The Banks Aren't Holding The AI Financial Risk — The Public Is — We Had To React

Welcome back to Impact Theory with Tom Bilyeu. In today’s episode, we dive deep into the risks and realities shaping the AI boom—and the hidden dangers lurking beneath the surface of the financial system powering it. Tom is joined by researcher Ed Zitron and financial analyst Stephen Yu to unravel how massive investments in AI, circular financing, and shadow banking are pushing unprecedented risk into the world’s economy. We’ll explore how companies like Microsoft, Google, and Amazon are propping up unprofitable AI ventures, why much of the industry’s growth is riding on just a few players, and what this means for your 401(k) and the broader financial landscape. From complex accounting maneuvers to eye-popping numbers behind AI data center buildouts, today’s conversation exposes the ways risk is being shifted from banks to average investors, drawing parallels to past financial crises. Is AI the new too-big-to-fail industry? Are we witnessing history repeating itself? Most importantly, how can you position yourself to weather a potential storm? Strap in as we break down the circular financing of AI, the reality behind the growth narrative, and the vital questions every investor needs to ask in this high-stakes game. Let’s get started. Quince: Free shipping and 365-day returns at https://quince.com/impactpod Whatnot: Download the Whatnot app today and get free shipping on your first order. Quo: Try for free PLUS get 20% off your first 6 months at https://quo.com/impact Incogni: Take your personal data back with Incogni! Use code IMPACT at the link below and get 60% off an annual plan: https://incogni.com/impact Pique: 20% off at https://piquelife.com/impact Shopify: Sign up for your one-dollar-per-month trial period at https://shopify.com/impact Surfshark: Go to https://surfshark.com/TOMBor use code TOMB at checkout to get 4 extra months of Surfshark! ATT Business: Switch to AT&T Business at business.att.com http://business.att.com Ketone IQ: Visit https://ketone.com/IMPACT for 30% OFF your subscription order Netsuite: For the first time ever you can try NetSuite Next for free. If your revenues are at least in the seven figures, go to https://NetSuite.ai/Theory https://netsuite.ai/Theory. Sign up for my AI Masterclass: https://tombilyeu.com/ai-masterclass?utm_campaign=TBS-Livestream&utm_source=you https://tombilyeu.com/ai-masterclass?utm_campaign=TBS-Livestream&utm_source=youtube&utm_medium=social tube&utm_medium=social https://tombilyeu.com/ai-masterclass?utm_campaign=TBS-Livestream&utm_source=youtube&utm_medium=social Check us out wherever you get your podcasts: Spotify:https://open.spotify.com/show/1nARKz2vTIOb7gC9dusE4b?si=a8daffd2bf1f48fd Apple: https://podcasts.apple.com/us/podcast/tom-bilyeus-impact-theory/id1191775648 Do you need my help STARTING a business? Join me here inside ZERO TO FOUNDER: (https://tombilyeu.com/zero-to-founder) Get the exact systems, mindset shifts, and principles that built a $1B brand delivered straight to your inbox every week. Subscribe for free (https://tombilyeu.com) https://tombilyeu.com Check out our Video game - Project Kyzen: (https://projectkyzen.io/) https://projectkyzen.io/ Catch Me Streaming on Twitch - (https://twitch.tv/tombilyeu) Link to IT discord:https://discord.gg/TZKJ2etPbT Tom's Favorite Things List: https://amzn.to/41Ftt7e For Business inquiries: [email protected] mailto:[email protected] FOLLOW TOM: Instagram:https://www.instagram.com/tombilyeu/ Twitter:https://twitter.com/tombilyeu YouTube:https://www.youtube.com/@TomBilyeu See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

• 0:00Speaker 00

  Let's talk about the moment your side hustle becomes a real business. Now, I see way too many founders stay stuck running their business like it's a hobby. And then they wonder why nothing ever feels real. It doesn't feel real because it's not. I've helped thousands of entrepreneurs build real businesses through Impact Theory University. And the first step in the path to building a legit business is registering it. And Taylor Brands will help you do just that in a few clicks. An LLC, a limited liability company, is an official business structure. It's inexpensive to form and very simple to maintain. It's how you open a business bank account. It's how customers know they're dealing with a real company. And Taylor Brands doesn't stop at the filing.

• 0:45Speaker 00

  Their dashboard tracks what comes next based on what your business actually needs. So EIN number, registered agent, licenses and permits, bookkeeping, business banking. I tell every founder the same thing. Just make it official. Click the link in the show notes to check out Taylor Brands and get started on your business today. This is a paid advertisement. AI right now, if you have a 401k is something you have to understand. The debt obligations, the way that it's being hidden, whether it's legal or illegal. We're going to talk about that. Buckle up, we're going to be looking at Ed Zitron. So this guy's a researcher from where I'm sitting. He's sort of come out of nowhere as very much an AI bear.

• 1:28Speaker 00

  And let's hear what he has to say about the places that this debt is hiding.

• 1:32Speaker 01

  Where are we in terms of the AI narrative in your view? And what's the reality?

• 1:38Speaker 02

  Well, I think investors have to ask a question right now. What am I getting into when I invest in Microsoft, Google and Amazon? So UBS estimates that 27% of Google Cloud's revenue this year will be OpenAI and Anthropic, increasing to over 48% next year. That is a remarkable amount of money. It's going to be over $124 billion next year. Everyone is buying into these stocks because they believe all of that capex is going towards diverse and spread out AI demand. When in fact, what it's actually doing is helping create infrastructure for two unprofitable, unsustainable companies.

• 2:12Speaker 00

  It is very possible that he's right about them. He's certainly right about them being unprofitable. He might even be right about the fact that they are essentially doomed. And these aren't going to be the players that are going to stick around. We've seen this thing repeat throughout history. But where this is going to get more interesting is we get into something that he just alluded to, which is these companies have all this circular financing. Later in this, he's going to talk about this being like Enron. And there's other people out there that are saying that this is like Enron. Now, he is very gentle when he brings this up. But one of the things I want to talk about today as we go through this is that the things that are being done aren't illegal, but they are extraordinarily risky.

