The Four Horsemen of the AI Apocalypse | TCAF 257

The "Rot Economy" of AI [0:00]

• Ed Zitron, founder of EZPR, critiques the current AI boom as a "rot economy" driven by unsustainable capital expenditure and media hype.
• He argues that companies like OpenAI and Anthropic, despite projecting massive future revenues, lack current profitability and are propped up by venture capital and hyperscaler spending.
• This creates an "illusory demand signal" for AI infrastructure, particularly GPUs, leading to astronomical data center builds that are financially precarious.

  "The problem isn't finding well I mean actually it is becoming one now this a lot of this was I think that the um the one big customer the 16 one I think it was spacex"

The Media's Role in AI Hype [0:00]

• Zitron, a former PR professional, became disillusioned with tech companies' focus on growth over users and their rapacious monetization strategies.
• He observed a pattern of hype cycles, from Clubhouse to the Metaverse and crypto, exacerbated by social media and media coverage.
• AI, he suggests, is particularly special due to the 2022-2023 tech depression, offering a new avenue for growth and business creation.

  "And the thing oh yeah of course 2020 they were selling virtual land I said I know but I can't virtual real estate"

OpenAI and Anthropic's Financial Precariousness [0:00]

• OpenAI and Anthropic are presented as unprofitable startups that require constant fundraising to sustain their operations and massive compute commitments.
• Their projected revenues for 2030 are astronomical, but their current financial performance and the increasing cost of compute raise serious doubts about their viability.
• Zitron highlights that Microsoft's AI revenue is heavily dependent on OpenAI, making the startup an existential liability.

  "And on top of that they're taking up 90 of ai infrastructure this is creating an illusory demand signal because right now it's difficult to get gpus"

Hyperscaler Dependence on AI Startups [0:00]

• Major hyperscalers like Microsoft, Google, and Amazon are increasingly reliant on spending from OpenAI and Anthropic for their cloud revenue growth.
• Analyst expectations suggest hundreds of billions of dollars in cloud revenue are tied to these two startups.
• Zitron argues that without this AI-driven spending, the growth stories of these tech giants would significantly falter.

  "The problem is is that for now everything you'd look a quarter two quarters in the future wow money not money good money here"

The "Run Rate" Scam [0:00]

• Zitron criticizes the use of "run rate" as a metric by AI companies, arguing it's an undefined and untrustworthy measure of business performance.
• He contrasts it with "annual recurring revenue" (ARR), which is a more established metric for subscription-based businesses.
• The acceptance of "run rate" by investors is seen as an insult to their intelligence, masking the true financial instability.

  "The problem is is anthropic and open ai are both analyzing token spend which is not a record it's not a software subscription it's not recurring"

Nvidia's Central Role and Risks [0:00]

• Nvidia is at the heart of the AI story, with its GPUs being essential for AI development and data centers.
• However, a significant portion of Nvidia's revenue comes from a few hyperscale customers, making it vulnerable to their spending shifts.
• Zitron questions the sustainability of Nvidia's growth, given the immense capital required and the potential for demand to plateau or decline.

  "The point i made about microsoft's ai revenue is where this comes in so fiscal 26 which just ended 34.33 billion dollars of ai revenue"

The Data Center Build-Out Bubble [0:00]

• The construction of AI data centers is described as a "physical infrastructure build-out bubble," analogous to the 1873 railroad boom.
• These data centers are massive, power-hungry facilities requiring enormous investment.
• Zitron expresses concern about the speed and scale of this build-out, especially given the uncertain demand and the strain on power grids.

  "Each one of these even the smaller ones are like is that the oracle one yeah it's one of the one of the ones in texas and –"

The "Neoclouds" and Debt Dependence [0:00]

• Companies like CoreWeave, which rent AI GPU compute, are described as "neoclouds" that are "horribly unprofitable" and heavily reliant on debt.
• They need to perpetually raise billions of dollars to fund their operations and GPU purchases.
• Zitron believes that if these companies cannot raise debt, it could signal a major problem for the AI ecosystem.

  "And so one of those companies doesn't raise money it's going to be they try and they fail interesting"

The "Rot Economy" in Action: Meta and Google [0:00]

• Zitron criticizes Meta (Facebook, Instagram) and Google for prioritizing growth over user experience, deliberately degrading their products to increase ad revenue.
• He cites internal documents and antitrust trial evidence to support his claims of intentional product deterioration.
• This "rot economy" approach, he argues, makes products worse to extract more value from users.

