Better Offline

AI Is Worse Than The Dot Com Bubble: Part Four

12 min
Jan 30, 20263 months ago
Listen to Episode
Summary

Ed Zitron concludes his four-part investigation into why the AI bubble is worse than the dot-com bubble, arguing that AI startups are selling fundamentally unreliable software backed by unsustainable GPU infrastructure costs. He predicts massive financial collapse across Nvidia, major tech companies, Oracle, and venture capital, with societal harms exceeding the dot-com era due to widespread deception about AI capabilities.

Insights
  • AI bubble combines worst elements of both dot-com and current era: massive capital waste, unreliable products, and infrastructure with no alternative use cases once bubble bursts
  • Nvidia faces inevitable revenue collapse (40-90%) because GPU demand is entirely dependent on unprofitable AI spending that cannot continue indefinitely
  • Major tech companies (Microsoft, Google, Amazon, Meta) have accumulated hundreds of billions in depreciating GPU assets that will face massive impairments
  • AI psychosis is a widespread phenomenon where users convince themselves LLMs work through prompt manipulation, creating false productivity narratives
  • Unlike dot-com bubble which created useful infrastructure, AI data centers will become empty shells with no viable alternative use cases post-collapse
Trends
Unsustainable GPU infrastructure economics will force major tech companies to write down billions in assets within 2-3 yearsVenture capital funding for AI startups will effectively freeze for years post-bubble burst due to massive lossesPrivate equity and banking sector exposure to $178.5B in US data center deals poses systemic financial riskWidespread credibility collapse among tech influencers, executives, and media who promoted AI capabilities beyond current realityShift from asset-light to asset-heavy business models has made mega-cap tech companies vulnerable to infrastructure depreciationOracle's $248B lease obligations and $56B debt for OpenAI partnership creates bankruptcy risk if OpenAI failsGenerative AI adoption narratives driven by manipulation of tool outputs rather than genuine productivity improvementsData center construction projects will face massive cancellations and defaults across real estate and construction sectors
Topics
AI Bubble Economics and SustainabilityNvidia Revenue Collapse ScenariosGPU Infrastructure Depreciation RiskTech Company Asset ImpairmentsVenture Capital Funding FreezeData Center Construction DefaultsOpenAI Financial ViabilityLarge Language Model ReliabilityAI Adoption Narratives vs RealityDot-Com Bubble Comparison AnalysisOracle Bankruptcy RiskPrivate Equity Exposure to AI InfrastructureGenerative AI Productivity ClaimsTech Industry Credibility CrisisSystemic Financial Risk from AI Bubble
Companies
Nvidia
Central to collapse thesis; predicted to lose 40-90% revenue when AI spending stops, causing S&P 500 impact
Microsoft
Faces multi-billion dollar impairments on GPU assets acquired for AI infrastructure investments
Google
Faces multi-billion dollar impairments on GPU assets acquired for AI infrastructure investments
Amazon
Faces multi-billion dollar impairments on GPU assets acquired for AI infrastructure investments
Meta
Faces multi-billion dollar impairments on GPU assets acquired for AI infrastructure investments
Oracle
Faces potential bankruptcy due to $248B lease obligations and $56B debt for OpenAI data center deal
OpenAI
Will run out of money within one year; $300B 5-year deal with Oracle unsustainable; IPO would result in investor losses
Anthropic
AI startup that if goes public will collapse under weight of bullshit assumptions and burn all partners
CoreWeave
GPU infrastructure provider whose collapse would trigger cascading failures across AI ecosystem
People
Ed Zitron
Host and primary analyst; conducted extensive research into dot-com bubble to compare with current AI bubble
Quotes
"I'm really fucking worried. I'm not doing this to be contrary and I'm seeing these things and I'm worried more people aren't seeing them."
Ed Zitron
"There is no squaring the Nvidia circle. It is inevitable even if AI somehow worked out that companies will have to stop buying as many GPUs and Nvidia cannot afford for them to slow down at all."
