Odd Lots - Why Paul Kedrosky Says AI Is Like Every Bubble All Rolled Into One

Episode Date: November 14, 2025

In recent weeks, there's been renewed anxiety about the sustainability of the AI boom. This is partly due to comments from OpenAI CFO Sarah Friar about a possible role for a government backstop in the AI infrastructure build out. We've also seen the stock market wobble, with many major tech names hit hard. But even with all these concerns, we continue to see new announcements all the time. Just this week, Anthropic said it would spend $50 billion on data center development in the US. So are we actually in a bubble? Our guest on this episode believes we are -- and not just any bubble. According to Paul Kedrosky, a longtime VC currently at SK Ventures, the AI bubble is like every previous bubble rolled into one. There's the real estate element. There's the tech element. And, increasingly, there are exotic financing structures being put in place to fund it all. And then on top of that, there's talk of government bailouts and backstops. In this episode, we walk through some of the math that would be required to justify all this spending, and how the seemingly existential stakes of 'winning the AI race' is causing an unsustainable investment binge. Read more:AI Startup Cursor Raises Funds at $29.3 Billion ValuationPoint72’s Drossos Sees AI Boom Driving Gains in Asian Currencies Only Bloomberg - Business News, Stock Markets, Finance, Breaking & World News subscribers can get the Odd Lots newsletter in their inbox each week, plus unlimited access to the site and app. Subscribe at  bloomberg.com/subscriptions/oddlots Join the conversation: discord.gg/oddlotsSee omnystudio.com/listener for privacy information.

The news doesn't stop on the weekends. Context changes constantly. And now Bloomberg is the place to stay on top of it all. Hi, I'm David Gurra. Join us every Saturday and Sunday for the new Bloomberg this weekend. I'm Christina Rafini. We'll bring you the latest headlines, in-depth analysis, and big interviews. All the stories that hit home on your days off.

And I'm Lisa Mateo. Watch and listen to Bloomberg this weekend for thoughtful, enlightening conversations about business, lifestyle, people, and culture. On Saturday mornings, we put the past week's, events into context, examining what happened in the markets and the world. That on Sundays, we speak with journalists, columnists, and key political figures to prepare you for the week ahead. Join us as soon as you wake up and bring us with you wherever your weekend plans take you. Watch us on Bloomberg Television. Listen on Bloomberg Radio, stream the show live on the Bloomberg business app, or listen to the podcast. That's Bloomberg this weekend.

Saturdays and Sundays starting at 7 a.m. Eastern. Make us part of your weekend routine on Bloomberg television, radio, and wherever you get your podcasts. Bloomberg Audio Studios, Podcasts, Radio News. Hello and welcome to another episode of the Odd Lots podcast. I'm Joe Wisenthall. And I'm Tracy Allaway. Tracy, covering the AI boom is actually reminding me a little bit of the tariff boom in April, simply because every day there are new headlines.

Just today we're recording this November 12th. Anthropic commits $50 billion to build AI data. centers in the U.S. So the advanced model companies are vertically integrating more to build their own data centers. Every day, some new development. Yeah, it's becoming pretty hard to keep up. So I think we're probably just going to talk in terms of billions and trillions. We're just going to say lots and money is going into the space. But the way I've been thinking about it is, okay, at this point, everyone agrees that the AI buildout is super expensive. And all these companies are spending massive amounts of CAPEX to do this. And I,

I'm starting to think that AI CAP-X is kind of like the Schrodinger's cat of markets in the sense that it could either be a massive strength for these companies because the CAP-X is so expensive and it takes so much money to build out. And so anyone who manages to do it kind of builds a moat around their business or it could be a massive weakness, right? If you're spending all this money and then that doesn't end up generating the revenues that you actually need to justify it. And going back to the Schrodinger's analogy, it seems like we just don't know what's going to come out of the box, right? Like it's simultaneously a strength and a weakness. And until we build out AGI or whatever, like we're just not going to know. I totally right. There's so much at stake here.

And obviously we know the numbers are absolutely enormous. They're staggering. And we could talk about them too. The financing structures are also very interesting. Yeah. You know, it's one thing if you just have meta. or alphabet, and they make a ton of money already, and they're spending money on data centers. Whatever, that's one thing.

It's another thing when you start seeing these SPVs where the hyperscaler puts in this amount of money and then the private credit puts in this equity, and then they borrow a bunch, and then there's all these questions about the payback. And we think of tech as from years and years as basically being this equity story. And when it becomes a credit story. Yeah. And when, you know, people are talking about quoting Oracle CDS. I always forget these companies even have CDS.

because I'm so unused to thinking of big tech companies as credits. So when I see people starting to tweet Oracle CDS charts or core weave CDS charts, it's like, okay, we are in a different level of capital intensity here. Right. And some of those swaps have been going up lately. I'm going to say one more thing. Thinking back to the 2008 financial crisis, I remember the economist at Raymond James, I think it was Jeff Sout, who went on to become a very big name. Yeah, we should have him on the podcast. But he made the point that historically, when you had real estate crashes, property crashes, it was usually because of a problem in the economy. But then what happened in the run up to 2007, 2008 is the housing market crash became

the proximate cause of the troubles in the economy. And if you think about how much money is being spent on AI right now, again, billions, trillions, possibly of dollars, it's very easy to see how. AI could morph into a problem for the wider economy. For the real economy. Totally. Just on this note, and then we'll get into our conversation, the Center for Public Enterprises out with a great report today called Bubble or Nothing by Edvait Arun, pointing out one of the things that makes data centers interesting is how they sit at this intersection of essentially industrial spending and real estate. It's an interesting asset class for its own right. So much to talk about.