  "The point i'm making is however you feel about this ethically it has a natural end point where you engine you change the company from making a product that you sell and you turn it into a growth engine"

The Problem of AI Cost vs. Value [0:00]

• Despite the immense cost of AI development and infrastructure, Zitron argues that the current output is "unremarkable" compared to its price.
• He questions the value proposition of LLMs, noting their tendency to make mistakes, "hallucinate," and require constant updates.
• The cost of intelligence has come down, but the models themselves are becoming more expensive to run due to increased token burn.

  "The thing is you can't divorce the technology from that you cannot what it does today is unremarkable compared to its cost and it's about what I would expect"

Divergent User Experiences with AI [0:00]

• Zitron notes the starkly different experiences users have with AI, contrasting his skepticism with those who find AI miraculous for research and automation.
• He uses AI for research and code generation but remains unimpressed given the immense costs involved.
• The efficiency gains, he argues, are often overstated or come at a high price.

  "The thing is that doesn't impress you I doesn't impress me for the cost for like a trillion dollars and tens of billions of dollars to get to a better search"

The "Existential Threat" Narrative [0:00]

• Hyperscalers feel an "existential threat" that if they miss the AI wave, their existing data center businesses will become worthless.
• This fear compels them to double down on AI investments, even if the economics are questionable.
• Zitron believes this is a self-perpetuating cycle driven by a fear of being left behind.

  "They feel that way I get that and I think what's happened is maybe they've realized but they now realize they have to double down because everyone else is doubling down"

The Role of Private Credit [0:00]

• Private credit is a significant warning sign, with a substantial portion of insurance annuities and pension funds invested in AI data centers.
• The lack of transparency in private credit makes it difficult to quantify the full extent of the risk.
• Zitron is concerned about the potential for a systemic crisis if these investments sour.

  "You've got multiple problems you have open ai and anthropic you bail these out okay but that doesn't fix the fact that they won't be able to afford to pay 440 billion dollars in three and a half years"

Oracle's Troubled Position [0:00]

• Oracle's stock performance and financial situation are presented as a bellwether for the AI industry's instability.
• Despite its involvement in AI infrastructure, Oracle has not seen the same stock surge as other tech companies.
• Zitron points to Oracle's history of flat revenue and aggressive cost-cutting as indicators of underlying weakness.

  "Look at oracle stock it's down 70 still on the mat even as as uh software stocks are bouncing"

Potential "Pale Horses" of the AI Collapse [0:00]

• Zitron outlines several potential triggers for an AI market collapse: a hyperscaler pulling back investment, a major AI startup going insolvent (like Perplexity), a neocloud failing to raise debt (like CoreWeave), or a poorly received IPO.
• He also points to the increasing cost of debt and GPUs, making it harder for companies to secure funding.
• Nvidia's own financial statements, showing growing accounts receivable and inventory, raise concerns about demand sustainability.

  "The pale horse could either be a major hyperscaler decides it wants out of the race a major hyperscaler has its debt downgraded a la oracle"

The Scale of the Problem: Trillions at Stake [0:00]

• The sheer scale of investment in AI infrastructure—potentially trillions of dollars—is difficult for ordinary people to comprehend.
• Zitron argues that this massive expenditure is not justified by current demand or the value AI provides.
• He warns against relying on historical analogies like the dot-com bubble, suggesting the current situation is far more precarious due to its physical infrastructure and private credit components.

  "The data center's not being used and the data center debt not being paid that is going to be what shocks people"

The "Mania" and Lack of Genuine Demand [0:00]

• Zitron describes the current situation as an "activity bubble" and a "hype cycle," even if some investors aren't treating AI companies as "teflon."
• He questions the sustainability of Nvidia's high margins and forward earnings multiples, given the underlying demand issues.
• The reliance on a few key customers and the increasing cost of production suggest a fragile foundation.

  "The difference and the problem with data sensors these things are not really useful for other things gpu compute is not really that useful outside of this"

The Future of AI and Potential for Change [0:00]

• Zitron is not optimistic about a breakthrough that will magically fix the AI industry's economic issues.
• He dismisses the AGI (Artificial General Intelligence) narrative as a distraction.
• While he doesn't want to be right about a potential collapse, he believes the current trajectory is unsustainable and could lead to significant economic fallout.

