Better Offline - Monologue: The AI Data Center Overbuild

Episode Date: September 25, 2026

In this week's Better Offline monologue, Ed Zitron talks about how $200bn to $300bn of AI GPUs are sitting in warehouses, how hyperscalers like Microsoft are overstating their AI capacity, and why the AI data center overbuild is magntiudes worse than the Dot Com era fiber buildout.  Where’re All The AI Chips? https://www.wheresyoured.at/wherere-all-the-ai-chips/Dario quote: https://www.dwarkesh.com/p/dario-amodei-2#:~:text=Basically%20I%E2%80%99m%20saying,that%20much%20compute  Save $10 off a year of my premium newsletter: https://edzitronswheresyouredatghostio.outpost.pub/public/promo-subscription/gzqwkv54e1 YOU CAN NOW BUY BETTER OFFLINE MERCH! Go to https://cottonbureau.com/people/better-offline and use code FREE99 for free shipping on orders of $99 or more. --- LINKS: https://www.tinyurl.com/betterofflinelinks Newsletter: https://www.wheresyoured.at/ Reddit: https://www.reddit.com/r/BetterOffline/  Discord: chat.wheresyoured.at Ed's Socials: https://twitter.com/edzitron https://www.instagram.com/edzitron https://bsky.app/profile/edzitron.com https://www.threads.net/@edzitron Email Me: [e-mail] omnystudio.com/listener for privacy information.

This is an IHeart podcast. Guaranteed Human. Didn't catch the latest Roland Martin unfiltered podcast? Here's what you missed. There should not be a single law enforcement agency. It does not have body cameras. It's real. Black farmers have been under attack.

This is just the latest example of them just slapping DEI on anything. It's raw. And I'm sitting here going, they are playing y'all for fools. Catch Roland Martin's daily commentary on the Black Information Network. And download Roland Martin Unfiltered on. on the IHart Radio app, Apple Podcasts, wherever you get your podcasts. The new NFL season is here,

and you should be listening to NFL Daily as we march along to Super Bowl 61. It is in the name, NFL Daily. You'll have fresh content in your feed every day all season long. That game-winning drive, maybe it's Herbert, maybe it's Gino, maybe it's Mahomes. We'll have the highlights. If Fernando Mendoza or any of the rookies are balling out,

will break down the tape. Join me, Greg Rosenthal, and an all-star cast of co-hosts as we preview and recap every game. Listen to NFL daily on the IHeart Radio app, Apple Podcasts, or wherever you get your podcast. Whether you're a seasoned NFL fan or new to the game. There's one place to keep up with all of it. The League's newest podcast. The NFL Report, hosted by me, Andrew Siciliano.

It is your home for everything football. Breaking News, expert analysis, game picks, and hear from your favorite players too. Join me at an all-star cast of experts for everything you need to do. know from around the league. Get new episodes of the NFL report Monday through Friday all season long. Listen to the NFL report podcast on the IHeart Radio app, Apple Podcasts, or wherever you get your podcasts. The World Cup is over, but if you caught the soccer bug and aren't sure what to do now,

check out our podcast, Big Kick Energy. I'm Cameron Dicker, the kicker from the Los Angeles Chargers, and I'm here with Timmy Tillman, current L-AFC star player. So, we know a little something about kicking balls. Each week we'll tackle things like who's winning the Premier League. What do you think is harder? Scoring a penalty kick or scoring a field goal? Join us each week for a quick chat with your soccer-loving

bros about the best sport in the world. Soccer, just don't tell my coaches I said that. Listen to Big Kick Energy on the I Heart Radio app, Apple Podcasts, or wherever you listen to podcasts. Hey, it's Alec Baldwin. This season on my podcast, here's the thing. I talked to actor Stephen Root. I'm a people watcher.

You see that guy going down to the ring, yeah, I can use that. And Open Igloo co-founder, Alia I love staying on top of what is going on in our city, what is on renter's minds, and taking all of that knowledge to build a platform that hopefully is going to make New York City better in the long run. Listen to Here's the Thing on the IHeart Radio app Apple Podcasts or wherever you get your podcasts. What up y'all. This is your main man, Memphis Bleak right here. Host of that rock solid podcast.