We could never do it justice in one episode, but that means we got to do more. Anyway, I'm very excited for today's episode. We really do have the perfect guest. Someone who's been writing about this for a long time, someone who's just been writing about the internet and all things for longer than any of us, someone who's been blogging and investing for far longer than either of us or anything like that, way more knowledgeable about how these businesses work than most, very focused on the data center buildout. We're going to be speaking with Paul Kodroski. He is a fellow at the MIT Institute for the Digital Economy, also a partner at SK Ventures, and longtime internet, blogger, writer, yapper, et cetera.

Yeah, yeah, I guess. Someone we've never had on the podcast before. So Paul, thank you so much for joining us. Hey, guys, thanks. Good to be here other than the blogging part. No, it's all. It was you're a true pioneer in that. And it's impressive that you still write with the output that you do.

At some point in the last year, I feel like you really got laser focused, or maybe in the last two years, really got laser focused on the data center story is. This is where the action is. Yeah, I did. And in part, just because I caught myself by surprise with it, it was weird. I was looking at first half GDP data, actually first quarter GDP data earlier in the year. And, you know, this has become now a commonplace that people know this. But I hadn't realized what a large fraction of GDP growth in the first quarter data centers were.

It was on the order of 50 percent, much larger. If you included all sort of externalities, all the other things that data center spending in turn kind of accelerates. And then obviously the same thing was true in the second quarter. And it was, I got back to thinking about my dog. And my analogy is that, as one does. As one does. And I got to hear like, my dog barks when the mailman comes to the house and keeps barking, and then the mailman goes away.

And I'm convinced he thinks he makes the mailman go away, right? He has this really screwing up causality. And it's like, dude, if you don't bark, he goes away anyway. This is part of the job. They just go away. And I think about macro policy in the same way that if you don't understand the drivers of GDP growth, you're likely to think that whatever it is you would most like to be caused. GDP growth is doing that. So in the case of the U.S. in the first half of the year,

you know, this puzzle was, well, maybe it's tariffs. Maybe tariffs are actually contributing to it. Maybe consumers are much more resilient than we expected. And as it turns out, a huge factor, probably the largest factor, was this sort of unintentional private sector stimulus program, otherwise known as data centers. And for me, that all started, so that started this puzzle of understanding this sort of, just commensurate size, the consequences of that size and the acceleration's consequences in terms of where the money is coming from and all sorts of other things. But just to reframe in terms of something you guys were already talking about in this, I think, is super important in understanding why this particular episode is likely to

turn out to be historically really important. When you say episode, you're referring to this podcast episode. You're not referring to the broader episode of AI Data Center? No, entirely just the podcast. Okay. Who cares about Data Center? It's a 10-year anniversary of oddlodge. So the reason why it's going to be historically important is because for the first time, we combine all the major ingredients of every historical bubbles in a single bubble. We have a meta bubble, no pun intended for meta.

We have real estate. You guys just talked about this, right? Some of the largest bubbles in U.S. history had some relationship to real estate. We have a great technology story. Almost all the large modern bubbles have something to do with technology. We have loose credit. Most of the major bubbles in some sense have a loose credit aspect. And then one of the other exacerbating pieces that some of the largest bubbles thinking about even the financial crisis is some kind of notional government backstop.

You know, think about the role in terms of broadening home ownership in the context of the real estate bubble and the role that Fannie and Freddie played and loosening credit standards and all of those things. This is the first bubble that has all of that. It's like we said, you know what would be great? Let's create a bubble that takes everything that ever worked and put it all in one. And this is what we've done. So it's got a speculative real estate component. It is probably one of the strongest technology stories we ever have back to rural electrification in terms of a technology story.

We have loose credit. You guys talked about what's happening with respect to not just the role of private credit, but how private credit is largely supplanted commercial banks with respect to being lenders here. So we have all of these pieces that have all come together at once. And I think in terms of framing what's going on right now, it's really important to understand that it brings together all of these components in ways we've never seen before, which is one of the reasons why the notion that we can land this thing, on the runway gently is nonsense.

I love that framing. The meta bubble is perfect. Also, I had an epiphany earlier. I already told Joe, so you can attest to this. But I realized private credit kind of supplanted shadow banking as the term, right? Like after 2008, we called it shadow banking. And then at some point, it flipped to, I guess, the coupler private credit. Shadow banking always sounded sinister.

Right. In a way, the private credit doesn't do it. Someone figured that out. And they're like, well, now it's private credit. I like to think if it's a kind of financial witness protection program. It was like, oh, you're those guys. That's great. I understand now who you are.

Yeah, it's kind of like that. And it's now like one point, whatever it is, $1.7 trillion is the size of, which is that many components of the orthodox lending market combined in terms of the private credit industry itself. So that's a huge new piece of this that sometimes escapes notice how big it is and why it emerged. So all of those pieces. Yeah, it's stunning, the growth that we've seen.

Let me ask a very basic question. before we go further. But one thing I've been wondering is Joe mentioned that anthropic headline that we heard before. We've seen meta raising financing for data center builds, all that stuff. Why do these massively profitable and cash-rich companies have to raise financing at all? Well, they don't, but there's these irritating shareholders out there who get all pissy whenever you start diluting earnings per share too much and diverting it towards a single source. Now, that's not the case with private companies, obviously, but by the same token, Open AI doesn't have the luxury of having cash flows via which they can do any of the things we're describing. So Anthropic Open AI and everyone else,

they have no option other than to do exactly what we're describing. It's a different story with respect to how what percentage of Google's free cash flow or Amazon free cash flow that they want to continue to divert towards data centers. So in terms of the privates, this is the only option that they have. The public's obviously increasing the hypers increasingly, we got up to the point we're around $500 billion, sort of 50% of their free cash flow is going directly towards spending on data centers. And that's obviously a point at which, you know, we have other things we have to do with free cash flow, and including having some of it be earnings per share. And so increasingly, it's become the option. You see what Matt is doing recently with respect to its SPVs.

