Compute Is Revenue

back in march i published a bullish substack essay into a collapsing tape. software stocks were getting butchered. hyperscalers accused of losing their minds. nvidia was falling like a broken momentum trade while missiles were raining down across iran and every idiot on television suddenly became a geopolitical strategist.

but price was saying something else.

software was no longer scarce.

that was the whole point.

once code starts writing code the scarcity moves upstream into the physical machine. power. transformers. cooling. fibre. systems that cannot expand fast enough once demand arrives all at once. compute stops supporting revenue and starts becoming revenue itself.

from bar select in gustavia trader mike and i walk through that transition in real time. mike sitting perfectly still watching the machine while i pace around conducting imaginary charts in the air. none of those exchanges are invented. we’re very different traders staring at the same pressure points from opposite ends of the same bar.

eventually the market caught up. of course it did. the same hyperscaler capex once described as reckless suddenly became visionary once price turned higher. same reality. different price. the market had already decided while everyone else was still trying to sound clever.

this episode is really about constraint. who has it. who doesn’t. and what happens once intelligence itself becomes industrial infrastructure. 

copper carries current. fibre carries light. 

the winners stop looking like software companies and start looking like electricity grids.

i also go somewhere else entirely. bitcoin. derivatives. synthetic scale. optionality. and the uncomfortable possibility that conventional investing strategies increasingly guarantee an average life.

my friends. if you enjoy the episode share it with someone who watches price instead of headlines. subscribe. leave us a review. and come join us before the crowd notices the world has already repriced itself. summer acid camp aug2-6th in st barts.

remember, if you don’t own assets, you are the asset.

hugh.

Transcript

publish and be damned

SPEAKER_00 0:00

Ladies and gentlemen, this is another episode of The Asset Capitalist. I've this is my third attempt to record this. This is gonna be the one. And I want to talk about a paper that I published literally in the Middle Ages, it feels like that. I I wrote this paper and I published it on the 3rd of March in the large year 2026. I was bullish. Yes, I, the ASA Capitalists, I was bullish. Let me tell you, let me tell you who's not bullish. The S B 500 was almost certainly not bullish. On that day that I published to the Substack, the SP closed down 2%. The SP had erased all of its gains, it was down on the year. Nvidia had fallen from I don't know, 210, 215 to 160. It was unloved. That wasn't even the worst of it, let me tell you, because I had to wait another and do two weeks for of stocks falling, of ridicule, humiliation. I believe the saying is publish and be damned. I felt damned. Right. I'm doing sections. I have sections. I have prepared sections for you. Not for you, for my sanity.