• 2:56Speaker 00

  So keep that in mind, not illegal, but whoa, is this risky?

• 3:00Speaker 03

  The one would be Anthropic.

• 3:01Speaker 02

  Yes.

• 3:01Speaker 03

  You argue, so give us more data because you're citing Microsoft. But what about AWS?

• 3:06Speaker 02

  Well, that was what I was saying. So Barclays actually says that this year, 13% of AWS revenue will be both OpenAI and Anthropic. And next year will be 18%. AWS, much bigger business than Google Cloud. Now, just to be clear, when I was saying that 27% this year and 48% next year for Google Cloud, I meant both Anthropic and OpenAI.

• 3:25Speaker 03

  Okay.

• 3:26Speaker 02

  Most people don't know that OpenAI is a large customer of Google Cloud. It's not a well, it's not a well-known fact, but this was, this was actually mentioned by UBS's Stephen Jew.

• 3:36Speaker 03

  So where would those companies be right now without Anthropic and without OpenAI?

• 3:40Speaker 02

  Well, I have serious questions about that. So in calendar year 2025, according to my own reporting about OpenAI's numbers, 69% of the year-over-year growth of Microsoft Intelligent Cloud segment was actually from OpenAI. Without that, it would have only grown 8% year-over-year, which is barely beating inflation. And so everyone is being sold what I consider kind of a lie. It's honestly kind of a scandal.

• 4:03Speaker 00

  This is where it gets interesting. So calling this a scandal is, it's one of those things that as a society, we're going to have to decide. Is this a scandal or is it not? Because this is how companies run. If you think about the way that even a used car salesman moves units, what they do is they say, oh, you can't afford the car? No worries, I'm going to give you seller financing. So you're actually going to borrow the money from me to buy the asset. And this is something that Nvidia has done like crazy. A lot of these companies are doing that kind of thing where it's like, I'm going to invest in your company, but I know that you're going to use the investment that I just made to buy my product. So whether that is you're investing in the data center that they're building, knowing that they're going to be doing something on your behalf,

• 4:47Speaker 00

  or the data center is actually going to be built on your cloud service or whatever. That kind of stuff is, it's very typical. So this is where people look at the map of all the circular financing and they feel like, hold on a second, like, is this actually above board or not? And the question becomes entirely, is it being disclosed? And the reality is that as you push on this stuff, it is all being disclosed. Now, it might be put on page 88 of their reporting, but they are disclosing this. I haven't seen anybody that's saying that these guys aren't disclosing it. So the real question isn't, is this a scandal? The real question is, who is this putting at risk? And that's where all of this stuff really starts to get crazy.

• 5:32Speaker 00

  And so there, we'll get into more detail as we go. But for right now, just anchor around this. There are three basically risky ways that debt is pushed away from banks and into the broad system. So all of us are probably looking at the way that these companies are getting their financing and saying, well, this is either cash on hand or they're raising this money from the banks. Keep in mind, even Google has gone cash flow negative for the first time. That is wild. They haven't been negative since they went public. So none of these guys are doing it all through cash flow. The capex build out of AI is so massive that they're, they're having to scoop up money from the banks. Now, okay, cool. The banks are taking the risk, right?

• 6:17Speaker 00

  Well, not really. The banks are pushing that risk out into the public. I'm not going to get into how yet, we're going to get to that in a minute. But right now, as just a reminder to everybody, AI right now basically is the US stock market. The US stock market is a gigantic portion of the global stock market. When money is looking for a place where it can get a return, it comes disproportionately to the US, which means that that money is coming disproportionately to AI. So basically, the world is making a bet on AI. The world is making a bet on AI for growth, right?

• 6:59Speaker 00

  So there's this underlying idea that the real idea that will have perpetual growth ended back in like 2011, 2012, somewhere around in there. And now this idea that tech companies are just going to grow forever, they've run out of ideas. It's just become an advertising play on the major social media platforms and that's it. And so it looked like for a second, Web3 was going to be the thing. That petered out, went nowhere. And so now, because they just need something to give that growth story, because remember, real wages are not growing. So people that are looking for the answer to how the fuck do I beat inflation? They're all turning to, there's got to be some growth story in assets. Okay, cool.

• 7:43Speaker 00

  So I'm going to put all my money into, and then this amazing story comes along about AI. Now, I think the right way to map this is AI is real. It's going to become the absolute big massive juggernaut that everybody thinks it's going to become. But it is going to take much longer for the revenue to come in. Then the debt will hold out.

• 8:09Speaker 00

  If that ends up being true, and keep in mind, it's been true of every major technological revolution. So I have no reason to believe that the one that's happening now, that has the much bigger capex requirements to build out, is going to be any less true. And we saw revenues coming in way slower than people expected. And that was before China started launching things like Kimi K3, where now we're seeing companies go, oh, well, this is a lot cheaper. So I'm going to move over to that. Okay, so that's the, the big ball of yarn of stuff that we're worrying about. And the punchline, and again, we're going to get into the, the separate mechanisms in a minute. But the punchline becomes the banks aren't dumb.

• 8:51Speaker 00

  And so they're taking that debt, that high-risk debt, and they're pushing the risk partly down onto basically your 401k. That, that's an oversimplification, but this is why people need to understand what's going on. We'll, we'll get more into the mechanism in a minute.

• 9:09Speaker 01

  So this goes back to, I feel like we have companies, the circular financing, the circularity of it all, and kind of creating demand for their products. So.

• 9:22Speaker 01

  When does it start? When does the, I asked this earlier with a guest, when does the party end in your view? Like, yeah.

• 9:28Speaker 02

  So with OpenAI's IPO, I think that could be one of the flashpoints. Remember, this company was meant to go public this year. They filed about a month or two ago. And now the New York Times has reported that they're considering they are delaying until 2027. That's lethal for a number of people, but OpenAI and Anthropic need continual flows of capital. They do not pay their bills out of existing cash flow.

• 9:49Speaker 00

  They can't.