We bring in other participants, create new financing vehicles, and then we play this entertaining game of it's not really our debt. It's in an SPV. I don't have to roll it back onto my own balance sheet and then bring in new lenders, new private credit firms, and others. And so that's the reason, obviously. It's partly because of the scale. It's partly because the privates who have no other option. And it's probably we've kind of capped out the public companies in terms of the fraction of free cash flow that they feel as if they can spend with impunity on these projects. Explain to us, for those who don't know, you know, again, SPV, one of these terms that we really haven't heard in a while. And there's nothing inherently bad about an SPV, except that you only hear

about them typically after there's something, you know, some sort of crazy. Right, which is weird, obviously. Tell you, how would you, in the broad strokes, how would you characterize what these financing vehicles are? So mechanically, it's just a way of making sure that I don't have to roll debt onto my balance sheet, but legally it's a structure into which I and my partners contribute capital that in exchange for which they retain legal title to the project that we've created, which allows us to all contribute capitalists, but not have to put it back on my balance sheet

and therefore not to have that debt rated, which is really the key. Now, if you look at the actual intrinsic, say, for example, the recent meta project that they did in conjunction with Blue Owl, it's wild in Byzantine. It looks like something you might have seen in, what was that in Harry Potter, the forest with all the spider webs? It looks a little like that, right? Where everything's connected to everything, and all I know is there's something in here's going to get me.

So there's incredible complexity, but at the core, it's a mechanism via which I can raise more capital and keep it off my balance sheet by creating a legal entity that controls the actual data center, and I don't, therefore, have to put it back, roll it all back onto my balance sheet and have it rated. Now, there's weird intricacies, obviously. So, for example, what happens if at some period in the future this thing isn't performing the way we expect? Who owns it at that point?

Is there a payment exchange? Does it become metas? Does it become blue owls? Does it become someone else? And these things will turn out to matter. Right now, no one cares. If you go through some of the documents on these things, it's not entirely clear what the recourse payment will be if and when it ever has to revert back to another owner and it's not going to be

held on to by the SPV. And I think this will turn out to be really important four or five years down the road. But right now, nobody cares. I'm Francine Lacqua, an award-winning journalist. And I've got a new podcast, leaders with Francine Lacqua from Bloomberg Podcasts. I've interviewed everyone from heads of state to fashion icons about the news of the moment. But I've always been curious, who are these people as leaders? I don't think there's one right. way to be a leader. Make decisions. A poor decision is always better than no decision. Listen to new episodes every other Monday. Follow leaders with Francine Lacroix wherever you get your podcasts. So number one, the lifespan of data centers is actually not that long. I can't remember the exact

estimate, but maybe like three or four years, something like that. And then also you have this risk that tenants are sort of rolling through and no one knows what that actually means for the structure of the debt and you kind of get this asset liability mismatch. Yeah, so I'll start with the first one first. So this gets into something Michael Burry was tweeting about the other day, which was sort of entertaining, that back about four years ago, tech companies changed the depreciation schedule

for the assets inside of data centers. They extended them somewhat. Now, that wasn't an error. The reality is that data centers used for the purposes like at AWS, where you've got a big S3 bucket and I'm storing data inside of it, Those things, generally speaking, the assets are long-lived. I'm not running them flat out. These are not streetcar racers that I'm running around inside of the data center.

These are relatively inexpensive chips that I'm using for really mundane purposes, like storing large amounts, terabytes, exabytes of data inside of S3 buckets. So it's not unreasonable to say their lifespan's fairly long. They're not being taxed that heavily. So pushing out the depreciation schedule makes a lot of sense. But that was coincident with the emergence of GPU-driven data centers, using products like chips from Nvidia, and those have much shorter lifespans. So depending on the uses, so there's two different reasons why the lifespan and therefore the depreciation schedule of a GPU inside of a data center is very different.

So the reason most people think about is, oh, well, technology changes really quickly, and I want to have the latest and greatest, and therefore I'm going to have to upgrade all this time. That's important, but it's probably about equal, if not maybe slightly less important than the nature of how the chip is used inside the data system. center. So when you run using like the latest, say, an Nvidia chip for training a model, those things are being run flat out, 24 hours a day, seven days a week, which is why they're liquid cool, they're inside of these giant centers where one of your primary problems is keeping them all cool. It's like saying, I bought a used car and I don't care what it was used for. Well, if it turns out it was used by someone who was doing like LeMont's 24 hours of endurance with it, that's very different, even if the mileage is the same as someone who only drove into church on Sundays, right? These are

very different consequences with respect to what's called the thermal degradation of the chip. The chip's been run hot and flat out, so it probably, its useful lifespan might be on the order of two years, maybe even 18 months. So there's a huge difference in terms of how the chip was used, leaving aside whether or not there's a new generation of what's come along. So it takes us back to these depreciation schedules. So these depreciation schedules change just as the nature of how the lifespan of the chips changed dramatically. I can use something for storing things in S3 buckets for a long time. Six to eight years isn't unreasonable.