Ed Zitron
"This is a fucking calamity. I am fucking terrified because everything I've read about the past makes me certain that this will be worse."
Ed Zitron
"I think large language models are a global gaslighting experiment, where the test is to see whether inefficient and questionably functional software can convince you that it works."
Ed Zitron
"The AI bubble is happening in a remarkable era of digital information and connectivity that allows us to process and analyze details at scale and with speed and everybody who inflated this era should be made to experience inextricable permanent shame."
Ed Zitron
Full Transcript
This is an I Heart Podcast. Guaranteed Human. Today, new BT group customers only 62% UK availability terms apply. Quarzo Media. Hello and welcome to Bear Off Line. I'm your host Ed Zitron. Today is the final part of Bubble Week and our four-part dot com bubble special. I am wearingly invested in this show. I put so many hours into it. This is the most I've done in the January. I'm having the time of my fucking life. But this four-part has been rough and it's not been rough for the recording. It's rough because every time I go over these details, they fill me full of anxiety. I don't enjoy being right here. Sure I get the satisfaction. Sure I get the clicks and all that. But like... It sucks. I'm really fucking worried. I'm not doing this to be contrary and I'm seeing these things and I'm worried more people aren't seeing them. And I'm worried that the after effects will be so much worse than the dot com bubble. So this four-part was somewhat drawn from a premium piece I did a few weeks ago called inventively. This is worse than the dot com bubble. I really suggest you subscribe to my premium newsletter or be in there. But if you don't, I've had a little mustard on it and truncated it for this because it needed to be different in voice. Nevertheless, you should subscribe. But when I was writing that, I actually went into it trying to prove myself wrong. I spent hours and hours reading old articles through broken paywalls and archived.org links in the hopes that I'd find some sort of positive spin. A way that the AI bubble wasn't as bad, some sliver or nugget of joy, some shred of hope. And everything I've read has made me more worried, not less. The real problem is that there are similarities, but only with the worst parts of both bubbles. We've got stupid amounts of money being sunk into stuff. We've got stupid companies that want a IPO. We've got hopeless startups due start by credulous venture capitalists and marked it by even more credulous journalists. Except this time we have the worst margins this side of a vineyard planted in Las Vegas, Nevada. And despicable boosters that will go to war with you to try and prove you're wrong using ghost stories, foul retails and bixie dust. I realize I'm being a little cavalier and colloquial about this and that's because I feel emotionally drained is the way I'd put it. Things are so much worse than I thought. I really thought I'd find something good. I dig into the past and find that they were direct. Obviously similarities that would say, okay, these GPUs will be useful. There will be something we've built here. I can't find it. I'm actually now convinced that things are going to be much worse. The Dock-Hornbubble was relatively distributed and created useful infrastructure. The ideas of some of the most prominent Dock-Horn startups were valid, but based on stupid hypergrowth business models that didn't make sense and were only sustained through mass hysteria and ignorance. By comparison, AI startups are fundamentally illogical selling unreliable software that's best known for its consistent mistakes in active harms back by the corrosive and ever-worseening infrastructural costs of GPUs. And they're the customers of the largest company on the stock market that can only keep its valuation on which the markets depend by selling more of it and expensive to buy and more expensive to install item that only appears to lose its customers money. There is no squaring in video circle. It is inevitable even if AI somehow worked out that companies will have to stop buying as many GPUs and Nvidia cannot afford for them to slow down at all. They must speed up and they must do so forever. There's also no other business model for GPUs that scales. None. There is no use case. This is not useful infrastructure. Once the AI bubble bursts, these data centers cannot be used for other things, at least not things that are going to make more than a couple of single digit millions in total. They will have to be emptied out and replaced with useful gear. They will be those that are finished and I think there's going to be a bunch of done construction. They will be powered shells which are just empty buildings with power attached. And really I don't think that's going to be across the board. I think we're going to have a lot of failed construction. And throughout all of this bullshit, the rest of the Magnificent 7 has