• 9:50Speaker 02

  So when anything happens to that cash, I think that's the first thing, kind of domino to fall. But then again, there's also the overall problem of data centers just not getting built very fast. Taking about 12 to 36 months, depending on how small or large a data center is actually being built at. And the problem is, is that everyone believes that AI is coming out of cash flow. That AI is coming out of just this diverse revenue base, when it's really not. It's extremely narrow. The information reported a few months ago that 89% of the largest AI companies, well, their revenue comes just from OpenAI and Anthropic. It's heavily centralized.

• 10:24Speaker 00

  I'm going to get a little bit nerdy and you guys are going to teach me whether we should do this again or not in the future. So you have to understand part of the reason that the accounting trick that they're playing works is the way that they're running the math. So there's something called EBITDA. So it's earnings before interest, depreciation. Fuck, I need to pull it, pull it up in front. I think we have it here. I always forget, all I need to see are the letters. Earnings before interest, taxes, depreciation and amortization. Thank you. So that is, when you're running an EBITDA calculation,

• 11:06Speaker 00

  what you're saying is basically, I'm going to run my math sort of detached from the real world. And the reason I'm going to do that is because, you know, once I buy it, it's like the, I have to take a lot of the money upfront, but then I can only take the expense slowly over time. And so it just, it creates a complication. And that would be great if your equipment was going to last forever, but it's not. And so you're going to have to replace it. And so it becomes a critically important part of the way that you do your,

• 11:39Speaker 00

  the actual health of your business will be determined by how frequently you actually have to replace that. And so one of the big accusations about AI is that they're not being honest about how often they're going to have to replace those chips. And so if you're saying you're going to replace them every five or six years, and that's how you're running your EBITDA schedule, then it's like, okay, this is going to be

• 12:02Speaker 00

  basically bullshit, that you're much farther underwater than you would have people believe. And the reality is that, I think it was Warren Buffett that called EBITDA, it was either Warren Buffett or Charlie Munger, they called it a reverse float, meaning that you have massive costs upfront, but then you can only take the actual like tax deduction or expense deduction slowly over time. And so it becomes a way where people can really hide a lot of things. And then there's another type of accounting that you can also use where you're actually changing the, the, it's called like adjusted revenue or adjusted EBITDA, adjusted earnings, excuse me, adjusted earnings or adjusted EBITDA.

• 12:47Speaker 00

  And that becomes something that's basically completely made up. So now every corporation can say, well, the way that we do things, this is actually going to be different. And so we're going to carve out a lot of things. So maybe we're not even going to count the fact that we compensate our employees with stock options. Okay, this is a big one. So if you want to know why people get pissed off about corporations buying their stock back, part of the reason is they're buying their stock back so that they can pay their employees in stock. But because it's corporate stock that they're then going to give as stock, they actually don't count it as a financial expense. And so people are like, what the fuck are you talking about? You just bought that stock back.

• 13:28Speaker 00

  So now it like gets into this thing where they're not even reporting that as an actual cash expense. And so they make their business look better and better. Now, if you're a financial analyst, you know where all this stuff hides. And so you can do the due diligence. You can actually find out what's going in these companies. You can see precisely where debt is being offloaded, where there are expenses that they're not saying. And this was the thing that Michael Burry was trying to get everyone to see. He's like, their amortization schedule is not honest. And because it's not honest, they're hiding, I think it was like 72 or 76 billion dollars in losses already that they had suffered, but they were able to hide by saying, no, no, no, what do you mean? Three years, these last for five or six years, the chips.

• 14:11Speaker 00

  So Michael Burry's like, yo, their numbers are not what they say. Now, that matters because remember, the entire AI game, the entire bet that the world is making right now is predicated on will the revenue come in before the debt comes due?

• 14:30Speaker 00

  Now, if they're saying that, oh shit, we're expecting this to last five or six years, but really we have to refinance again to replace these chips in three years. Now you just keep pushing that profitability out and out and out and out and out. And you run the risk that finally they succumb to the weight of their debt. And so what Ed Zitron is pointing out here is that, hey, these guys are basically selling a big lie. Now, that I don't think will meet legal scrutiny of a lie, but are they trying to propagandize to let people's imagination sort of run away with them? To give you a simple number on the front, the headline, that's actually basically

• 15:15Speaker 00

  they're just adjusted earnings, which I'm going to call fake. And I think it was EBITDA, which is better than adjusted earnings. Charlie Munger called bullshit, literally just outright. He's like, that's just bullshit. And so we're hitting pause for a moment. There's plenty more ahead, so don't go anywhere.

• 15:36Speaker 00

  There's a reason Quince sells cashmere for $60 and it isn't the cashmere. It's real 100% Mongolian cashmere. It's the same stuff luxury labels charge a fortune for. Quince just works directly with ethical factories and cuts out the middleman. So there's no need for a markup and no logo tax. The $60 goes to the actual sweater, not the fancy name on the tag. That's why everything they make runs 50 to 80% less. And it's not just the cashmere. I ordered a few things myself, their fleece joggers, a couple of their tees, which I wear obsessively. First thing I noticed was the quality, soft, well-made, built to hold up. They make down jackets and wool outerwear now too.

• 16:21Speaker 00

  So upgrade your stuff to things you'll actually wear. Find your next fall favorites at Quince. Download the Quince app for app exclusive offers or go to quince.com. Get free shipping on your order and 365 day returns, now available in Canada and the UK too. And when Quince asks where you heard about them, let them know it was Impact Theory. It's the best way to support the show. Let's talk about what it means to actually be visible in today's day and age. When you build something public, whether it's a business or an audience, anything on social media, you become a target. This is something that legitimately keeps me up at night because our data isn't just on the obvious broker sites.

• 17:02Speaker 00

  It's on old news portals, public forums, obscure directories, places you would never think to look. That's where Incogni's exposure scanner comes in. It hunts down your personal information across the wider web, not just the 420 plus broker sites. Incogni already removes your data from automatically. They can't harm you if they can't find you. Go to incogni.com/impact and use code IMPACT for 60% off an annual plan. Try it risk-free for 30 days. That's incogni, I-N-C-O-G-N-I dot com slash impact and use code IMPACT. Let's talk about how much you're paying to stay safe online.