But if I'm doing the Lamont's endurance equivalent with the GPU, it might be 18 months. That's a huge difference in terms of the likely lifespan of a product that I'm depreciating over a very different period. And so that's a huge part of the problem here with respect to understanding the intrinsics in terms of how data centers can and can make money, how you have to think about the likely CAP-X requirements because of this much shorter lifespan of the underlying technology. And then talk about the tenancy rollover risk, I guess we might call it. Yeah. It's really interesting. So one way to think about data centers is as giant apartment

buildings, right? They're essentially gigantic pieces of commercial real estate with a bunch of tenants. Sometimes there's a lot of tenants. Sometimes there's only one. Sometimes Google bought the whole apartment building and just moved in. Or it's a giant office building. They just moved in. It's all theirs, right? So think about it in those sorts of terms. And the reason why, as a sponsor of a data center, I might take a different view on how many tenants I want. Again, you think about it in terms of what can I get Google to pay versus what can I get someone who's a much flightier tenant to pay? Well, I can get the flightier tenants, more of them and diversified as all leasing inside the data center, paying higher lease rates for GPUs over the

period of tenancy than I can get a Google to pay. Why? Because Google's got great credit. They don't have to pay very much, and they know they don't. So if you look at the commercial real estate data, the cap rate, the blended cap rate for the largest data centers that are tenanted by hyperscalers is horrible. It's like 4.8, 5.3%. It's like a, why don't you just buy a treasure? By a treasure. What are the world you're doing? So what happens then is people start blending in more different kinds of tenants to Tracy's point as an effort to try and improve the yield, the cap rate on the underlying instrument, which is the data center. So all of this should start to sound familiar because it's this idea of if I blend together all of the,

of these different tendencies, I can increase the yield of the securitized instrument, but that also changes the risk profile of what comes out the other end, which just takes us to things like the increasing usage of these things in asset-backed securities, which are these tranche securities that have all the different pieces. We have different layers associated with it, and that's a reflection of, well, there's different tenants inside these data centers, and people want different exposures to risks. So I may only want to buy the senior tranche. You may want to buy the mezzanine, and Tracy may want to buy the equity.

Can I just say, I know we already said this, but Paul is truly, truly the perfect guest. I remember reading his coverage of subprime and securitization in like 2008, and so having someone who's able to synthesize that experience with what's going on now is just fantastic. I kind of can't believe we're doing this again. I know. I mean, look, again, there's nothing inherently wrong with SPVs. There's nothing inherently wrong with tronching, right? Like a lot of these things are very intuitive.

etc. But it is still a little weird how central this is and how it's the same old, there's nothing new. I mean, on some financial level, it feels very familiar. No, there's nothing new into the sun. But I think that point's really important. It's not that tranches are evil. It's not the securitization is evil or that asset-backed security or project finance is evil. No, no, all of these things are terrific pieces of the arsenal whenever you're actually raising money for projects. The issue start to arise at the scale. which is what you guys have already alluded to. But the secondary piece, which again will sound painfully familiar to the financial crisis,

is there's a flywheel that gets created at the back end of this. So once you start securitizing the yield-producing assets in the form of these tranche securities, the people who are purchasing those things don't give a rat's ass what's going on inside this AI. I joke all the time that a lot of these people can't spell AI. They don't care what's going on inside the data center, right? it could be, you know, the World Hide and Go Seek Championships are going on in there. I don't care as long as it generates yield and I can securitize it. Well, it's very much analogous to what's happened in prior periods like

this, where again, you get this secondary flywheel effect of let's just create more of these things because our customers want more and they're really easy to securitize. And look, it's backed up by meta and Google or whoever else. Well, so this actually brings an important point. I mentioned this great report out from the Center for Public Enterprise. One of the things that they pointed out is in this market environment, where everyone has just be, you know, there's this sort of AI pixie dust, but also just the reality, if your revenues are surging, the market probably loves you. Like, talk to us about the unit economics here. Like, is the incentive for all the players essentially to just grow the top line as much as possible, even if these aren't, whether

we're talking about inference on a per token basis, even if these aren't particularly. particularly profitable. How are you thinking about the unity economics of some of these businesses and how that could eventually perhaps sort of, you know, come home to Roos, so to speak. Yeah. So the term of art, obviously, is these things have negative unity economics, which is a fancy way of saying that we lose money on every sale and try to make it up on volume, right? I mean, that's the problem here. So, but that's okay. I mean, we've had lots of things, Amazon and at early days had negative unity economics. You can get past that. And as an aside, I'll say right here, all of the things I'm saying isn't to say that, you know, AI is some kind of, you know, furry Tamagotchi thing that's just a fad.

Of course. AI is an incredibly important technology. What we're talking about is how it's funded and the consequences of doing that in terms of what's going to happen with respect to the businesses and the return on those businesses, right? So the unit economics are dire for a bunch of reasons, have mostly having to do with the more tokens you have to produce. the costs rise more or less linearly with the demand on the system, as opposed to an orthodox software business where the more people who use my service, the more people across which I can spread my relatively fixed costs. That's not the way that, for the most part, current generation large language models work, costs rise linearly or sublinearly with the number of users, which

makes for really crappy unity economics. And that's a big part of the problem. So from there, you get to the question of, okay, so what does it have to look? like in terms of making it look profitable. There's lots of ways to back into this. You can do bottoms up models that would suggest that like if every iPhone user on on Earth paid 50 bucks, that at work, we could have around a $400 billion, $500 billion annual stream of revenue flowing. Well, that's not going to happen, but it's worth pointing out.