And each week, we bring in you exclusive looks inside music. everything happening inside the culture. Some of the best conversations were the biggest names in the game. You had a little bit more pressure to be good. And they didn't care if you was a girl. Like, you wasn't getting no points. If you're not trash, you better be not just good. You better be great.

Listen to Rock Solid on Black Effect Podcast Network, iHeartRadio app, Apple Podcast, or wherever you streamed podcast. Hey, everybody, it's me, Debbie Kama Bell. This season on my podcast, who's with, me, I speak with Bruce Lee's daughter, Brandon Lee's sister, and my friend, Shannon Lee. Back to your initial point of stewarding my father's legacy and giving my energy to that. You're lending your life to stewarding somebody else's life. Listen to Who's With Me with with Debbie Chmobel on the IHeart Radio app, Apple Podcasts, or wherever you get your podcasts.

AllZone Media. Hello and welcome to this week's better offline monologue. I'm your host, Ed Zittron. As ever, subscribe to the newsletter, Premium, you know all that good stuff. I'll have the links in there. And this week I want to talk about actually a free newsletter I put out called Where Are All the AI Chips? And it turns out that the answer is either in warehouses or unpowered data centers. In an investigation with The Guardian, I should own over there, worked on it with me, she's awesome.

I reported that Microsoft only had 2.2 million GPUs in service. far less than people believed. And while I couldn't report it at the time because I still had to run down some leads, the total power of the chips was estimated about 1.993 gigawatts of capacity, with the overall cost of those GPUs installed somewhere in the region of $50 to $60 billion.

A few weeks ago, Bloomberg reported that despite reporting that Microsoft had added a gigawatt in each of the last three quarters, and by reporting, I mean the things that Microsoft literally said on their earnings call, that only two gigawatts out of their ownings call, total 12 gigawatts of data center capacity was specifically for AI. In other words, despite having spent over $265 billion on capital expenditures since the beginning of 2022, Microsoft has only put

about $50 billion worth of GPUs into service, and I estimate that as much as $106 billion worth of GPUs and associated hardware are now sitting either unpowered in data centers, or incomplete data centers or in warehouses, and that very little AI computers coming online throughout the entire industry. Microsoft's CEO, Satchin Adela, had sort of admitted this in November of last year, when he told an interviewer that he had a bunch of chips sitting in inventory that he couldn't plug in, and that's a quote, though he didn't mention the sheer scale of the warehousing, or indeed how little he'd actually turned on. Well, I'm a curious little critter, so I went looking for more evidence of the problem outside of Microsoft, and found it buried in the balance sheets of

multiple hyperscalers and neoclouds under the construction in progress line on the balance sheet, which is specifically where companies bury their uninstalled GPUs and incomplete data centers. And this number has grown across a ton of them, well, by a remarkable amount. Google, Meta, Oracle, Amazon, SpaceX and Tesla, neoclounds like Corweave and Iron, and co-location companies like Core Scientific and Applied Digital, all have about $374 billion of construction in progress, a figure that's likely lower than the two. true number because Amazon's contribution, which is about $71 billion, is only current as of the end of 2025, and there have been two now, very soon three more quarters of that. And the company only reports

annual, like I said. A large chunk of that larger CIP number is Google's, which sits around $122 billion, which is truly shocking. And in every case, by the way, the number has grown, every single time it's been reported for the last 12 quarters, with some fluctuation in the case of Amazon, because of their logistics operations, but really right now, it's only growing. Now, not everybody reports construction in progress, so that number doesn't include Microsoft, Firmus, Sharon, Equinex, Nebius, poolside, private operators like Vantage, any of the sovereign AI buildouts in the Middle East or Europe, the private projects built for Open AI and Anthropic, or any of Oracle or META's off-balance sheet construction projects, which were likely, I think, numbering in tens of billions of

dollars and because these are special purpose vehicles kept off of their balance sheet, that wouldn't be in construction and progress. Now, if I had to guess, the total construction and progress number is somewhere between $400 and $500 billion. And based on everything I've analyzed, I estimate that approximately 50% of all AI hardware and chips that have been sold is being warehoused. And will take more than two years to fully ingest, which means that basically everything you've seen in Vida's sell has been a pre-order campaign that arrives sometime in 20,

2028, 2029, or maybe beyond that. In other words, I think Invidia is made somewhere between $200,300 billion in revenue on stuff that's been sold probably 24 to 36 months in advance and that any scarcity around AI compute is a result of most of the capacity being sold immediately to Anthropic and Open AI, leaving very little for the rest of the world, with more capacity taking agonizingly long to build and not being prioritized unless you're one of the AI labs.

