Pundit ClaimsAI claims by Gary Marcus and Ed Zitron, with the evidence

Ed Zitron Where's Your Ed At Published

The AI Industry Is Losing

We recorded 31 claims from this post. 17 were checked against evidence. 14 can’t be checked. 19 more are held. Read the post at its source.

Claims from this post (31)

In the order they appear in the post. Each quotation is the speaker’s words in this post. A label such as “Supported” or “Can’t be checked” is the claim’s current assessment; its page has the evidence and the reasons.

  1. Assessment: UnresolvedSaid

    “All of this is to say that Oracle has massively leveraged itself for the benefit of one company, OpenAI, and if that company can’t pay its bills, it’s fucked.”

    What we tested: Oracle has massively leveraged itself for OpenAI and will be in severe trouble if OpenAI cannot pay its bills.

  2. Assessment: UnresolvedSaid

    “If Microsoft, Google, Amazon and Meta decide that it’s time to stop spending $30 billion or more a quarter on GPUs, RAM, storage, and data center construction, that’ll tear a hole in the side of what people assume is a permanent supercycle.”

    What we tested: If Microsoft, Google, Amazon and Meta decide to stop spending $30 billion or more a quarter on GPUs, RAM, storage and data center construction, that will tear a hole in what people assume is a permanent supercycle.

  3. Assessment: UnresolvedSaid

    “Hyperscalers — who will see their capex outpace their cashflows as of Q3 2026 — have had such poor returns on their investment in AI that none of them will actually disclose their revenues outside of vague “run rates,” which means that all of this investment is effectively based on the idea that something completely different will happen in the future.”

    What we tested: Hyperscalers' capex will exceed their cash flows as of Q3 2026, and their AI returns are poor enough that they don't disclose AI revenues beyond run rates.

  4. Assessment: UnresolvedSaid

    “Said future will have to make them at least $2 trillion in brand new revenue by 2030, because if it doesn’t, effectively all of that capex will have been spent to prop up Anthropic, OpenAI, and whatever it is that Meta is doing with its chatbots.”

    What we tested: Hyperscaler AI capex needs to produce at least $2 trillion in new revenue by 2030, or it will have effectively been spent propping up Anthropic, OpenAI and Meta's chatbots.

  5. Assessment: UnresolvedSaid

    “Those millions of H100 and B200 and B300 GPUs are not going to usher in a digital God, they are not going to create recursive self-improvement, they are not going to be the fulcrum to adding $600 billion or more in brand new revenue to current services”

    What we tested: The current GPU buildout will not produce a 'digital God', recursive self-improvement, or $600 billion+ in new revenue for current services.

  6. Assessment: UnresolvedSaid

    “Anthropic and OpenAI have raised a little under $300 billion combined since 2019, but I estimate their true cost is at least $500 billion given hyperscaler capex investments that were necessary for them to exist, and that’s before you consider the $340 billion or more that Oracle is spending to build out the 7.1GW of “Stargate” data centers for OpenAI.”

    What we tested: Anthropic and OpenAI have raised a little under $300 billion combined since 2019, but he estimates their true cost is at least $500 billion once the hyperscaler capex needed for them to exist is counted, before Oracle's $340 billion-plus Stargate buildout for OpenAI.

  7. Assessment: ContradictedSaid

    “Outside of OpenAI, Microsoft may as well not have an AI business.”

    What we tested: Excluding OpenAI, Microsoft has essentially no meaningful AI business.

    What happened: More than 20 million paid seats establishes substantial commercial adoption of a Microsoft AI product beyond OpenAI's purchases of Azure compute. It does not establish standalone Copilot revenue or profit. (Microsoft, )

  8. Assessment: Can’t be checkedSaid

    “meaning that, at least in that quarter, OpenAI likely represented around 70% of Microsoft’s AI revenue, and I’d be surprised if that dramatically changed in the year that followed, given that OpenAI’s inference spend was $3.648 billion in Q1FY26.”

    What we tested: At least in Q3FY25, OpenAI likely represented around 70% of Microsoft's AI revenue, and he would be surprised if that changed dramatically in the following year.