Like, that would do it. But it gives you a sense of the kind of scale of what at a consumer level, for example, it might have to look like. People come at it from the other end. One of my favorite ways that people come out is to say, well, we could create a viable model here if you think, this was in the JPM call last week. I don't know if you guys saw the summary of it, but it was huge fun for the whole family listening in. So one of the ways they backed into it was a top-down model where they said, well, the global TAM for human labor

is $35 trillion. I love the global TAM. I said that was right up there with saying, like, if I reduce humans to their chemical components, here's what I can get for you. Well, this was Steve Eisman's line, which was like, beware of anyone that mentions Tam. Right, right. No, exactly. And so then they play the next step is, of course, to say, well, imagine we can get 10% of that, right? Which is obviously one of the oldest cliches. It's like saying, you know, I'm going to get 5% of the Chinese market. No one ever gets 5% of the Chinese market. This doesn't happen. So the same thing won't happen with global labor. But if you were to do, you do the math on that, those kinds of numbers get you to a weighted average cost of capital

basis to a reasonable return on current and planned expenditures with respect to AI data centers. If you assume we're heading to about a $3 or $4 trillion, a number, which is kind of the, I think it's around the number that most people put out there, which I think is a completely wrong number, but nevertheless, that's the kind of number in what you'd have to do to get there. So you can get there from a bottoms-up model by making some really unreasonable assumptions about the total numbers of subscribers and what they pay. You can get there from a top-down model. You can also get there by thinking about it purely in terms of industrial users.

I think about purely API users, but pretend retail users of AI don't exist and say, you know, Anthropics projecting $70 billion in revenue in 2028, something like 35% of their current revenues. Most of their revenues today are from their API. 35% of that is from software developers. That split between two large users, copilot and cursor. And so, you know, we can model that out. Everybody has to become a software developer and we can make the math work. The problem is it's got huge fragility, right, in customer concentration risk.

So a cursor disappears as a user of Anthropics API, and you just blitz. blew out 15% of your revenues because they're gone and they've done something else. And as it turns out, Cursor two weeks ago announced that they were trading their own internal model that you could use for software development. You wouldn't have to call the Anthropic API. So you can think about all these different ways to get there, but they all have a lot of built-in fragility with respect to either. So we all become software developers and we all subscribe to Cursor. Just going back to the used car analogy that you mentioned before, when we're thinking about all this financing of the AI CAPEX spend,

is it useful to think of GPUs essentially as the collateral? The problem, yes. Or what would you call the collateral in this case? So what ends up happening, the collateral in this case is the GP. There's no question it is the GPU. The issue is this disconnect, this temporal mismatch that you alluded to earlier with respect to the duration of the underlying debt and the assets that are producing the income that allows me to pay for the debt, right?

So we've got this probably unprecedented temporal mismatch with 30-year loans and two-year depreciation on the underlying collateral, which is essentially the GPUs that are the income-producing assets. And so that creates this constant refinancing risk because I'm going to continually have to turn over the base. And we've seen this many, many times. Right now, it's easy to turn it over, but in two years it may not be possible. There's a wave of refinancing is coming in 2028 in many of the more speculative data centers.

Will they be able to turn over their debt and refinance all the GPUs? Today they could, today is in 2028. So that's the inherent problem, is this structural temporal mismatch between the income-producing assets and the duration of the loans. And it gets worse. If you think about it in more holistic terms, think about it in terms of one of the other gating factors here that's driving all of this is the scarcity of energy supply. It's really difficult.

You can hook them up to the – well, it's actually kind of turned into a bit of a joke. I can hook you up to the grid, but I can't give you power. I don't know if you saw the recent episode with the Oregon Public Utilities Commission. Amazon had three data centers that they connected to the grid, and it was kind of like the Oregon PUC said, oh, you want power too. Oh, wow. We can't help you with that. We can't help me with that. So now there's a complaint in at the Oregon PUC from ADS, Amazon's the digital services group that runs AWS, complaining that we now have data centers, but we have no power. It sounds a little bit like a winter storm hazard or something, but it's a structural problem with respect to the inability. We can connect people, but we can't provide them with power. So the next stage, is, and this takes back to the collateral problem and the temporal mismatch, is that people are doing behind the meter power, they're building natural gas, or if you're Fermi, you're saying wild things about nuclear power, and you're saying, okay, I'm coming with my own power.

You don't need to connect me to the grid, because I'm going to power this myself. That creates two or three different issues, but among the more important is, think about how long-lived an asset, a natural gas plant is. This is not something that's got a five-year lifespan, and we just truly wave goodbye. This is going to be running probably 25 to 30 years. And the only thing, your ability to forecast, we know the cost of the natural gas plant, but in terms of the cost of the center and its ability to generate enough income to pay off the loan associated with the natural gas plant,

God help you if you think you can sort that out, because what you've really got is a huge likelihood of a stranded acid out there, natural gas plants that no longer useful for powering these things that they were built for. You can get the news whenever you want it with Bloomberg News Now. I'm Amy Morris. And I'm Karen Moscow here to tell you about our new on-demand news report, delivered right to your podcast feed. Bloomberg News Now is a short five-minute audio report on the day's top stories. Episodes are published throughout the day with the latest information and data to keep you informed.