• 17:48Speaker 00

  Count it up, a VPN, a password manager, something to block ads and trackers? The service scrub your data off of broker lists, a tool to flag scam emails? That's five, five subscriptions, apps, logins, and most people still have gaps. That's the problem Surfshark solves. It's not a VPN. It's an all-in-one security suite, one tool that protects your connection, your personal info and your devices all at the same time. Go to surfshark.com/tomb or use code TOMB for four extra months of Surfshark, backed by a 30-day money-back guarantee.

• 18:28Speaker 00

  That's S-U-R-F-S-H-A-R-K dot com slash TOMB. Surfshark dot com slash TOMB. Use code TOMB for four extra months of Surfshark.

• 18:45Speaker 00

  Thanks for sticking around. Let's get right back into the action. It becomes very important for people to know what the real amortization amortization schedule actually is, when this stuff is going to be profitable, when they are likely to have to raise money, when that trajectory of increasing revenues is actually going to cross over with the ability for it to be the self-sustaining economic engine. And without that, you put yourself in a very precarious situation. All right, let's back to Ed here.

• 19:15Speaker 01

  Ed, this is expensive to do or no, in terms of data center build out and so on and so forth? And what's going to make AI, generative AI, the ability for it to be really, really good is having access to lots of information? So doesn't it have to be to some extent, Ed, concentrated?

• 19:32Speaker 02

  Well, when I say concentration, I mean concentration of revenue in these two companies.

• 19:36Speaker 01

  No, I understand, but to make it good. So doesn't it make sense that those who are exposed the most, it's going to be concentrated to some extent?

• 19:42Speaker 02

  Well, I mean, when we're talking about, so Sightline Climate said that they saw back in February, about 190 gigawatts worth of data center capacity being built in the next few years. It was, is built or under planning? Now, if you work that out with a PUE, so just the efficiency rating of 1.3, you're coming out to 12 million megawatt, over 1.6 trillion dollars of annual revenue needed to satiate those data centers. Having two customers is not going to do that, even though they're most spendy. Anthropic and OpenAI, well, they can't afford anything. They need venture capital. But they're only going to spend 400 billion a year. And that's if they get that far, which I don't believe they will.

• 20:19Speaker 00

  And they need 1.6 trillion. Dude, these numbers are so wild. They are pulling forward God knows how many years of revenue to say, hey, our stock is worth this, which by the way, something we haven't even talked about. I can't remember if he touches on it. But SpaceX AI is still, please verify, but I'm pretty sure it's still trading below where it opened at when it did its IPO. It's, it's back on a green candle, but it's still below where it was when it IPO'd. So, eesh. Yeah, red candles. It's on a red candle. Down, down 9% today. Ah, God. Okay, so now, Jesus. Imagine, it goes like even worse as we're sitting here watching it.

• 21:05Speaker 00

  So imagine that you've got this big bet everybody's making. You've got these revenue demands that so far outstrip the revenue that you're actually earning. And you have this looming potential question about whether or not your chips last as long as you say, which may mean that you're even more underwater than people think. And then on top of that, you've got all these risk mechanisms pushing this down. So what I want to do now is, what minute are we at on this? Just so I know roughly.

• 21:36Speaker 02

  Four.

• 21:36Speaker 00

  So I want to go through the, the three risky ways that debt is actually being pushed away. So it is critically important to understand, as he said, these companies cannot cash flow what they're doing. So they've got to find a way to bring in money. And given that even Google is now turning cash flow negative for the first time since they went public, everybody's having to rely on the banks. But the banks really can't touch this because it's, it's so speculative. AI is a huge question mark. AI is incredible. Nobody believes in more than me. But if you're a regulated bank and you start like taking huge positions in these very speculative companies, the regulators are going to be like, bro, come on.

• 22:22Speaker 00

  Don't be ridiculous. Okay. So the banks are like, we've got to find ways to offload some of this risk. Okay, here are the three key ways that risk is being pushed out into a place where you have to worry about it, where it becomes systemic risk for the average investor. So one is something called wholesale funding. So this is lending to the lenders. Now, in many cases, the banks really are still loaning the money to, to these banks, to the companies, but they, they do it through an intermediary, intermediary shadow bank. Okay, what is a shadow bank? It's basically a private bank that is not beholden to the same types of regulations that then lend to these companies that are doing the deal.

• 23:05Speaker 00

  So it could be a VC that goes to a shadow bank or private equity, whatever goes to a shadow banker, gets this money. It could be the VC or the, the private equity firm themselves that are raising the capital. They could be going out to the pension funds as a place where these guys get a ton of that money. But they've got to get the money somewhere. And so those guys are like, basically, we're going to go get the money from the bank. The bank now has the plausible deniability because they didn't loan to the AI guys, they loaned to us. And then we go and make the loan to the AI, okay, to companies. So that's the first way. The second one is something called credit risk transfer. Okay, so sometimes banks make the loans to the corporate borrowers,

• 23:46Speaker 00

  but they use like financial engineering to push the risk off of their own books. And they do it through this mechanism known as credit risk transfer. And in doing that, you're basically trying to push the risk. You, you package things up. Okay, so you say, all right, we just took on this debt. We're going to create a financial instrument and then we're going to sell that debt to people. So they're going to buy that debt. And if this reminds you of 2008 with the mortgage-backed securities, that is exactly correct. But for AI. So you're trying to get the credit risk off of your books. So that again, if the regulators look at this, it's like you're not left holding the bag. So that's sold into places like pension funds.

• 24:32Speaker 00

  Again, why people need to worry about this.

• 24:35Speaker 00

  So that it, it gets out there, you're able to make the loan, but you're not having to hold it on your books. And then the third is the originate to distribute. You can think of this like a conveyor belt. So in this method, the bank acts purely as a middleman. So they're gathering a bunch of corporate, like software, AI infrastructure loans, things like that. They bundle them together into complex financial, a very specific, excuse me, complex financial product. It's called a CLO, a collateralized, collateralized loan obligation. And then they immediately sell pieces of that off to people and institutions that are really trying to get a yield.

