. Well, you generate a massive tax. bill under the IRS has this thing for the AMT. So under AMT, you just bought something cheaply, your options, exercised them cheaply for a low price, and then you, in turn, have been generated, given something with value. It doesn't care that you never got any cash out of it, that the company never went public. So I know of people who take... That's right.

So I know of people, many people, who've taken out multi-million dollar loans to exercise their options and purchase them and then are going to be faced. if this thing doesn't go public this year with multi-million dollar AMT bills because the company didn't go public. So this has happened... Sorry, go ahead. Is it usual that a company...

This is not a leading question. Is it unusual or usual that a company delays their IPO what appears to be twice? Because they were meant to do it in September, then it's October, now it's November. It's uncommon.

In general, with a blue chip company. So a company that's seen to be a market key company that the entire sell side is, you know, beating each other up to underwrite. It's more likely that it happens on schedule than not. So this is really unusual. And so there's all kinds of reasons you might anticipate why, but it's very fraught as it gets closer to the end of the year, not least because of the, you know, the political consequences in the U.S. given that this thing is kind of a, you know, laser target, but also the AMT consequences for employees, this IRS AMT issue. And

And further, obviously, the consequences for the company itself in terms of whether or not the valuation even looks like the kinds of numbers that they've been looking at. Because, again, this is complete, complete hearsay speculation. But a vendor who's fairly close to the company was told me that they had at least underperformed their expectations in the third quarter. I have no idea if that's... Anthropic did? Anthropic, that's right. So they had underperformed their expectations. Now, again, this was not from someone at Anthropic.

This was from a third party who was with intimate knowledge of what the company was doing as a provider to them on a very large scale. I'm likely my eyebrows nonetheless. I'm like... Yeah, yeah, yeah. Which I thought, which is obviously really interesting, isn't necessarily particularly surprising in the context of the numbers we keep seeing out there

in terms of customer concentration. Well, also with token maxing as well. You had that era where everyone was blowing a bunch. That's right. That's right. So this incredible customer concentration among customers who were token maxing and who have increasingly cut back. And so whenever you have two or three customers making up 60 plus percent of your revenues and they were following a policy of token maxing or whatever else, it's not particularly surprising that you get this incredible pulse of revenues and growth at one point. And then you get a kind of air pocket later in the year. And that's exactly what it seems like we may be seeing. Yeah, and I mean, I don't know if you saw the SB Energy S-1. I did. Yeah. What a dog.

But also, they have now pushed their IPO. It kind of feels like we're at the latter stages. I guess I've been saying that a while. But this feels like when you've got multiple IPOs being delayed with weird stuff happening around them, like there was the anthropic story with the financial times they should know better, where it was like, oh yeah, they have 80% gross margins if you remove many of the, the EBBT problem. Erding's before bad things again.

So my general view on all of these things is, and it sort of goes back to this guy, Hyman Minsky, who talked about these sort of Minsky moments in financial markets, where financialization of the phenomenon takes over and detaches from any underlying reality, that's when things accelerate and then break. And that usually requires the public markets.

It usually requires people to be dumping stuff into the public markets, investors dumping stuff. And so that's the moment we're at now. A year or two ago, there wasn't a public market component to this in terms of directly related to the frontier companies themselves.

It was mostly obviously the GPU and memory providers. And so we're now in this moment where the companies at the center of all of this are in this kind of Minsky moment where the financialization of what they're doing is so large and so broad. It's affecting everything.

But at the same time, they're all very eager to dump the shares and get them out there at a credibly large scale to retail investors. And that's the moment where I think everything becomes, starts to crack and break because you realize that this is that financialization Minsky moment where everything breaks down. And you're seeing a little bit of it already with the delays and everything else. And I mean, I don't know if you saw it, but there was even back in September,

was it in September? I saw that there was at least one action. acquisition that Anthropic walked away from, that it was apparently in its late stages of trying to get done. That's right. So, oh, let's Descartes, right, yeah, Descartes. And so, again, not to overread into these things,

but what tends to happen as people approach IPO, they feel kind of omnipotent that I have like a printing press in the basement and I can buy anything and do anything. And when you see them start to pull back and think, you know what, that printing press isn't as powerful as I once thought it was. That's a really important tell about what they're seeing inside the business. this because the printing press in the basement doesn't look as all powerful as it once did. Well, as we wrap up, though, I wanted to talk briefly about this OpenAI non-IPO and the fact

they're raising at either one. Well, they want to raise at $1.5 trillion, which is amazing. So amazing. But what's great about it is apparently investors tried to preempt him and say, we'll give you 1.2, according to the New York Times. But then Open AI said, no, we want 1.5. Is that not, even 1.2 feels like it, isn't 1.5 kind of a suicidal private valuation based on what they'll get on the public markets? It's obviously impossible to know what demented things public markets will do if you float. And this is the key. It depends on the size of the flow