the oracle had already decided

SPEAKER_00 1:19

Section one, Oracle. Oracle was almost like the the Oracle stone. I predicated this whole entire essay around the plight of Oracle. Why? Because at least Oracle had the blinking decency to stop falling by the time I was publishing my my little note on sub on, did I mention on Substack and Patreon? Oracle had fallen from $400 down to $135. In the world of markets, that is called absolutely fucking butchered. At least it was lying there when I published, like half half conscious, six weeks later, it was lying there, like having a cigarette going, I got this. Thank you, Oracle. You helped me. And today at 190, Oracle, my darling Oracle, I think you've got a chance challenging the all-time highs before the craziness is over. Anyway, the Oracle drawdown, it was kind of like something that was it was like drawing me closer. It was like saying like I needed an opinion because the sell-off wasn't behaving normally. The market wasn't behaving errant, bad businesses. The market had a it had almost like an a left bank, a Parisian existentialist sense of of dread. The market was beginning to hate good and bad. And Mike, Trader Mike, if I was to adopt the language of Elton John, Trader Mike was flickering like a fucking candle in the wind. Pull that thought. You see. Code was beginning to write code. And that's a very big fucking deal. Compute was becoming revenue. I've just invented, me not personally, but in 2026, we've invented a new lexicon. Compute was becoming revenue. And back at the beginning of March, this was all beginning to take shape. Prices, of course, eventually, after four weeks of ridicule, eventually stopped falling. And and they rallied 14%. And NVIDIA, and NVIDIA, NVIDIA made an all-time high. I think it's now 2.30. And that's why I was thinking my thoughts turned to that paper that I published on the 3rd of March. So I thought it might be fun to do that paper, Bar Select, and the end of the accounting age. And I want you to think of sitting in the bar in Gustavia with Trader Mike and me and getting a sense of what it felt like. So yeah, I forgot to mention that 900 American and Israeli airstrikes, we took out Israel at the same time as I was writing my great essay. And then I was publishing an essay telling anyone who would read it that the technology story was the technology sell-off was not the story. But Iran was not the story either. Constraint was the story. Constraint was this physical, unglamorous, it was this world of 18 months in the future, 18 month leap times. The constraint for a gas turbine, 18 months for a gas turbine, that nobody really on television was discussing back then. They weren't discussing it because you can't draw a yield curve around a transformer shortage. The way it seems you can do it around a barrel of oil. And of course, the Western elite, you know, the press, our friends, we're reaching for the fox in the box narrative. The crafty mullahs, the great Satan, certain to be outwitted again by the ancient Persian cunning. It's a fantasy that the foreign policy establishment, they sell it to feel I don't know why they sell it, because they need to feel sophisticated about a region they just seem to be so palpably out of touch with. Regardless, the strikes arrived, the paper was released, and the Iranian retaliation seemed modest at best. A declawed cat making the sound of a lion. The market flinched, of course, but the flinch wasn't really about Tehran. What was it all about as a capitalist? That was my assertion at the beginning of March. It was not Tehran, it was not hyperscalers, it was constraint. Let's discuss the constraints that Mike and I discussed at the end of February.

american technology stocks were dying

SPEAKER_00 5:52

I want to say to you, the bigger deal was American technology. American technology stocks, at least at that moment, it felt like American technology stocks were dying. The magnificent seven, the stocks that had carried the entire bull market on their shoulders for what felt like a decade. Stocks that capitalized at amounts greater than G7 economies. They were down hard. Microsoft at the end of March was trading at levels last seen at the end of 2021. Ball Street Journal called it a $1.6 trillion meltdown. A SaaS apocalypse. Market was looking at these stocks and it worries about margin compression. But a compression so violent, it would look biblical. It would be like an exterminator event. Markets are biblical in their imagination, perhaps. And the cruelest part wasn't really even in the size of the fall. It was that the companies themselves, the companies whose share prices were falling, were reporting really decent numbers. Revenue was growing, customers were renewing happily. Guidance was mostly intact and for hire. Market didn't care. Market was looking into the yonder. And the market was asking whether these businesses. Market was like having a cigarette and go saying, I just don't know if you deserve to exist. Why? Because AI had arrived, not a chat box. AI had arrived as an engine. And can I say to you again? Compute was becoming revenue. AI was now capable of doing what your expensive enterprise subscription. AI was faster, cheaper, and crucially, it was without human beings sitting in front of the screen. That's the key to this. It was the elimination of the lab rat trained by Salesforce and Adobe and all the others. The human had been the glue, the moat protecting these one-of-a-kind cash monsters. The software nodes, they actually had a social network. The software brought you in contact with other humans from different parts of the organization. And that bred great loyalty, if you will, around the fountain, the office fountain. And yet January and February of this year was when the language changed. Agentive agents entered my lexicon. I'd never spoken like this before. A new citizenry of workers that didn't need the social water fountain. Anthropic, unnerved people. You see, Claude was evolving so quickly that even I was using phrases like agentive workflows and agentive agents. I felt I could glimpse a world where the human interface itself was disappearing. And that again mattered because the human interface had been the glue, had been the moat to some of the greatest software companies. Motes were not built on code quality alone. Plenty of incumbents had mediocre code. Do you know what the moat was? The moat was behavioral, institutional. It was the assumption that large organizations would always require armies of human beings sitting together between the software and the outcome. That was the moat. People clicking, people chatting, people managing, reconciling, approving, learning. People. Software interfaces, people they could take them, they could leave them. They were tolerated because replacement felt impossible. And then it started, February and March, to dawn on smart investors that if you remove the people suddenly the mod, folks are less hell of a lot less formidable. And so that's why I say to you, by the time the bombs fell on Tehran, the technology market was already on its knees. And me, I was in the bar in St. Bart's drinking ice cafe latte. And I was turning bullish. Bar poor. Bar select. And that's what I write and what I record weekly. I meet people. I talk to everyone. And everyone knows me. I'm the acid capitalist. They have no idea what I talk about. And that's that's how I prefer it. My father, you know, he had that quality. His father too, I suspect, before him. You walk into a room and the room notices I don't think that's vanity, that's presence. It's how some people are built. Vibrations and all that. And there, in