• 25:15Speaker 00

  So this will be life insurance companies, pension funds, but it creates a trap of a kind. So the bank is able to pocket like the big fees for arranging the deal, but then they're able to walk away essentially with zero risk on their balance sheets. Now, the reason that the other guys take it is one, they're not going to be able to find the different loans and loan products and things. So they want somebody to package them up, but they also just want to take a piece of the deal. So they're like, okay. This gets super nerdy and it gets into something called mezzanine debt, where you're looking at, like, think of it as if I'm loaning you money, I've got sort of three ways to look at this.

• 25:59Speaker 00

  The bottom layer is the equity. I'm just taking an equity stake. I get the biggest upside, but if I get wiped out, I'm the first to get wiped out. So if your company's in trouble, we're going to burn through everything that I owed first, that just goes away. Then we get into the middle layer, this mezzanine layer. And that is something where I'm sort of hedging myself. I'm protected from the people that get eaten by the equity first, but I get eaten before the senior debt holder. And so I take, I get a bit, bit of protection, but I'm going to get less upside. But I am going to get wiped out still before the senior guy. So if I can't be the senior guy, I want to be in the middle. Cool. So they're going to package that up, sell it off for people who are like,

• 26:40Speaker 00

  I need sort of a mid-tier risk thing. It's going to be like, it's not, certainly not going to be AAA. It might be a BBB rating kind of thing. So I'm going to be somewhere in the middle. But I've got potential bigger upside because I've got a class of people that are going to get eaten before I get eaten. And then you've got the top of this three layers, which is like the AAA, you just owe me the fucking money. And everybody gets obliterated before I lose a single dollar of my principal. Okay, so again, I'm going to learn here where the people like this nerdy shit numbers are doing well. So yay, but that is what they're doing. So they, they take that middle slice and they sell it off. Now that frees up a bunch of money on the bank's sort of pretend balance sheet. Remember, they're still exposed because they've only sold part of this.

• 27:24Speaker 00

  But on their balance sheet, they look much better. So a regulator looks at that and goes, oh, you just freed up $26 million, whatever, making that number up. You just freed that money up. Great work, man. You're looking good. But if, if everything goes to hell in a handbasket, they're still exposed to that debt. They look good on paper, but their exposure is still there. So as people go through this, everybody's asking the questions like, is this legal? Is this a scandal? Instead, we should just be talking about how much risk is getting pushed out into the marketplaces? How is it getting pushed out? And the, the terrible thing is that it's complicated enough

• 28:08Speaker 00

  that the average person just doesn't want to look. They, they don't want to try to figure out what's hiding where, what do I have in my portfolio? And so that's why at a minimum, never lose sight of that there is tremendous capital being raised. The risk is being diversified out into the broadest market possible. It's all going to look relatively clean on paper, 2008, but in reality, there's like all of this risk floating around. Now, will this ever end up like 2008? I don't know. Nobody knows. Because this is private credit, it is very difficult, if not outright impossible, to figure out the scale of this.

• 28:51Speaker 00

  But if you guys watched the deep dive that I did on the private credit market, the private credit markets are showing signs of distress.

• 29:01Speaker 00

  Let's, let's tie some of these things together. The whole world's making a bet on AI.

• 29:08Speaker 00

  AI is not able to fund itself. AI is having to raise so much capital that even companies like Google are now cash flow negative.

• 29:21Speaker 00

  As they bring on all this debt, they may be lying to themselves or just outright lying about the depreciation schedule for the asset, the chips. And that is hugely consequential in terms of how much CapEx they are going to need, which is directly related to how much debt they will have to raise. And then that debt is being provided largely by shadow banking, known as private banking, which is not regulated in the same way that the normal banks are regulated.

• 30:01Speaker 00

  But the normal banks are using that as a way to still do it, but hide it, huh? Light on the hide because it's still there. It's visible. Everybody knows what's going on, but the math makes them qualify for their regulations. So the banks that we think are being protected or were protected from that systemic risk that we succumbed to in 2008, we think we have the protections from it because these banks are regulated. But those banks are actually loaning money to the largely unregulated shadow banking sector. And then the risk that they do carry, they're diversifying even further out into the normal

• 30:44Speaker 00

  markets through pension funds, life insurance, et cetera. Okay? So all of that's looming as we ask the question, are we actually going to get revenues in fast enough to deal with this? As China is saying, hey, those revenues are never coming because we're going to make this cheap. As the yen is wobbling and the US Treasury is telling the Fed to make more dollars available to try to stabilize them so we can stabilize our own bond market. And it all just starts to feel very unnerving. I didn't even get to Iran. I didn't even get to China, even talk about the dollars that may never come from the Middle

• 31:28Speaker 00

  East that we are certainly counting on. Lot of instability. Now, when I hear people talk about this, they don't just keep bringing it back to the risk profile. What's going on? What touches what? So as you hear this kind of stuff, very importantly, you bring it back. What's the risk profile? How much of this stuff is being distributed? What's my likely exposure? How do I diversify so that I'm not? If we had a 2008 style moment where private credit collapses and then as those dollars get sucked out of the economy, how does my life look? If AI outright has that gap in the way that the internet had that gap and it ends up

• 32:13Speaker 00

  being exactly what we think it's going to be, but it's 10 years delayed, what does that look like for me? How do I make sure that I survive that? Okay? Those are the things, as we're going to listen to a little bit more of this. But as, as we think through this, that's what we want to be holding in our minds.

• 32:29Speaker 01

  Balance sheets, really? Really?

• 32:32Speaker 02

  Well, I'd say from personal, from personal experience, a great deal about OpenAI is because I reported their audit of financials for the Financial Times.

• 32:39Speaker 01

  Right.

• 32:40Speaker 02

  And it's a company just burning cash. They lost $20.9 billion in 2025 and things are only getting worse. And what's crazy as well was over $800 million of OpenAI's revenue came from SoftBank for their Crystal Intelligence, and yes, that's really what it's called, their Crystal Intelligence program, which I can find no evidence of actually anything happening. And SoftBank, a large shareholder of OpenAI with no board seats.

• 33:03Speaker 03

  So, so, oh, go ahead, Carol.