So what does all of this mean? First of all, basically any announcement of how much capacity a hyperscaler has built is suspicious. The fact that Microsoft is intentionally obfuscating how much capacity, AI capacity, I mean, it has built is a sign that the entire AI industry is doing so, and that the available AI capacity is much, much, much smaller than we think. All of those estimates of 12 or 15 gigawatts coming online? Bullocks. Whank?

Tosh. I don't know. I can't remember other terms for it. Second, if Microsoft, one of the most well-capitalized and experienced data center developers in the world, is having trouble bringing the majority of its AI capacity online, everyone is. Really, I think Amazon is the only company, maybe Google, that has comparable experience, let alone more. And I think Microsoft was also the first AI supercomputer powered by GPUs back in 2020 for OpenAI. There's probably someone who did it before, just to mess with me. Nevertheless, if they're failing, I think everyone is.

And lastly, any beliefs that anyone has about the insatiable demand for AI compute are completely distorted because very little actual capacity is coming online, like gigawatts less than people believe, and almost all of it is flowing directly to Anthropic and Open AI, creating an illusion of scarcity when it's actually just too unsustainable venture-backed AI labs sucking up whatever exists. And this also means we're in a terrible overbuilt situation. Right now, everyone building more AI data. a center capacity is doing so because of the massive revenue backlogs across Microsoft, Google, Amazon, Oracle, and Corweave, ignoring the fact that anywhere from half to 75% of those backlogs

are directly from Anthropic and OpenAI, neither of whom can actually pay for them, and any hyperscaler signing contracts to give that capacity to them can probably cancel the agreements. To the outside world, everything looks like a thriving industry with tons of demand, both for AI GPUs, which they think are turning into money immediately, and for AI compute. with Microsoft's clearly crooked statements around adding a gigawatt of capacity every single quarter, creating the illusion that its revenue growth is from adding all of that capacity and from spending all of those capital expenditures and from diverse demand on top of it. Rather than mostly from OpenAI's compute spend, which made up 70% of Microsoft's AI revenues in fiscal year 2026,

and 7% of its overall revenues in that same fiscal year. To be very specific about what I mean, people are assuming that lots of computers coming online every quarter and that all of that computer is being immediately rented out for high rates. As a result, they think that building more AI data centers is an obvious choice and akin to printing money because they believe there's unbelievable amounts of pent-up demand waiting to be met and that meeting that demand will be as simple as bringing a data center online which Microsoft is, from the outside, proving can be done a gigawatt a quarter. except it's not AI data center capacity that's coming online.

What's actually happening is very little capacity is coming online for AI services, and what little does is going straight into the hands of two companies with near-infinite resources provided by venture capital, and in some cases the hypers themselves, which allows them to sign contracts to take up massive amounts of capacity, and indeed fill those revenue backlogs. Microsoft is saying that it's bringing data center capacity, not AI-specific data center capacity, online a gigawatt at a time,

but it appears the actual AI capacity is, at best, coming on at a rate of maybe tens or hundreds of megawatts a quarter, and that really is the best case. Perhaps there are others that are doing it faster, but I actually can't find evidence of them. I'll be honest that I have had this story in my head for weeks and was hesitant to jump the gun because of its ramifications, as it appears that everyone's got this story wrong.

This is magnitude's worse than the dot-com bubble. To make the obvious comparison, this is akin to the fiber-optic cable not even being laid in the ground, or only being halfway connected, except the actual demand for AI infrastructure is so thoroughly distorted by OpenAI and Anthropic that there's anywhere from five to 50 times the amount of capacity in planning than there is demand for it. And even then, that demand is inflated. To make matters worse, it's not going anywhere after this. There is no situation where this is going to be used at the scale it's built. It will also be just as expensive to finish an AI data center in two or three or five years

and just as expensive to run those GPUs. And who knows how many generations will be in there. But I don't know, in my opinion, my esteemed opinion, I don't think we make it past Vera Rubin or Feynman, which is the generation after Viboruban. And even then I have questions about whether Feynman actually ships. Who knows, though? Things are crazy.