how trader mike reads the machine

SPEAKER_00 10:34

Bar Select, sitting, as always, by the entrance, as usual, trade a mic. I won't give you his last name, but he prefers that I don't. He's a private soul. Forty years in markets, he's done it all. He spends half the year in St. Bar's and he trades. And when I say he trades, he watches what he calls, he calls it the machine. He watches it like a hawk. He's the friendliest, kindest guy you can possibly meet. But during market hours, he is impossible. Doesn't look up when you approach. I mean I call that unfriendly. He's like the invisible man that sees everything. He watches markets the way a predator watches watches the weather, expecting something to happen. Mike and I, we are opposites. I can't sit still when things get get interesting. I've got voltage in my lower spine. My hands conduct imaginary charts in the air. I pace. I shout, I'm vulgar. I say hello to the slender lady at the bar, and I mean it, and then I'm back. I either act manically or I die. There's no middle ground in my mind. But Mike, Mike doesn't move. Mike accumulates 30 years of building a stillness that most people mistake for detachment. It's not detachment. Mike has total, total attention. If I generate the heat, Mike generates the coordinates. Now maybe I should have said this earlier.

i am the acid capitalist; genius is staying lucky.

SPEAKER_00 12:05

I am Hugh Hendry. I am the ACE capitalist. 15 years running global micro hedge fund money at the sharp end. Am I famous or infamous for being right before anyone was comfortable being right? Was I insufferable about it? Probably. I've been called brilliant. I've been called a cunt. I've usually been called both by the same people, usually in the same sentence. Humble is for people who got lucky and they try or they need to disguise it. Me, I'm determined to stay lucky. Mike, Mike has a rule.

mike has a rule.