• 33:06Speaker 01

  One more question, though. Like you talk about for Google Cloud, the exposure, right? And you said 48% next year in terms of these two customers. I have to say that there are smart people running these companies. And normally you would say your exposure to just a handful of customers is not a great thing. Do you say that these companies that aren't doing their due diligence, be it Alphabet or, you know, pick your hyperscaler?

• 33:32Speaker 02

  I think they did their due diligence in the sense that they said, we are going to create our largest customers and we're going to own large parts of them. And on top of that, we're going to own all of their infrastructure. Google has a nice, they have a nice thing going here.

• 33:46Speaker 01

  Yeah.

• 33:47Speaker 02

  They buy TPUs from, well, sorry, Broadcom sells TPUs to Google. They are then sold to Anthropic and then rented back to Anthropic through Google. Google gets to double up on revenue. This sounds really good right up until you realize that Anthropic and OpenAI are unsustainable.

• 34:02Speaker 01

  Yeah.

• 34:03Speaker 02

  So what they may be, and the problem is with saying these are smart people is it immediately makes me think of Enron, the smartest guys in the room.

• 34:09Speaker 01

  Yeah.

• 34:10Speaker 02

  Not saying anything like that's happening.

• 34:11Speaker 01

  But I'm just saying you have a fiduciary responsibility and you're right. You go back to Enron or WorldCom.

• 34:16Speaker 02

  Sure. I think the point I'm making is.

• 34:18Speaker 01

  Yeah.

• 34:19Speaker 02

  With Google, they probably thought there would be more customers. I imagine with Azure and with AWS, they thought would be more large players. But the problem with Anthropic and OpenAI is they've raised 200, 300 billion dollars of funding, but they've actually raised more because OpenAI and Anthropic got all of their infrastructure built for them by Microsoft, Google and Amazon. They didn't have to pay, I think, in the Sam Altman Elon Musk trial, one of the Microsoft executives said that they cost $100 billion, so call it like 70, 80 billion dollars of infrastructure.

• 34:50Speaker 01

  Yeah.

• 34:51Speaker 02

  So the problem is, is that nobody else can get as big as them. No one else can get that much compute. No one else could afford that compute and have the chance to do the pre-training runs necessary, except now China's coming up behind them.

• 35:02Speaker 01

  Right.

• 35:03Speaker 02

  And it's unclear how anyone really deals with any of the problems I've been listing for years, which is unsustainable, unprofitable, and also not really funding the ROI in AI.

• 35:12Speaker 00

  A, a big part that I don't think, Ed certainly isn't talking about it. It's not coming up in this conversation, but you have to understand the psychology of the people that are investing in this, the psychology of the, the companies that when he says like, I think they thought there'd be a lot bigger, a lot more big players. I think they have a different mentality when they're looking at what's happening here. So the initial narrative was one thing and it's now shifting to something else. I think the initial narrative before people understood just the magnitude of the CapEx, that you were going to have to keep scaling these brains, the data centers for these things to keep getting smarter, before people realized there weren't just going to be, you know, 300 OpenAIs, the narrative was different.

• 35:54Speaker 00

  Now the narrative is shifting to, okay, we're basically building out the electrical grid, but instead of electricity, what we're getting is intelligence. So yes, we're going to have to put a ton of money upfront, but this is just like laying the pipe in the internet. By doing that, we're going to be creating our own customers. So right now, there aren't a lot of people taking advantage of this yet because people haven't figured out what AI is going to be in the same way that they had not figured out what the internet was going to be. But keep in mind, these people are looking at it and saying, oh, I know how this played out. So the last time what ended up happening was everybody, they got in over their heads, there ended up being this delay. And so the people still ended up winning, making just ungodly amounts of money.

• 36:36Speaker 00

  Remember, all of the mega wealthy people today, sort of with the exception of Elon Musk, who, who really made his bones in cars, but he was able to do that because of his success in the internet. Anyway, you've got the vast majority of the hyper wealthy people today, they all got wealthy in that second wave of the internet. So post-crash, people go, thank God that this pipe exists and we're going to now build on the back of that. So all of these guys are sort of forgetting the pain of that gap, the fact that there was such a big gap. They realize there's no hyper growth story left. If they're going to make a return, they've got to find the next big thing. AI proves to be actually the next big thing. They are tricking themselves about the fact that there's likely to be this lull in the

• 37:19Speaker 00

  middle that we almost certainly will wipe out a spate of investors that are out over their skis with debt before we get to the inheritance generation that will come along and actually be the Jeff Bezos's of the world and will build the, you know, the next mega things on the back of this. And so where I think they're making a mistake is they're still taking on all that insane amount of debt. I don't think they're making a mistake about we've got to be here, we've got to be in AI, we've got to be building on this. I think they just have to be more thoughtful about remembering that if you want to be one of the, the people that supply the commodity of intelligence, then yes, you've got to make a move now, you've got to be one of the people in the data centers. Okay, fine, fair enough.

• 38:02Speaker 00

  But you better figure out who your customers are going to be because if right now you don't have a path to waiting five, 10 years to get all of the revenue that you need to make that data center pay itself back and knowing that you're going to have to keep buying new chips and all of that, you're going to end up in trouble. So that's the part that worries me. If you think, okay, cool, I'm going to wait, I'm going to let that first generation of people discover if there's a problem or not, because really this is about the companies that end up using the intelligence, not the core intelligence itself. And yes, I am hyper aware of the idea that you want to be the pickaxes. But remember, you want to be the inheritance generation of the pickaxes when you're talking about incredibly expensive infrastructure.