Yes, there are other products like this from a variety of news organizations. But they usually rerun their radio newscast. throughout the day. That's not what we do. We create customized episodes that can only be heard on Bloomberg News Now. And we don't wait an hour to publish breaking news. When news breaks, we'll have an episode up in your podcast feed within minutes. So you're always getting the latest stories and developments. Get the reporting and the context from Bloomberg's 3,000 journalists and analysts we're all over the world. Listen to the latest from Bloomberg News Now on Apple, Spotify, or anywhere you listen. news is that Daniel Yergan said this on the show. You know, the back orders for natural gas

turbines, like if you ordered one today, you would probably get it in 2030. So the good news, I suppose, is that at least you don't have to have the turbines sitting there for years. Like, I don't know, I don't know if that's good news at all, but there are such, you may never get it anyway. You may never get the gas plant built anyway. Someone will be stuck with the bill. But it kind of raises, this goes back to Tracy's question earlier. This raises a really interesting things. So like honestly, what the F are all these people doing who are announcing these giant funding transiting? I think of it like people all showing up at the OK corral at once. And it's like, dude over there has one guy and I got two. Yeah. That guy's go, oh, two. That's not a knife. This is a knife. Yeah. But it's this deterrence.

It's this deterrence program that's going on. Don't even imagine spending 50 because I'm spending 100. Yeah. There's no point in you doing any of those. That's very, this game theoretic. Well, this also worries me because you hear so many people framing this as like an existential competition, right? And once you start calling something existential, the limit on spend, well, it becomes unlimited, right? It's about survival, so you'll spend anything. That's why the conversation has turned in recent weeks to the one entity that actually, at least in theory, can print as much money as possible. Right. That's the, you know, the Sarah Friar's accidental foot and mouth thing earlier in the week. But that's right.

But that's, again, goes back to my original point about what makes this bubble unusual. It's this element that not only is there a kind of backstop, but there's actually a notion of wrapping it in the flag. We have to win this competition. We have to do what it takes. This is existential. It's us versus China. And it's not just the U.S. doing this. I was talking to some Canadian policymakers just earlier this morning.

Exact same thing going on there. We have to build out a domestic industry. Same thing in the U.K., same thing in Germany. And so there's this idea around. the world, that sovereign AI is something that's incredibly important. So this government backstop isn't just mythic. It's global. It's this idea that we all have to win. We all have to win, which obviously can happen. But the government's playing a role in that that PIC can create this kind of limitless force of capital. You know, so one of the things that's going on,

and maybe it's part of the same, the sort of maximalist strategy mentioned Anthropic wants to get into data center. So everyone's sort of looking at how they can expand vertically. Can I own the data centers. I think, you know, Sam Altman has talked about owning chips or owning a semiconductor fab at some point. Like, maybe that'll be part of the story. Who knows? There's one thing that I don't, I'm sort of curious. I'd love to have your take on. There was at the end of September, meta announced a deal to buy compute from CoreWeave, one of these neoclouds. I don't totally get that because meta has its own data centers, et cetera. Do you have some intuitive sense about what an established hyperscaler needs a neocloud for in this arrangement, what coreweaves can supply that

meta can't build on its own or buy on its own? Nothing. So that's the answer. So here's what's going on. This is what's going on, is that there's this form of hoarding going on. So what's happening is, is people saying, you have capacity, I can lock that up, I'll lock that up. And because I can't lock it up yet by building a data center quickly enough, I'll lock it up in the marketplace.

So once you start thinking of compute as a hoardable commodity and what people are doing is trying to hoard it, control it before someone else can do it, because until they bring on their own excess capacity, that's really what's going on in a lot of these transactions. This is a way of making sure that I may not need this, but you sure can't have it. And so there's an element of compute hoarding going on across the map because of, you know, this backlog in building data centers that may or may not ever get built. So that's the answer. The answer isn't that they care at all about whether or not they're going to run. giant workloads on any particular neocloud provider. It's the idea of hoarding capacity in making sure that no one else can have it, like trying to, you know, like the Hunt brothers and getting a corner on the silver market. You know, I want to go back to China because it is true that the U.S.

and China seem locked in this existential race for AI supremacy, but they seem to be taking very different approaches to it. And in the U.S., it's all about spending as much money as you can, developing these, you know, state-of-the-art, mostly closed-source models, whereas in China, it seems to be much more about rapid adoption and creating open-source models that just get out into the market much faster and much more cheaply. And so I'm curious, like, which of those approaches do you think it's going to win here? Yeah, so that's a really good question. So I think it's going to be something closer to the Chinese approach, but not for the reasons they expect. So the reason is because I'll reframe what the Chinese are doing slightly.

So I'll say that instead of it just being a sort of an example of open source, I don't think that's the right way to think about it is they're using this kind of distillation approach increasingly where there's kind of a, you think about it like, okay, I'm a sales manager. I don't want to train all my salespeople. I'm going to train this dude and they're going to train all the sales, but that's distillation, right? You train the trainer. I train somebody who trains something else.

and the something else in this case are these smaller models. So that approach of kind of training the trainer really speeds up the process of creating new models because I distill them. I train them out of other models that are really compute intensive like Anthropics or opening eyes or whomever else is, right? So the notion is there are huge efficiency gains to be had in training. And the Chinese are showing the huge efficiency gains to be had. And the one way to think about it is that the transformer models,

that underlie large language models that are so computationally intensive, went from the lab to the market faster than any product in technology history. So they're absolutely bloated and full of crap, right? So these things are wildly inefficient. There's all kinds of other ways to do the same sorts of things, one of which is distillation. So what you're really seeing is a kind of an accident of history that we came down, the U.S. came down this path that led directly out of the original transformer paper in 2017.