SPEAKER_00 12:37

He trades the first two hours of every day because they tell him whether the machine is in execution mode or discovery mode. In execution mode, every bounce is a gift to the sellers. In discovery mode? Discovery means uncertainty, and uncertainty can be traded. Mike says to me, index is theatre. He barely moves when he talks shit like that. The real information is in the internals, in whether buyers are allowed to hold ground or whether they're immediately suffocated. That's the tell, according to Mike. Me, I switch off. I switch off when Mike draws on like this. It's dealing bullshit. Never been my thing. But truth be told, the Monday of that week it did feel like execution. Every rally, savagely, savagely strangled. Then Tuesday, Tuesday, February, uh Tuesday, February 24th, end of. SP was up a percent. As that, more than one percent. Software stocks bouncing. Not because software had finally been saved. More likely the selling had simply become too heavy. The machine needed to reset its grip. That's what Mike thought. But he said to me, if it if it repeats this tomorrow, we're in. Because one day moves mean nothing to Mike. Two days and we're beginning to talk about behaviour. And Mike, he got the second day. Because Wednesday the SP was up again. NASDAQ again was stronger. One day means noise. Two days means the room is changing. Thursday was a stinker. Nvidia destroyed earnings. Destroyed everything. Nvidia's stock price dropped five percent. I told you in February and March, markets were uncaring. Position was too crowded, shouts Mike. One stock has become so fucking enormous that the entire ocean moves with it. It's not fundamentals. Mike was being unusually animated. And he started to to change, recalibrate his risk book. Because Mike lives inside tiny observation windows. One week, one day, sometimes less. He's attracted to behavioral stress, panic through option pricing. Moments where trading becomes so terrified of further downside, markets merely stabilizing can produce outstanding short-term returns for a trader like Mike. His whole game is basically children in the back seat of the car asking, Are we there yet? That's Mike. He's not pretending to predict the destination. He's trading the exhaustion of others traveling on the same highway. And this bit is crucial because when I published at the beginning of March, the markets kept falling. Nvidia kept falling too. Nvidia fell all the way to 160, down almost 40% from the high in the previous October. The SP fell another 10% during the month of March. And yet today both Nvidia and the SP trade at all-time highs. The interesting part was what Mike was attempting to see on those days of the week. Because he he wasn't clearly seeing a bottom. What he saw was the first behavioural tremor of change, of a market beginning to exhaust its sellers. Mike was seeing the embryo of a significant turn that we have to discuss. The embryo of a turn as I was building my thesis. Everyone in Marcus was asking the wrong question.

compute is revenue. why scarcity moves from code to copper

SPEAKER_00 16:39

Peak anxiety back then was hyperscalers are spending too much, they're out of control, they'll never make a return. Amazon guiding towards $200 billion in Capex in one year. Google $180 billion. Numbers larger than the GDP of entire countries. And the market, collectively, the market was the market was having a nervous breakdown. Markets don't like things where they have no precedent. Mike, however, doesn't do nervous breakdowns. He asks a different question entirely. Not are we there yet? But what can't you simply buy more of? What takes years instead of quarters? Because bottlenecks, that's where pricing power lives. And Amazon, Microsoft, Google, the hyperscalers, Meta, maybe they were once software companies. Almost certainly they were software companies and very, very profitable and very, very cash-rich. Stay there more like electricity companies, converting power into computation and renting it back to everyone in the world. The issue, crucially for now, is you cannot scale electricity like you can software. AI demand exploding in quarters, exploding in days within weeks, but transformers, substations, transmission capacity, the grid. I mean Microsoft openly admitted it. Tens of billions in Azure demand, revenue, profits simply couldn't be fulfilled fast enough because the capacity for the power wasn't there. Real world constraints, real world physics. And this is when Mike said the line that stopped me pacing in bar select. Software

copper carries current; fibre carries light.

SPEAKER_00 18:38

used to be scarce because code was hard to write. Now code writes code. So maybe the new scarcity is copper. And suddenly the whole thesis took form, took shape right there in my head. The new scarcity isn't code. It's the damn infrastructure. Copper carries current. Fibre carries the light. I love that line. You see, everyone sees copper. Dracken Millip, my hero, he called it one of the great profit setups in the world. Fine, electrification, data centers, eight years of underinvestment by commodity mining stocks. Everyone can see that now. But Mike was already thinking one layer deeper. Copper carries current, fiber carries light. Because the AI boom wasn't merely consuming electricity, it was consuming bandwidth, clusters speaking to clusters, campuses speaking to campuses, terabits per second through strands of glass thinner than a human hair. Now, when millions of people begin using these systems simultaneously, suddenly the fiber optics that we talked about from a bygone era, they stop being telecom infrastructure and they start becoming the infrastructure of a modern civilization. A step change. Waiting for a pullback. He would see the CEO. And the CEO is good, sings. He's got a booty call. CMBC Darling. Mike was waiting for the pullback. That never came. The stock doubled while he sat there waiting for a better entry point. May that serve as a lesson to all of you. Mike's still furious. Mike should be furious about it. You know, if he'd bought it on that day, he'd have been buying it under 400 bucks. I think it's $930 today. You like it, you buy it. Mike can be cheap sometimes. And then Corning

corning: the ai gorilla.