  9. Assessment: Can’t be checkedSaid

    “There is simply no logical or rational reason to invest any further capex in AI, outside of the mistaken belief that OpenAI or Anthropic could actually afford to pay without Google, Amazon, or Microsoft handing it to them.”

    What we tested: There is no rational reason for further AI capex, since OpenAI and Anthropic can't pay without hyperscaler money.

  10. Assessment: UnresolvedSaid

    “Record sales across NVIDIA, Micron, Sandisk, SK Hynix, and Samsung are a direct result of an entirely speculative asset bubble, driven by the reckless and directionless capital expenditures of some of the largest and richest companies in the world.”

    What we tested: Record sales at NVIDIA, Micron, Sandisk, SK Hynix and Samsung are a direct result of an entirely speculative asset bubble driven by the reckless, directionless capex of some of the world's largest and richest companies.

  11. Assessment: ContradictedSaid

    “Anyone investing in data centers is building speculative capacity for demand that does not exist outside of Anthropic and OpenAI.”

    What we tested: AI data center demand does not exist outside Anthropic and OpenAI, so data center investment is speculative.

    What happened: The announcement identifies Meta as the customer and distinguishes $12 billion of dedicated capacity from up to $15 billion of additional capacity purchases. The dedicated component establishes substantial demand outside the two labs without treating the maximum value as realized spending. (Nebius, )

  12. Assessment: Can’t be checkedSaid

    “The current capital expenditures have, outside of the creation of OpenAI and Anthropic, been a near-complete waste. Microsoft 365 Copilot sucks. GitHub Copilot sucks. Google AI Overviews suck.”

    What we tested: Apart from creating OpenAI and Anthropic, current AI capex has been a near-complete waste, and Microsoft 365 Copilot, GitHub Copilot and Google AI Overviews suck.

  13. Assessment: Can’t be checkedSaid

    “Amazon Rufus sucks, and Amazon should be investigated by the SEC for suggesting it drove $10 billion in “annualized revenue” in Q3 2025, because it most assuredly did not.”

    What we tested: The SEC should investigate Amazon for claiming Rufus drove $10B annualized revenue in Q3 2025, which it did not.

  14. Assessment: Can’t be checkedSaid

    “I refuse to buy into the premise that LLMs’ ability to generate code or replicate open source software is proof that these things will become a powerful, autonomous tool in the future”

    What we tested: He refuses to accept that LLMs' ability to generate code or replicate open source software is proof that they will become powerful, autonomous tools in the future.

  15. Assessment: Can’t be checkedSaid

    “This report is industry marketing framed as research, but uses deliberately positive framing and questionable data sources.”

    What we tested: Exponential View's AI revenue report is industry marketing framed as research that uses deliberately positive framing and questionable data sources.

  16. Assessment: UnresolvedSaid

    “So, yeah, not including anything from Q2 2026, Anthropic and OpenAI represent 68% of the $110 billion of AI revenue that Exponential View is trying to get people excited about.”

    What we tested: Anthropic and OpenAI revenue and compute spend account for about 68% of Exponential View's $110B trailing AI revenue figure, even excluding Q2 2026.

  17. Assessment: Can’t be checkedSaid

    “Instead, Exponential View says that AI is “real, big & fast” through a Pee Wee’s Playhouse of undefined models, datasets and alleged “quality grades” that helps feed a dangerous bubble further, and likely cons retail investors into further terrible decisions.”

    What we tested: Exponential View's report feeds a dangerous AI bubble and likely misleads retail investors.

  18. Assessment: Can’t be checkedSaid

    “They’re also tools intentionally built to produce engagement, and to make you feel you’re productive, even if you’re not.”

    What we tested: LLM tools are intentionally built to produce engagement and to make users feel productive, even if they are not.

  19. Assessment: UnresolvedSaid

    “what we’re witnessing is a widespread mental health epidemic caused by liars like Sam Altman, Dario Amodei, and their wealthy backers lying about the capabilities of AI”

    What we tested: Altman, Amodei and backers lying about AI capabilities have caused a widespread mental health epidemic among tech workers.