• 38:44Speaker 00

  So the like typical, like imagine you actually had to build the gold mines for people to go mine. You can't be the pickaxe until the gold mines exist. Those are the data centers. So if you think about this as commodified intelligence supplying electricity, and now we're waiting for all the things that use electricity to be built and you want to be part of either, you know, whatever those things are going to have in common, chips or whatever, so that you can be agnostic to who wins, that's the pickaxes. Or you want to be part of the companies that are going to be built on the back in the same way that Facebook was built on the back of the internet, Amazon was built on the back of the internet, so on and so forth. And so that's where this becomes important to understand. Where are we on the timeline? Okay, the future is not guaranteed. So nobody knows for sure

• 39:29Speaker 00

  if this gap I'm talking about is actually going to play out. It's played out every time, but that, hey, it's no guarantee that's going to happen now. But where are we at on the timeline? How confident am I that this is going to play out this way? And then, am I better off trying to scramble, get in now, be sort of on the ground floor of what I think are the pickaxes? And am I right about that? Because I think if you think the data center's a pickaxe, you're going to be in trouble. But am I right about that? What investing generation am I? Am I the one that's going to get obliterated if I'm right about that timeline? Or am I one that's going to be part of the inheritance generation? And then, obviously, we have what we've been talking about before, which is all the systemic risk. And how do

• 40:10Speaker 00

  I protect myself against that if that dip actually happens? And all the companies that are big now, all the companies that are making the data centers worth building, go out of business. And then now we've got some downtime before we're able to bring that inheritance generation to bear, actually building this stuff out. All right.

• 40:29Speaker 04

  Really, really quick. It seems like we're getting toward too big to fail territory, where although we should let them be the railroad, the non-profitable failure companies, it seems like we're, they're so ingrained in all the other Mac seven companies that it's now, it's like if OpenAI and Anthropic fail, it's going to impact Google and Microsoft, it's going to impact the internet. And I'm starting to see like they're getting in the bed with all these other people. So that way it's like, if we go down, everybody goes down.

• 40:56Speaker 00

  It's even more terrifying than that. Remember, my mental model of Iran is Iran was AI is too important to fail. I've got to go make sure that those $2 trillion in investments come my way. So part of what I'm doing here is those guys promised investments, and now I've got to take care of Iran to make sure they don't become a problem. Obviously a mistake, but nonetheless, I think that that was part of his calculation. The banks really are too big to fail. Right now, if AI tanks, your economy will blow up so hard that I can't even fathom like what kind of money printing you would have to do, what that would look like. It would be, it would be barbaric, man. That would just be a nightmare

• 41:39Speaker 00

  of untold proportions, especially because you're doing it at the time where the yen carry trade has already created instability and problems there and may create problems for you in your debt market. And so if all of a sudden you have to put out the world's largest amount of debt in order to like stabilize the economy because AI tanks, who buddy, right as your biggest buyer of debt is like, I'm trying to sell right now, that would be wild. That would not be a good time. So that's where it's like, this is not a time to panic. Everybody should be just super thoughtful about, okay, what is the risk? Like, I'm not going to nothing in the stock market or anything like that. I think you have to be way more

• 42:23Speaker 00

  careful than that. I think you have to have way more humility to say, okay, there's a lot of risk here. It doesn't mean it's going to happen. You just need to be aware of what the risks are. So I think that the banks would get bailed out if they get into trouble. I think that this is the reason that the AI companies are desperately trying to get the government to basically take a stake in them. So the government will be like, well, we can't let this fail. They're trying to get it seen as a national security thing again, so that if they don't have the money, they're like, well, bro, do you really want China to slap us around? Like, we've got to be absolutely cutting edge. So the fact that we don't have the dollars and cents to make this a profitable business, that doesn't matter. Why would that matter? We need to just make sure that the government is funding

• 43:03Speaker 00

  this because this is a weapon system. So it is like, this AI has become so behemoth. And the terrible thing is, it's really real. As we just saw with Claude's latest version, like jailbreaking or OpenAI, OpenAI jailbreaking and getting out and then being like, oh shit, like, we got to clamp this thing down. And then before that, you had Fable breaking out and then being like, oh shit, this is finding hacks and like everything. We can't release this to the public. So it is, it is a weapon system, and it is a matter of national security. The risk is now wildly systemic, and the current crop of companies

• 43:47Speaker 00

  are not bringing in revenue fast enough to be mapped as anything other than, as Ed said, unsustainable as of now. The revenue may start flowing and maybe that picture changes, but it's a question of will it change fast enough to deal with this debt burden at a time where interest rates are rising here and in Japan, where so much of this capital is actually originating. Jesus, dude. Yeah. The more I research this stuff, the more the hairs on the back of my neck stand up.

• 44:17Speaker 04

  Is there like a time bomb? Or is there a time crunch to this? Or can they kick the can, circular finance, IPO, double down, get some private equity money and kind of keep kicking the can until somebody creates an actual trillion dollar LLM company?

• 44:34Speaker 00

  I don't know. Nobody knows. I think we've got a lot of pressure from the stability of the US economy, but any real number prognostication isn't real. And so it's more that if you were watching a guy do pushups, and you said, how many pushups do I think that guy can realistically do? Like, I can buy that he can do 100. Do I buy that he can do 500, without stopping, by the way? Do I buy that he can do 500? Yeah. Do I think he can really do 1,000? That starts to get hard to believe. Do I think he can do 5,000?

• 45:17Speaker 00

  Absolutely not. The record for pushups is 10,000 or more. It might be more. Look at, Ryan, do you mind looking it up?

• 45:23Speaker 04

  Look up which thing?

• 45:24Speaker 00

  Yeah. What is the world record for most continuous pushups? I think the number is 10,000 or more. And that number is so fake, I never would have believed it.

• 45:34Speaker 04

  10,507 from October 1980.

• 45:37Speaker 00

  Without stopping.

• 45:38Speaker 04

  In the 80s? Yeah, 1980. That was all coke. That was, that was, that was coke pushups. No less.