I'll also be clear that, as I discussed with Paul Kodroski this week, that there is no post-bubble economy for GPUs, at least at the scale they're being built, because the cost of serving inference is similar to an airline with high fixed costs. You as an operator must buy capacity, hundreds or thousands of GPUs over a certain period of time, whether or not you actually have the demand. While there might be a situation where you could eke out a profit of some sort with the perfect amount of customers, anything above the demand you foresaw means an unstable and unreliable service and turning customers away at the door, and anything below consumes every ounce of margin you might have. Dario Amadei himself said as much in an interview back in February with

Dwarquish Patel, saying that in a theoretical scenario, he could buy one trillion dollars of compute that starts at the end of 2027, but if his revenue wasn't a trillion dollars, if it was even $800 billion, there was, and I quote, no hedge on earth that could stop him from going bankrupt. These are the underlying economics of every single company running an AI service, and their equal parts brittle and volatile. Thanks for listening. I'll see you next week. Can't catch the latest Roland Martin Unfiltered podcast?

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

Fools. Catch Roland Martin's daily commentary on the Black Information Network. And download Roland Martin unfiltered on the IHart radio app, Apple Podcasts, wherever you get your podcasts. The new NFL season is here. And you should be listening to NFL Daily as we march along to Super Bowl 61. It is in the name NFL Daily. You'll have fresh content in your feed every day all season long. That game winning drive, maybe it's Herbert, maybe it's Gino. Maybe it's Mahomes. will have the highlights. If Fernando Mendoza or any of those rookies are balling out, will break down the tape.

Join me, Greg Rosenthal, and an all-star cast of co-hosts as we preview and recap every game. Listen to NFL Daily on the IHeart Radio app, Apple Podcasts, or wherever you get your podcast. Whether you're a seasoned NFL fan or new to the game, there's one place to keep up with all of it, the league's newest podcast, the NFL report. Hosted by me, Andrew Siciliano,

it is your home for everything football. Breaking News, expert in out. game picks and hear from your favorite players too. Join me at an all-star cast of experts for everything you need to know from around the league. Get new episodes of the NFL Report Monday through Friday all season long. Listen to the NFL Report podcast on the IHeart Radio app, Apple Podcasts, or wherever you get your podcasts. Hey, it's Alec Baldwin. This season on my podcast, here's the thing.

I talk to actor Stephen Root. I'm a people watcher. You see that guy going down to the ring though? Yeah, I can use that. and Open Igloo co-founder, Alia Mohamed. I love staying on top of what is going on in our city, what is on renter's minds, and taking all of that knowledge to build a platform

that hopefully is going to make New York City better in the long run. Listen to Here's the Thing on the Iheart Radio app Apple Podcasts, or wherever you get your podcasts. The World Cup is over, but if you caught the soccer bug and aren't sure what to do now, check out our podcast, Big Kick Energy. I'm Cameron Dicker, the kicker, from the Los Angeles Chargers. And I'm here with Timmy Tillman, current L-AFC star player.

So, we know a little something about kicking balls. Each week we'll tackle things like, who's winning the Premier League this year? What do you think is harder? Scoring a penalty kick or scoring a field goal? Join us each week for a quick chat with your soccer-loving bros about the best sport in the world. Soccer. Just don't tell my coaches I said that. Listen to Big Kick Energy on the IHeart Radio app, Apple Podcasts, or wherever you listen to podcasts.

What up, y'all? This your main man, Memphis, Bleak, Rite. Hey here, host of that Rock Solid podcast. And each week, we bring in you exclusive looks inside music. Everything happening inside the culture. Some of the best conversations were the biggest names in the game. You had a little bit more pressure to be good. And they didn't care if you was a girl.

Like, you wasn't getting no points. If you're not trash, you better be not just good. You better be great. Listen to Rock Solid on Black Effect Podcast Network, I Heart Radio app, Apple Podcast, or wherever you streamed podcast. Hey everybody, it's me, Debbie Kamaabelle. This season on my podcast, who's with me, I speak with Bruce Lee's daughter, Brandon Lee's sister, and my friend, Shannon Lee.

Back to your initial point of stewarding my father's legacy and giving my energy to that. You're lending your life to stewarding somebody else's life. Listen to Who's with me with Debbie Kmartbell on the IHeart Radio app, Apple Podcast, or wherever you get your podcasts. This is an I-Heart podcast, guaranteed human.