seen their stock values soar off the back of pure hype because they weren't making a dime of profit from selling artificial intelligence services. And they've done so by acquiring hundreds of billions of dollars of physical assets. The whole point of these fucking companies was that they were asset lie and cash heavy. Now they're full of fucking GPUs that depreciate. This is a fucking calamity. I am fucking terrified because everything I've read about the past makes me certain that this will be worse. The harms will be longer and nobody seems to want to accept the true logical end point of my arguments. So I'm going to kind of lay them out. Depending on how right I am, I'm worrying about the following things. And video will lose anywhere between 40% and 90% of its revenue in a relatively short amount of time. Couple of quarters may be which will annihilate the value of the S&P 500 because the stock will go down because the market will go, ah! Microsoft Google and Amazon will face massive multi-billion dollar impairments within the next few years. Forgot to say meta but they're included too. Now this one is funny but will still be bad. Oracle potentially face bankruptcy or at least the severe horrendous series of layoffs and restructuring efforts. Try and write the shift. After agreeing to $248 billion in lease obligations and $56 billion in debt, primarily to build data centers to get the revenue from its $300 billion, $5-year-long deal with OpenAI, a company that's going to run out of money in the next year. Venture Capital is going to take such a severe haircut from the AI bubble that it will effectively kill any chances of raising a round off the series beefer many many years. I'm really worried about the reported $178.5 billion in US data center deals going underwater. I don't think they're going to be paid and it's going to lead to massive losses of private equity and banks which will lead to a chill in effectively all debt-free. It's going to be effectively all debt markets and massive job losses across the many many construction projects that have started as a result of the AI bubble. And if OpenAI and Anthropic can actually manage to go public, I think every investor is going to lose out as they collapse under the weight of their bullshit assumptions. Every partner, every company relying on them, every infrastructure partner, every AI start a bill on their tech will be burned. I also believe that we as a society will face a violent paradigm shift as the result of the AI bubble burst thing because it revealed how many people just don't understand stuff. The amount of bosses, influencers, movie stars, politicians and other individuals of note that have fallen behind generative AI and claimed to get in again that it can do things that it can't, but perpetuated myths that will lead to massive societal harms both 10 in the future have revealed themselves to be fucking frauds. Travis Perkins Stratford has now moved to Lentenstone. We've got everything from timber, bricks and blocks to loads of tools, plants and equipment for hire along with benchmarks, kitchens and joinery. We keep all your essentials in stock, deliver to site and upgrade trade deals too. Your new Travis Perkins branch is packed with even more products, more space and even better service. So for all your building materials, hire requirements and kitchen needs. Simply head to Travis Perkins and benchmarks on Joseph Ray Road, Lentenstone. Do we want matters for the trade? Master any task with the all-electric Ford e-Transit Custom Limited with lower running cost for your business. Selected dealer stock is now available with 0% APR on 4-year Ford options and a 7,000-pound customer saving until the end of March at participating dealers. Search Ford e-Transit Custom. Ready, set, Ford. Finance, subject to status. Turn up your taste buds with dolmyo-intensify pasta sources. From mild creamy garlic and black pepper. To bold smoky garlic and sundry tomato. And spicy smoked paprika and chili. Woo! Try dolmyo-intensify pasta sources with serious flavour. Dolmyo-yere. It's been obvious since 2023 that Generative AI flat out does not do what many people say it does or will do. And obvious since 2024 that the economics didn't work. Yet still, every single goddamn day we are surrounded by people screaming in our faces that AI is the future, that AI will replace us, that AI is the panacea to every fucking problem, even though it only seems to create new ones. I also believe that AI psychosis is a prevalent thing on a scale that few people want to discuss. There's something about the process of using a large language model that convinces people that AI is powerful because they manage to bonk it on the head enough times to make something useful come out. Seemingly intelligent people will spend hours and hours making Claude code burp out cause our useful cause our functional software. Claiming that the productivity gains are obvious even when if you sat down and worked at how long it took them and how