And the Chinese have said, yeah, we're not going to be able to do that. for a bunch of different reasons. But we don't have to do that because I can take this approach of distillation, which lets us get, and if you look at Kimmy, this sort of relatively recent open source these things are actually really effective and benchmark very well. And it's not surprising because they've been trained by really good trainers, which is to say some of the other models that are out there. But these are about efficiency gains, which should then ask,

the next question is, whoa, wait a minute, if there's all these efficiency gains ahead from training and training is 70% of the workload on data centers, hang on a second. And aren't we completely misforecasting the likely future the arc of demand for compute? And the answer is yes. And this is, rather than looking at it as an example of why China is doing something better for worse, another way of looking at is saying just refuted the approach that we're taking to training altogether because it shows how bloated and inefficient the approach we're taking

is. And yet we're projecting on that basis what future data center needs are. Part of the question, it seems to me, and this is where it gets a little bit philosophical, is what do these AI companies think they're building? Because one theory is like, well, maybe they're building business tools, right? Maybe they're building business tools of various sorts. And if they're building business tools of various sorts, that implies the possibility that eventually they get good enough.

This does the job, right? This makes it easier for this website. You can use an agent to book your travel and the technology works and we don't have to keep building it because we got to the point where it works. And then there is this other question of like, well, maybe they want to build something called AGI or ASI that's like so sci-fi, et cetera, in which case you could never get enough or simply having built the thing that allows you to book your travel or book a dinner reservation or translate a text or whatever, that's not nearly enough. You hear different things, but what do you think the builders at the cutting edge of these labs are going for? Is it really the sort of sci-fi building God cliche, or do they want to build profitable business tools? So it's the first thing until you challenged them, and then it's the second.

So what happens is if you have the conversation internally, they'll say, yeah, no, no, no, we're building this really effective productivity enhancing tools that will be used across a host of businesses. And these all sounds really good. But then when you walk through some of the math in terms of justifying the ROI on the spend, all of a sudden, then it turns into what I call faith-based argumentation about. AGI. And they say, it's like the greatest call option ever. Like, what would you pay for a call option that could get you anything? And it's like, well, wait a minute. This isn't a way of just just a faith-based argumentation. We're saying, you know, with the Uber call option for anything, you should be willing to pay anything for it. And obviously that kind of justification doesn't get you anywhere. So in-house, they'll arm wave a lot about these different models that will

emerge. Who knows? I had someone at NVIDIA tell me the other day that we really are just waiting for the Uber of AI to come along and show us the future. And I'm like, okay. But it's not an answer, right? Because in theory, if you're building a business productivity tool, then eventually you could solve your unit economics problem, right? If you're just trying to build a really great business opportunity, then it's simply, you know what, we don't have to build anymore.

It works. And then the cash flow just starts pouring in and the cost per token goes down. You can. And there's a bunch of that already happening. It's really interesting. But what's increasingly happening is the problems they're solving are really mundane. So it's things like I'm trying to onboard a bunch of new suppliers. Right now, the people have weird zip codes and they sometimes don't match up.

I have a dude in the back who fixes that. I'd rather have someone who could do it faster so they could onboard a lot more suppliers. Oh, it turns out these small language models are really good at that, these micro models like IBM's granite and whatever else. But those things require a fraction of the training are very cheap, are not going to justify. anywhere near the economics needed to pay for the current spend. And yet those things are almost very likely the future because it will be profligate token use

from micro models, often hosted internally, to do really mundane background tasks, not very glamorous, onboarding new suppliers, matching records. Yeah. Great stuff, just not really very exciting, but large language models are amazing at it and small language models are amazing at it and almost free.

And writing songs, right, Joe? And writing songs. I'm actually, I'm still annoyed that AI is like getting into art and music writing and all the fun stuff versus the stuff that I don't want to do like folding laundry to your classic example. Or matching customer records. So going back to the beginning of this conversation when we were just talking about the scale of AI investment and its impact on the U.S. economy, I'm pretty sure you are one of the ones who's described AI CAPEX as like a private sector stimulus program for the U.S. economy.

What are the actual consequences, either positive or negative, of having this massive private sector spend in the economy versus something, I guess, more typical, which would be a government stimulus or maybe growth driven by consumer spending or something like that? Yeah. So to an orthodox economist, the old line is like, it really doesn't matter what we pay people to do as long as we pay them, right? It's the idea of, I should be, I should be willing to pay people to dig holes in the ground and people over there to fill the holes back in again. It really doesn't matter as long as the money is out there in circulation, right? It's all just stimulus, right? So to that way of thinking, it doesn't matter because the money is all finding its way back into the economy. But I think that's obviously hugely misleading because in this context, these are investments created with an expectation of a return.

If they can't, then that flows backwards into all the entities that are built on that basis, whether it's private credit firms and their returns, the S&P 500, what is it, like 35% now, is AR-related, Mag-7, Megas, 10, whatever. 40%, 50% now the last two years return. So these is a massive negative wealth effect when you unwinded, not just in terms of the direct spending, but in terms of the wealth effect with respect to what people's holdings are. So this is not as simple as saying this has just been a wonderful stimulus program.

We're paying people to dig holes and filling them back in again. This is a wasting asset on something that's likely to be produced in quantities that we can never earn an economic return from, in part because of wildly flawed assumptions and projections about the future of demand for those units. notes. And so that's the deep structural problem. And then you can get into this whole question of like, well, if it's just private equity guys get hurt, you know, cares. Screw those guys. Right. And it's not, of course, because as we just talked about, it's in equity funds. It's firefighters and teachers money. Yeah, yeah, yeah. And it's in REITs now. Look at the larger holdings in REITs now increasingly our

data centers. And it's even in sort of sneaky backdoor ways. Like we're seeing increasing, I don't know if you guys are familiar with these new interval funds, they're appearing there. Oh, yeah. It's all over now. Paul Kodrovsky, we could, I have a million more questions I could ask you, but much like the race towards AGI itself, that would imply that we'll ever actually get to the end of this conversation. So how about we wrap here and then just plan on, you know, revisiting the conference six months, maybe three years. We just keep revisiting down the line where we are in the cycle.