SPEAKER_00 21:32

is where the thesis became almost embarrassingly obvious in retrospect. Everyone knew Corning as the gorilla glass that protects your iPhone. But suddenly at the start of this year, Zuckerberg Meta signs a six billion dollar infrastructure deal for optical cable. A six billion deal with Corning, the Apple iPhone. Phone company and the CEO of Corning says next year the hyperscalers will be our biggest revenue source than Apple and the iPhone. And the market stared at that information for weeks, and then suddenly the stall rose 54% in a month. Nothing to do with iPhone screens no more. And Mike, with all that wisdom, he turns to me and says, You don't build 500 megawatt AI campuses and connect you don't connect them with yesterday's bandwidth. That's when it became blindingly obvious to even me. The winners weren't building applications, they were building infrastructure. The losers built toll booths. The winners built the roads and the rails.

put the whole fucking thing into bitcoin.

SPEAKER_00 23:01

Let's talk about the kid. There's a kid in the bar, he's 28 years old. Is that a kid? Apparently he's got $20,000 to his name. And he's sitting quietly in the in the corner of Bar Select. He's watching Mike and me like we're conducting some strange financial auca struck. He hasn't yet earned the right to fully comprehend. But he's inching closer. He's eavesdropping. And I know him and I point at him and I'm like, you, you, I shout, put it all into fucking Bitcoin. No lectures about diversification, no 60-40 portfolio, no patient compounding, no responsible financial adulthood. Put the whole fucking thing into Bitcoin. And if you can't emotionally survive the drawdown, get comfortable being average for the rest of your life. I can be a cunt sometimes. But I meant it. I meant every damn word of it. Because risk is what money, what it really means to you. Risk is what you see when you look in the mirror. And $20,000 sitting safely in a savings account, that doesn't liberate anybody. And that's the tragedy. I mean they're they're taught right here. The old wisdom is that passive investing, which has been remarkably profitable in the context of the SP 500. And on average you've compounded at 10% per annum. Now that kid, if he compounds it 10%, his 20,000, in 20 years it's gonna be 135,000. Sounds a lot, but it is nothing. It's nothing. Again, no humble gloating, but an entry-level one-bedroom apartment will set you back one and a half, two million dollars today in St. Bart's. 20,000 is here to zero, or you better have a hero play. If the markets behave perfectly, and that's a huge if because from these elevated levels, the chances of you compounding your equity return at 10% from this point, I mean experts will tell you, they won't tell you it's impossible, but they'll say that there is a very high, like a 99.99% chance that your return over the next 10-15 years, maybe nothing. But let's say he's disciplined, conventional, and lucky, and he gets 10% compound over the next 20 years. He'll be 48 years old and he'll be sitting on 130,000, still unable to buy the cheapest entry-level apartment. Today that department is one and a half million dollars, arguably two million dollars. So when I say in this world, you're either the hero or you are the zero, that's what I mean. The conventional advice industry, mostly zeros, a beautifully manicured system for guaranteeing your own mediocrity. And Bitcoin,

bitcoin the financial hyperscaler.