much the API rates for the software would be if they weren't being subsidised, they could probably have just learned to code in the first place or paid someone else to do it. And no, please stop this thing of, oh it's cheaper than hiring a developer, did you go out and ask a developer, did you actually get a quote? Do you actually know what you're talking about? Do you know a fucking thing about you're talking about? You see, I think large language models are a global gaslighting experiment, where the test is to see whether inefficient and questionably functional software can convince you that it works because you manage to manipulate its handles with the right prompts, explaining away failures as you go, and getting increasingly agitated with those who don't agree with you, that you're living in the future. This series forced me to live in the past for hours and hours and hours and all it taught me is that people are desperate to compare this era to another as a means of explaining away the bad habits and species ideas of a tech industry run by fucking influences and management concerns. By comparison, the dotcom bubble is relatively sensible, built on good ideas executed by greedy people rushing to do things based on ideas they didn't understand and analyses that were utter bullshit. The AI bubble is happening in a remarkable era of digital information and connectivity that allows us to process and analyze details at scale and with speed and everybody who inflated this era should and by my fucking sword will be made to experience inextricable permanent shame for the horrors that follow as a result. And believe me, even if I'm half right, things will likely be so much worse than when the dotcom bubble burst. I don't see a scenario where anything survives the collapse of core weave let alone the collapse of open AI. I don't see a scenario where any more than a few AI stops survive, even if that happens, there won't be any customers with those GPUs. Not that anywhere near enough existed to fill even a tiny percentile of the demand that currently exists or for the capacity we've already built. On some level, I hope I'm wrong. I just don't know her anymore. I really don't. Thank you so much for listening. It's been an intense January with this with CES with the inti-tiponatural crisis. I'm probably going to take the week off next week and run a cool zone media rewind to give my aching brain and lungs a break. Thank you so much for listening. Who knows though some bullshit will probably happen because it's earning season when this runs. Some crazy shit will happen or maybe it'll be normal I could just rest. And if that happens I'll probably end up oinking and squawking into the mic as I always do. Regardless, you'll hear my voice one way or another. Thank you so much everyone. This has been Bubble Week. Thank you for listening to Better Offline. The editor and composer of the Better Offline theme song is Matt Orsowski. You can check out more of his music and audio projects at Matt or Salski.com, M-A-T-T-O-S-O-W-S-K-I. You can email me at easy at BetterOffline.com or visit BetterOffline.com to find more podcast links and of course my newsletter. I also really recommend you go to chat.whizyoured.at to visit the discord and go to our slash Better Offline to check out our Reddit. Thank you so much for listening. Better Offline is a production of Cool Zone Media. For more from Cool Zone Media, visit our website CoolZoneMedia.com or check us out on the iHeart Radio app, Apple Podcasts or wherever you get your podcasts. Crisp, vibrant and bursting with citrus. Villain-Maria's New Zealand Sauvignon Blanc is the perfect wine made to be enjoyed on every occasion. Whether you're soaking up the sun in your garden, hosting a backyard barbecue or unwinding after a long day. The zesty lime and lush tropical fruits are always delicious. Try Villain-Maria Sauvignon Blanc, a vibrant New Zealand wine that's perfect for every occasion. Available at all good wine retailers. Another party invite? Well, here's a way to make their big day feel even more special. With small shops in Etsy, you can discover thousands of original birthday presents. Like a handmade leather camera strap to celebrate your photography upset friend or maybe a custom top for that Pisces in your life. From the personalised to the practical, we've got you covered with millions of active listings to choose from. Birthdays don't celebrate themselves. Shop at Etsy.com and discover your perfect find today. Hey, Vee, we see we see need to come back here together. You'd love Morocco. We went hiking in the Atlas Mountains yesterday. Got you a little something from the market too. Mainly so I could use them as when you do GD baggage. Hiya! Hiya, Lee! Anyways, I hope you're all good. I'll belly later on the wifi. I want to show you the hotel. It's like a proper re-ad with a spa. Mishoo! 22 kilograms of baggage included. Nothing beats Jetsoo. Subject to availability in editions. This is an iHeart podcast. Guaranteed Human.