  "I don't see how this goes on as long as they they expect to be making hundreds of billions of dollars in compute revenue just from anthropic and open ai"

No Skin in the Game, But Emotional Stakes [0:00]

• Zitron admits he doesn't have direct financial investments in the market but has significant "emotional skin in the game" as a prominent AI skeptic.
• He believes the market is no longer driven by fundamentals but by "smoke signals" and speculative announcements.
• He advocates for a cautious approach, urging people to prepare for a potentially negative outcome.

  "I have emotional skin in the game I don't know what to say oh no I agree you have made you have become the face of ai skepticism"

The Unavoidable Reckoning [0:00]

• Zitron believes the current AI spending spree is unsustainable and will eventually lead to a reckoning.
• He argues that the system will eventually run out of cash, as the demand for AI infrastructure is not as robust as portrayed.
• The lack of transparency and the sheer scale of investment, coupled with the reliance on debt, create a dangerous scenario.

  "And now it's just how big is the system going to let itself grow before the system itself runs out of cash because there's enough cash there's enough but there are limits to how quickly you can mobilize it"

The Broader Economic Implications [0:00]

• Unlike the dot-com crash, the potential AI collapse could have a more profound impact on the broader economy due to the involvement of private credit and the sheer size of the investments.
• Zitron expresses concern about the widespread damage that could result from a downturn in this sector.
• He highlights the difficulty in quantifying the impact due to the opaque nature of private credit markets.

  "The numbers are so much bigger and because the private credit people are all in yeah and they are not nasdaq stocks these are banks and insurance companies and all right"

The Dangers of Data Center Debt [0:00]

• Zitron is particularly concerned about the data center build-out, describing it as one of the most ambitious and unstable infrastructure projects ever undertaken.
• He points to issues like local opposition, tax break changes, and the need for these projects to be completed on time and budget to generate returns for lenders.
• The quality of underwriting in private credit for these projects is questionable, with lenders often "winking at each other."

  "And you've got private credit with their wonderful underwriting quality where most most just people winking at each other and you've got them funding these big unstable difficult complex projects"

Nvidia's Accounting Shenanigans? [0:00]

• Zitron raises questions about Nvidia's growing accounts receivable and inventory, suggesting it might indicate "illusory demand" or artificial demand.
• He notes that Nvidia is engaging in vendor financing and renting back its own GPUs, which he sees as classic signs of desperation.
• He suggests that companies may push the limits of accounting to report revenue, even if no money has actually changed hands.

  "Why is it we like to think of them as backlog but that's the thing though why is this extending right as you're making more money is that because you're reaching the edge of when you can get money quickly or at all"

The Inevitable End of the AI Gold Rush [0:00]

• Zitron believes the current AI gold rush is unsustainable and will end abruptly.
• He argues that the demand for AI infrastructure is not nearly as large as being built out, making it a hype cycle.
• Unlike the dot-com bubble, there's no useful salvageable hardware after the crash, and the cost of running AI infrastructure will remain high.

  "There's no dot com bubble style fixer here ai gpus are not useful for other stuff and also if a data center is left on incomplete unbuilt it's going to cost just as much in three years to finish"

The Unforeseen Consequences for Ordinary People [0:00]

• Zitron expresses worry that the potential collapse of the AI bubble will disproportionately harm ordinary people.
• He notes that retail investors are heavily leveraged and piling into this market, which he finds "scary."
• He believes the system has grown too large and cannot sustain itself without eventually running out of cash.

  "And the thing is is like regular people are going to suffer on this one well you tell them that'll never that'll never change retail investors though are barreling into this"

A Call for Caution and Awareness [0:00]

• Zitron hopes that by sharing his analysis, people can understand the potential risks and consequences of the AI boom in real-time.
• He emphasizes that while he doesn't enjoy seeing potential economic hardship, he feels a responsibility to point out the dangers.
• His goal is to encourage critical thinking and preparedness, rather than blind optimism.

  "So we do put bear we've had sam koppelman on the show I wouldn't call him a bear but he is he did the blue energy piece back and I've had nick on on uh on the youtube live stream"

Where to Find Ed Zitron [0:00]

• Ed Zitron encourages listeners to subscribe to his newsletter, "Where's Your Ed At," at WheresYourEd.at.
• He also promotes his podcast, "Better Offline," available at BetterOffline.com.
• Zitron expresses his enjoyment of podcasting and hopes to return for future discussions.

  "Where's your ed dot at for the newsletter subscribe to the premium and of course betteroffline.com for links to the podcast dude you are a virtuoso at this you crushed it"