  20. Assessment: UnresolvedSaid

    “AI does not do what you have been promised it does, and those who are excited about it are excited because they believe it will replace you.”

    What we tested: AI does not deliver what it has been promised to tech workers.

  21. Assessment: Can’t be checkedSaid

    “The people who make these agents fantasize about replacing you with them, and want to use your data to do so. They are lying that it is possible, but they want you to be scared so you will use their products more.”

    What we tested: The makers of AI agents fantasize about replacing workers with them and want to use workers' data to do so; they are lying that this is possible, and they want workers scared so they will use their products more.

  22. Assessment: Can’t be checkedSaid

    “Any “autonomous” tool that requires specific prompting, harnesses, scripts and tooling to make it sometimes work autonomously is conning you.”

    What we tested: Any 'autonomous' tool that needs specific prompting, harnesses, scripts and tooling to sometimes work autonomously is conning its users.

  23. Assessment: UnresolvedSaid

    “As if I haven’t made it clear enough, the AI industry is losing. Their plans are not working, their products are not doing the things that they’ve promised, and though they intend to exhaust every available source of capital, they aren’t going to have enough money to do this forever.”

    What we tested: The AI industry's plans and products are failing and it will run out of capital to keep going indefinitely.

  24. Assessment: Can’t be checkedSaid

    “And no, AI is not “too big to fail.””

    What we tested: The AI industry is not too big to fail.

  25. Assessment: UnresolvedSaid

    “At some point Anthropic or OpenAI is going to slow down and then stop making models entirely because it costs too much money to train models, and said costs are only increasing.”

    What we tested: Anthropic or OpenAI will eventually slow and then stop training new models because training costs too much and keeps rising.

  26. Assessment: Can’t be checkedSaid

    “Even if GLM 5.2 is truly nearly as good as Opus 4.8, it did so by copying its outputs, which means that these models will likely only get as good as long as the foundation model companies keep training, which will only be possible if they can keep raising funding, which will become difficult if open source models eat their lunch in any meaningful way.”

    What we tested: Distilled open models like GLM 5.2 will likely improve only while frontier labs keep training, which depends on fundraising that open-source competition would undermine.

  27. Assessment: UnresolvedSaid

    “But they’re now going to have to slow-roll them, because Sam and Dario’s four or five-year-long scaremongering campaign has forced them into a situation where the US government demands oversight into their model releases at a time where the AI industry cannot afford to slow down.”

    What we tested: OpenAI and Anthropic will have to slow their model releases because of US government oversight prompted by their own scaremongering.

  28. Assessment: Can’t be checkedSaid

    “Considering that Mr. Altman can’t even con a bunch of bankers who were dumb enough to believe that SpaceX could 300x its AI revenue by 2030, it’s clear that the jig is up.”

    What we tested: Since Altman cannot even win over bankers who believed SpaceX could grow its AI revenue 300-fold by 2030, it is clear the jig is up for OpenAI.

  29. Assessment: UnresolvedSaid

    “Anthropic is a large language model company that loses billions of dollars that has subsidized accounts that allow users to burn $8,000 a month in tokens for $200.”

    What we tested: Anthropic loses billions and subsidizes accounts letting users consume $8,000/month of tokens for $200.

  30. Assessment: Can’t be checkedSaid

    “about the “cost of intelligence going down” (when it’s actually going up)”

    What we tested: The cost of AI 'intelligence' is going up, not down.

  31. Assessment: UnresolvedSaid

    “Anthropic and OpenAI have $1.1 trillion in compute commitments that are contingent on their continued growth, at a time when their customers are protesting their costs, at a time when the market is clearly saying “you are not worth a trillion dollars.””

    What we tested: Anthropic and OpenAI have $1.1 trillion in compute commitments that depend on their continued growth, while their customers protest costs and the market is clearly saying they are not worth a trillion dollars.

Held from this post (19)

These records did not pass every check, so nothing about them is published. Held is not a finding.