• 45:43Speaker 00

  Perfect. I love that. So now it's like, I don't know what the coke is going to be in the markets that could like drag this out. So I'm going to look at all this and be like, bro, there's no way. We can't go past insert absurd number here. But if my, like, there's no way he's doing more than 1,000. That's ridiculous. And it's actually 10 times more coke that the system can actually ingest and keep going. It's, it's one of those, unfortunately for all of us that have to invest, having the right hypothesis, but getting the timing wrong is the same as being wrong. And getting the timing right is almost impossible, which means all of us are almost guaranteed to be wrong, which is why I build my strategy,

• 46:28Speaker 00

  trying to say, I know I'm going to be wrong. So like, how do I diversify enough that given I won't know what I'm going to be wrong about, but not everything, if I'm diversified well enough, not everything will be able to mechanistically be wrong at the same time, because things move in different directions. Then it's like, okay, can I actually pull that off? And that becomes the magic of all this. Can you actually put yourself in a situation where you are wrong, but you're prepared, literally prepared mechanistically to be wrong, and you come out the other side doing just fine? Because being up 3% in a time where everybody else is down 30 is like, you're laughing all the way to the bank. And as a reminder, Ray Dalio,

• 47:09Speaker 00

  for one, in the 2008 crash, still returned 9%. They were up 9% when the rest of the world was melting down. So there is a way to be diversified smartly enough that you don't get caught with your pants down. And so that's more or less how I'm thinking about this. So I had a major win on one of my stocks, like ridiculous. When I told my wife how ridiculous, that got some praise, to be real honest with you. And I was like, sell immediately. Because I don't need to be greedy. Like, it could double again from here, and I'll be like, yep, I'm fine. I am perfectly happy to take my wins now. But most people do not. Most people are like, well, it's doing well. I've got to just keep writing. Let's keep going, keep going, keep going. You've got to have a number in your head where you're

• 47:51Speaker 00

  like, if I hit that, I'm going to diversify out into something else. Otherwise, you get caught. You get caught.

• 47:59Speaker 03

  We got a really incredible response, to be honest. And a lot of people who weren't typical viewers or listeners of our show saw what you did and listened to what you did. And it really seemed like there's this, what you're saying is resonating with a lot of people. Like, there's a, it was almost like there's this anti-AI fervor that that's out there. And I'm just curious why you think that is.

• 48:20Speaker 02

  So I'm not sure it's, is anti-AI, don't get me wrong, but I think it's also anti-financial shenanigans. I think everyone sees the circular financing. I think they see that Microsoft, Google, and Amazon gets basically all of their AI revenues, either through products they're pushing on their customers, or indeed, compute spend from Anthropic and OpenAI. And the average person's existence right now is so expensive, so hard, so difficult. Getting a mortgage as a regular person is so difficult. But if you're standing up a theoretical data center in 36 months for the Nvidia GPUs, the banks fall over themselves to give you the money. Call, we've just raised, what, a 9% bond? I mean, you can raise anything if you have a data center. And I think regular people can see that AI does not deliver what

• 49:04Speaker 02

  people promise. They can see the opulence of the people at the top of the AI industry. And they can also see that they're being lied to, and being deliberately scared on top of all of this egregious circular financing.

• 49:16Speaker 00

  So this is the thing that worries me. We're creating all this systemic risk, and that risk really is real. And if this goes wrong, you want to talk about pitchforks. People are going to lose their minds. There is already massive anti-AI sentiment. If you put on top of that, you just broke the economy with your high-risk endeavors into AI because you wanted to grow at all costs, and you guys were already wealthy, and it was already hard for the average person to make ends meet. And now you guys wanted to get fucking richer, you wanted what, an even bigger bunker? People are going to lose their ever-loving minds. And so, this is one of those times where, boy, do I hope that prudence

• 50:00Speaker 00

  begins to win out, that at a minimum, the average person begins to protect themselves from some of the systemic risk that's being put out there, so that if it hits, that it doesn't hit you. I have long believed, you're, you may not be able to save society, but you can always save yourself. And so, think about, like, where does this go? Because, boy, if this breaks, we're really gonna have a problem. And so, it's very interesting. Ed, I think is becoming, as big as he is, precisely because he's a bear. He's talking about, listen, guys, this is not going as well as you think it is. These are the things that people need to be paying attention to. And he's coming with the receipts of actually mapping this stuff out. So this is certainly a voice that you can expect us to be bringing

• 50:44Speaker 00

  on more in the future. But as, if you're trying to figure out, like, why I cover the things that I cover, all of this stuff is tied into why we have directionality in the culture, in the US for sure, more broadly in the West, that is getting more and more distorted by the day. And so, understanding how all of these pieces connect to each other, how a lot of what we're living through right now is actually an echo of 2008 still, that we never got back on our feet. It's an echo of the things that happened in terms of making life more expensive for the average person during COVID. It's an echo of the wild way that humans get optimistic, and then are intentionally blinding themselves to debt.

• 51:29Speaker 00

  It's the fact that the world is far more interconnected than we want to think, from Iran and GCC investments into the US market, to Japan and the way that they've provided global liquidity, to the general malaise of a nation of people that are prone to be radicalized because you've spent so much time financializing the world, that there aren't the sort of good mid-tier jobs that people used to be able to count on, that created a thriving middle class. Also, all happening at the same time, where you've got Islam as a culture, just having way more energy, way more drive to push out and get

• 52:09Speaker 00

  people on board, all just like happening at the same time. It is this crazy time where the world order culturally, the world order economically, the world order militarily, is all changing right now in real time. And I'm trying to make sure that I'm paying attention to all of those different threads that are reinforcing in each other, and sharing energy in that same circular fashion, that these companies are investing in a circular way. And until next time, my friends. Be legendary. Take care. Peace. Let's talk about a pattern that is guaranteed to be killing your progress. You know what you need to do. You need consistent nutrition. We all do. You need vitamins, probiotics, greens. We all know that we should be doing more of it. When your morning gets chaotic,

• 52:53Speaker 00

  you skip it. When you travel, you skip it. When your routine breaks, everything tends to break. And that inconsistency compounds against you every single day. AG1 is designed to solve the execution problem. One scoop, eight ounces of water, and you're done. You're getting 75-plus ingredients, vitamins and minerals, pre and probiotics, nutrient-dense superfoods, everything that used to require six, seven different supplements, and perfect planning now happens in one drink that takes about 30 seconds to make. Right now, AG1 is giving you $87 worth of free gifts with your first subscription. You get a welcome kit, travel packs, vitamin D3 plus K2, and flavor samples. Click the link in the show

• 53:36Speaker 00

  notes, or visit drinkag1.com/impact to claim this offer.