As long as we haven't been turned into paper clips, I'm good. Yeah, that's the, no one talks about the paper. That's the nightmare, clipy. I feel like that was, no one talks about the old school paperclip maximizer stuff. Everyone's on to more esoteric fears. I know. People have moved on. Yeah. We need to worry.

Did anyone ever try to securitize Clippy? They didn't, right? I don't think so. No. No, they never did. Thanks, Paul. Okay, thanks, guys. Paul's so good.

That's a lot of fun. He's so good. Here's my highest form of praise for an Oddlott's guest. I am going to go back and read that transcript from beginning to end. That is a very good, that is a very good practice to do. Wait, you're not going to listen to it. No, I'm going to read it. I can't listen to it.

I just listen to it. No, no, I can't. I need to read it. I can't listen to our episodes. No, I just, you know, I think there's a lot, there's a lot more to do on all this topic. But the financing in particular and some of these arrangements, it's just incredible how the speed with which I guess I would say the financing has gotten interesting. Do you know what I'm saying? Oh, yeah. I think like a data center project 10 years ago in Microsoft AWS thing just seemed like a fairly straightforward.

It's probably more complicated than I appreciate at the time. But basically straightforward, we make this money and part of it is going to go to building more data centers to, you know, serve, you know, Amazon Prime streaming or whatever it is or some client thing or whatever. And then the degree of complexity with these SPVs and roll over risk and depreciation schedules and tranching of who. It's gotten very interesting, very fast. Life finds away. Life finds away, yeah. That was my terrible, terrible impression. I think that's absolutely right.

One thing I would say is the fact that a lot of these big supposedly cash-rich companies are doing this through SPVs that effectively preserve their balance sheet and their cash flow so they can do something else with it. I mean, a lot of companies use SPVs, sure. But I do think it says something about the scale. Right? Like there's a scale problem here where if all your spending was appearing on balance sheet, investors might think very, very differently about your company. And then the other thing I would say is, I still think the comparing contrast between the U.S. and China and their approaches to AI, both of them, I think, would agree that this is an existential problem of some sort or an existential competition. But they're following very different paths. Yeah. And it does seem to me like the arc of history kind of leans towards stuff becoming cheaper. I think the arc of history bends towards China, is what I thought you were going to say. Well, that too. That too. But it bends towards, you know, people generally. really want the cheaper thing and they want the thing that's like available now. And China seems

to be going for that. The counter argument is that if you're going to use an open source model for some purposes, you have to supply your own electricity, right? You have to supply your own inference. You've got to host on your service. Like you still run into some constraints. And so rather than having it be on whatever, whoever else is data center, you've got to find a way to run it yourself. Yeah, okay, but China has a leg up in electricity too. Which was the point that Jensen Wong made. I mean, part of the reason, like, there's so much talk about this these days right now is that the industry insiders are saying a bunch of weird things. Paul mentioned the Sarah Fryer comment.

Oh, yeah. And she said when she sort of had to walk back, but then she said the same. Wasn't there that Sam Altman thing? Then there was the Sam Altman thing where he was asked, how are you going to pay for all this? And he said, look, you want to sell your shares or not? Which is like the interviewer probably thought he was. Little defensive. Obviously, Jensen Wong, talking about how China was going to win.

maybe he was saying that because he wanted to catalyze more action on solving some of the electricity problems in the U.S. But, you know, the very people at the center of this are saying things right now that, you know, what's interesting too is, you know, this bullwhip phenomenon. Everyone, as Paul described it, he didn't use the word bullwib, but when everyone is trying to get their hands on the same gear, you've got to wonder how it's sustained what the other side of a bullwhip could look like. I don't know.

We just got to do more episodes on this. Yeah, we have to. So we leave it there for now. Let's see if it there. All right. This has been another episode of the Odd Thoughts podcast. I'm Tracy Allaway. You can follow me at Tracy Allaway.

And I'm Jill Wisenthall. You can follow me at The Stallwork. Check out Paul Kodroski's writing at Paul Kodroski.com. Follow our producers, Carmen Rodriguez at Kermann, Dashel Bennett at Dashpot and Kail Brooks at Kail Brooks. And for more oddlots content, go to Bloomberg.com slash oddlots with daily newsletter and all of our episodes. And you can chat about all of these topics 24-7 in our Discord. Discord. gg slash oddlots.

And if you enjoy odd lots, if you like it when we talk about the AI private credit leverage, subprime, economy, nexus, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes absolutely ad-free. All you have to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening.

This is Tom Keene, inviting you to join us for the Bloomberg surveillance podcast. It's about making you smarter every business day. I'm Paul Sweeney. We bring you complete coverage of the U.S. market open. We cover stocks, bonds, commodities, even crypto, all the information you need to excel. And I'm Alexis Christophers. Bloomberg surveillance also brings you the analysis behind the headlines. We do that through conversations with the smartest names in economics, finance, investment, and international relations. We do all this live each and every weekday, then bring you the best analysis in our daily podcast. Search for Bloomberg Surveillance on Apple, Spotify, YouTube, or anywhere else you listen. On the East Coast, listen at lunch. And on the West Coast,

listen as soon as you wake up. That's the Bloomberg Surveillance Podcast with Tom Keene, Paul Sweeney, and me, Alexis Christophorus. Subscribe today, wherever you get your podcasts. Bloomberg's Surveillance, Essential Listening, each and every business day.