load bearing. Right? That's the thing. So if I do something like, I do a SpaceX and I float a tiny sliver of my potential public, eventual public issuance, you can get any price you want in terms of an eventual valuation. And so that'll be the key. It really comes down to how cynical Sam wants to be in terms of whatever he decides to float. Because at a small enough float, you can do anything. You can get, I can get, you know, any valuation I want in the public markets. Of course, what happens then is at a small flow, when I then bring the rest of the stock out, six months, the employees all leave lockup and all that stuff comes to market, well, then the doc, of course, promptly craters, or at least craters in anticipation of it. And so,

you know, my expectation is the higher the valuation we see today, the smaller the float will be, and more cynical the eventual IPO will be in terms of trying to manipulate the price. Yeah, and I think the, my thing is, as well as just a suicidal round for investors to put money in. Oh, of course. And because SpaceX is $2 trillion. Yeah, yeah, yeah. So you'll never, you have to think about it in terms of, if I'm a large institutional investor, not a retail investor. Retail investors, God bless them. But if I'm putting it like, say,

$50 to $100 million or something like this, some giant sovereign amount of money into this at this price, my expectation has to be that I'll be able to sell fairly quickly post-IPO at a significant bump-up. Well, fine, let's say I can get the bump-up in price by floating only a small amount of the stock.

You'll never be able to sell hundreds of millions of shares into that bump-up because there simply won't be the liquidity given the small size of the flow. So you'll be able to, paper say you made some money, but in practice, you're stuck. Yeah, it, this feels, it all feels like to round up everything, like there is going to just

be something that snaps and everything kind of falls apart, because these things don't tend to fall apart because, oh, I, like everyone decides to pull back for ethical or moral reasons. It's part of the system breaks and every, yeah, part of the system breaks and everything falls apart around it. Right. And this is it, because you can take it that next step and say, if it's this large of a P, if it's distorting inflation and it's distorting the tenure and it's distorting world trade, take that the next step. That's causing, and I use this term loosely, you know, intelligent policymakers, because I'm not sure those exist anymore. But it's causing policy to be distorted because if you think, oh, you know what?

Tariffs didn't have any effect on world trade or, oh, you know what, despite the U.S. consumer being very weak, the high inflation, so we must be. be something strong. You pursue all kinds of bad and self-destructive policies because you don't actually understand the mechanics of what's happening. And so as a result, the crackup becomes sharper. There's a sharper delineation between the now time and the after time because of this colossal misunderstanding of the nature of what's actually going on in the economy. And this wild unwillingness of people to connect all the pieces, even at the Fed. I mean, the Fed's recent commentary on economic outcomes is so narrowly siloed and so backwards looking and trying to reinterpret what's happening in the economy in terms that might have made sense 40 years ago, but they don't

make sense in the context of what we're seeing. So people really don't have a good handle in terms of understanding these drivers, and that makes this inevitable break, you know, much more consequential. And also means that people are going to be surprised. Oh, absolutely, right. Yeah. And And, of course, we'll have this crazy moment where even if you feel that way, the constant people are always so nervous about it. Well, what happens if on a Saturday I lean over the fence and my neighbor says he bought stock in this pre-IPO and it did so well? People are just congenitly unable to walk away from that and not feel like I missed out on something, this phomo problem. And so people, despite all of the things I'm saying, are going to walk right into this and then the crack will happen and it'll be a lot of who could have known. That's such a shame. Well, Paul, on that happy note, thank you so much for joining me on the show this week. Thanks, Ed, good to be here.

And everyone, you'll get me on a monologue this week, of course, on Friday. Thank you, as ever, for listening to Better Offline. Thank you for listening to Better Offline. The editor and composer of the Better Offline theme song is Mattosowski. You can check out more of his music and audio projects at Mattisowski.com. M-A-T-T-O-S-K-I.com. You can email me at EZ at Better Offline.com or visit Better Offline.com to find more podcast links and, of course, my newsletter. I also really recommend you go to chat. Where's Your Ed?at to visit the Discord and go to R-slash Better Offline to check out our Reddit.

Thank you so much for listening. Better Offline is a production of Cool Zone Media. For more from Cool Zone Media, visit our website, coolzonemedia.com or check us out on the IHeartRadio app, Apple Podcasts, or wherever you get your podcasts. Hey, everyone, it's the Jonas Brothers with the Hey Jonas podcast. We've been catching up with some great friends, Paul Rudd, Michael Bublay, Seth Myers, and making some new friends while getting into what's really like to have sisters, like Alex and Asgen-Nerl. She likes to control everyone and tell everyone that to do it.

Okay, well, control is a harsh word. Plus, we've been taking your calls, listening to your voicemails and giving advice. Listen to Hey Jonas on America's number one podcast network, IHeart. Follow Hey Jonas and start listening on the free IHeartRadio app today. Hey, Jonas is sponsored by Edible Arrangements. Order yours today at edible.com. Didn't catch the latest Roland Martin unfiltered podcast? Here's what you missed.

There should not be a single law enforcement agency. It does not have body cameras. It's real. Black farmers have been under attack. This is just the latest example of them just slapping DEI on anything. It's raw. And I'm sitting here going, they are playing y'all for fools. Catch Roland Martin's daily commentary on the Black Information Network.

And download Roland Martin Unfiltered on the IHart Radio app, Apple Podcast. Or wherever you get your podcast. This is an IHeart podcast Guaranteed human