SPEAKER_00 26:08

let me talk about Bitcoin because Bitcoin rhymed with the entire thesis of the software and the hyperscalers. Let me tell you why. Because everyone thinks Bitcoin is a technology story. I don't. I think it's a hyperscaling story. And and I I go way back to 1982 and and the launch of the first SP futures contract. Most people think it was just another financial product. It it wasn't, it was an immense premeditated, let's call it synthetic scaling. The futures market allowed American equities to become something much greater in import and size than the underlying cash market itself. That's what was recognized by the geniuses on Wall Street in the late 1970s, that introducing a futures contract onto the SP would transform the US and rightly cement its position as head of the financial citadel, global citadel. It's the same logic as fractional reserve banking. Small base layer, massive synthetic expansion built on top. Basically talking about an infinite ocean of high-powered liquidity, fractional reserve banking, lifting futures into the stratosphere. It's size, my war is all about quantum of size. And fractional reserve banking, when you plug it into a futures contract, boy, you pump size. Don't believe me? Back then, back then SP was like a puny one trillion dollars size. Today's more than 50 times that. And so when I think Bitcoin futures, I think the same play. The financial market is not hiding its tracks, it's laying them down, it's laying down plumbing right in front of you. Because Bitcoin, it cannot fulfill its destiny as a cute little one trillion dollar asset that lots of guys in Anorac and computer keys discovered 15 years ago. If America is to retain its hegemon, it has to retain the world's currency. And it's going to require more synthetic scaling, especially for digital gold. It needs derivatives, collateral systems, futures architecture. Not America, but America and where it sees its place in deploying Bitcoin requires derivatives, collateral systems, and futures architecture. And when I think of

the kid with twenty thousand dollars.

SPEAKER_00 28:52

that in my sana moment, I'm thinking we are looking at planetary monetary plumbing. The plumbing of the planet, the monetary plumbing of the planet being plugged into the Bitcoin matrix. That is how you secure the macro hegemon of dollar cash.

be a trader and take the fucking money.

SPEAKER_00 29:15

Anyway, Mike, with the kid, with the Bitcoin futures, with the SAS apocalypse, the 1.6 trillion wiped off, the market dismissing wonderful software company earnings, the market feeling very existentialist, full of dread, bombs dropping on Iran, the Strait of Hormuz seemingly closed, the Americans setting out foxed by the crafty mullahs. Mike turns around to me and completely seriously he says, You should personally have five million dollars in Bitcoin. What are you gonna do about it? You see, for over a year he's been trying to convince me to sell this house Blanc Blue. And Mike loves my house. You see, what Mike recognizes, what he knows, he knows I bought the land. And again, these are numbers. This is this is not humble gloating. But I I bought the land in St. St. Bard's in 2013 for three million dollars. And I spent another three to four million building building the house. So seven million all in. And it's not just capital, it's become my identity. And we discussed London in the flight from the lunatic asylum. And I think I spoke about how property in these safer financial jurisdictions. I hope I don't rue saying that. But we we spoke about how there was scarcity in jurisdiction like St. Bart's and it was bidding prices higher. And St. Bart's prices have certainly outperformed what is now regarded as the abundance of problems stoking the London market. And so Mike, of course, was able to reduce the whole thing. He has a steely look thing, he pulls me close and he says, sell it today, and you get maybe over 40 million dollars. You make 30 net after everything, after you've paid off your mortgage, after you've paid your taxes. 30 million dollars, he says, that's not money. That's world-class optionality. And the motherfucker, he just stares at me and he says, Oog, you are supposed to be a trader. Take the fucking money. Mike tortures me like that all of the time. He says, a man needs a mission. Yeah, I say, tell that to my girlfriend. Right. Let's end this damn thing.

a colossal mistake; it was not.

SPEAKER_00 31:52

Colossus. Has this paper, has this video been a colossus error? I hope not. So meanwhile, while while Mike is trying to liquidate my emotional attachment to my physical and very real and very beautiful property here in St. Bart's, the AI machine is it's accelerating. It woke up, it's accelerating again. How is this possible? Because on May the 6th of this grand year, Elon Musk leased his Colossus to anthropic. And this is when my little essay stopped feeling like a technology yarn and it started to feel historical. Colossus was not a normal data center. Normal companies build 20,000 GPU clusters over two years. Elon, crazy motherfucker, built more than 200,000 Nvidia units in 120 days. That is a hell of a lot of power. That is a small as a city, as a city consuming electricity. Tesla, megapacks, liquid cooling. This is private infrastructure sovereignty. This is a small country. This was not a company expanding capacity. This was somebody privately constructing an industrial civilization before everyone else. And the extraordinary thing was just three months earlier, the protagonists were at loggerheads, each saying that the other was more evil, more dangerous. And then suddenly they're they're renting Elon's big fucking machine for like six billion dollars a year. Why? Because physics doesn't care about ideology or theology. Imagine building the transcontinental railroad. Let's

taggart transcontinental.

SPEAKER_00 34:05

go already Anne Randon at the struggled. Imagine Dagney had just discovered Howard Rock the lunatic. Imagine Howard Roque had already completed the entire transcontinental railroad before Dagney had even asked. Steel late, stations finished, locomotives waiting. It's like waking up on Christmas morning as a child, and there's a model railroad running across the floor. And anyway, and you say to Elon, I need it, and he says, fine, rent it. That's what Anthropic found in Memphis.

anthropic's growth escaped the lab and was replicating fucking fast

SPEAKER_00 34:40

Their revenue growth had escaped the infrastructure designed to support it. Anthropic's revenue growth had escaped the infrastructure plans designed to support it. Anthropic's revenue began to resemble a virus that had jumped the laboratory and was replicating very fucking fast. They'd planned for tenfold growth. First quarter of this year, they achieved 80-fold. Software businesses are just simply not designed. Software businesses are not supposed to grow like biological outbreaks. And because of that rate limit surged, the big customers, the guys buying tokens, panicked. People were fucking angry. And suddenly the supposedly reckless infrastructure project of

the embarrassment of clarity

SPEAKER_00 35:35

Elon, the Colossus, the Colossal Era, became the single most valuable asset in the universe. That's the embarrassment of clarity. Everyone thought Colossus was excessive, crazy mad, and it turned out demand itself was excessive, crazy mad. And that's what Mike and I were really discussing in Bar Select, that last day of February, that third day of March. Not software, not chat boxes, but constraints. Find the bottleneck. Own the bottleneck. Hold it while the market calls you insane. Because once intelligence becomes manufactured, everything upstream of intelligence becomes strategic. Everything power, glass, cooling, copper, bandwidth. That's where the new gatekeepers live. Jensen said it directly, and

compute is revenue.

SPEAKER_00 36:38

almost nobody in financial media understood the implication. Compute is revenue. Not compute enables revenue. Not compute supports revenue. Compute is revenue. In the old world, the server rack was it was a cost center. Remember, we had Michael Bury neck deep in the server rack examining the depreciation policy, telling us that the rack was being made obsolete quicker and quicker, that profits were overstated. In that old world, in Michael Bury's world, the server rack was a cost center. In the new world, the world we entered this year in 2026, the server rack is the product. The factory no longer manufactures the thing being sold. The factory is the thing being sold. And that changes everything. The hyperscalers understood this before the market did, which is why they spent $600 billion, which is why they'll spend a trillion dollars over the next 12 months. They were not being reckless, they were buying civilization scale positioning. And that's why the market was wrong for five weeks. And that's why it'll be right for the rest of the year. The same reason compute is revenue. Find the constraint on compute and own it. That's the only trade. Because civilization always reorganized themselves around whatever becomes scarce oil, railroads, electricity, and now compute.

somewhere in memphis. the take down.

SPEAKER_00 38:23

Somewhere in Memphis, a privately owned machine is consuming enough electricity to power a city. And somewhere in Kentucky, factories are manufacturing glass strands thinner than human hair. And somewhere in St. Bartholomew, two men sat at a bar and realized the world had already changed. I am you, Andrew. That was another asset capitalist. A discourse on the end of the software age from the hallowed bar, the bar select in Gustavia. I thank you, people. If you're liking the show, the human touch is the greatest touch. Tell the machine you love the show. Subscribe, give us four stars, check us out on Substack. Come. Come on the 2nd of August. Come and be my friend. I promise I will be yours forever. Bezus, hugs, and kisses. Good night from St. Bart's.