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

Ed Zitron Where's Your Ed At Published

NVIDIA Isn't Enron - So What Is It?

We recorded 27 claims from this post. 17 were checked against evidence. 10 can’t be checked. 15 more are held. Read the post at its source.

Claims from this post (27)

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: Can’t be checkedSaid

    “And yes, while I believe that one day this will all be seen as one of the most egregious wastes of capital of all time, for the time being, Jensen Huang may be one of the most successful salespeople in business history.”

    What we tested: The generative AI buildout will one day be seen as one of the most egregious wastes of capital of all time.

  2. Assessment: UnresolvedSaid

    “Yet as dodgy and weird and unsustainable as this is, it isn’t illegal, and it certainly isn’t Enron, because, as NVIDIA has been trying to tell you, it is nothing like Enron!”

    What we tested: NVIDIA's circular deals with neoclouds like CoreWeave, Lambda and Nebius are dodgy and unsustainable but not illegal.

  3. Assessment: Can’t be checkedSaid

    “Nobody seriously thought that NVIDIA was like Enron before (though JustDario, who has been questioning its accounting practices for years, is a little suspicious), because Enron was one of the largest criminal enterprises in history, and NVIDIA is at worst, I believe, a big, dodgy entity that is doing whatever it can to survive.”

    What we tested: NVIDIA is at worst a big, dodgy entity doing whatever it can to survive, not a criminal enterprise like Enron.

  4. Assessment: UnresolvedSaid

    “To be clear, NVIDIA is quite different from Lucent. It has plenty of money, and the circular deals it does with CoreWeave and Lambda don’t involve the same levels of risk.”

    What we tested: NVIDIA is quite different from Lucent; it has plenty of money and its circular deals with CoreWeave and Lambda carry less risk.

  5. Assessment: Can’t be checkedSaid

    “In fact, before NVIDIA put out a letter saying how it was nothing like Enron I would have staunchly defended the company against the Enron allegations, because I truly do not think NVIDIA is committing fraud.”

    What we tested: Zitron truly does not think NVIDIA is committing fraud.

  6. Assessment: UnresolvedSaid

    “You have deals with both Lambda and CoreWeave to guarantee that they will have compute revenue, which they in turn use to raise debt, which is used to buy more of your GPUs.”

    What we tested: NVIDIA guarantees compute revenue for Lambda and CoreWeave, which they use to raise debt to buy more NVIDIA GPUs.

  7. Assessment: Can’t be checkedSaid

    “More than likely NVIDIA is doing relatively boring vendor financing stuff and getting people to pay them on 50-60 day time scales — probably net 60, and, like it said, it gets paid upfront sometimes.”

    What we tested: NVIDIA is more than likely doing relatively boring vendor financing with roughly 50-60 day payment terms.

  8. Assessment: Can’t be checkedSaid

    “It’s entirely possible that everything NVIDIA is doing is above board. Great! But that doesn’t do anything for the deep pit of dread in my stomach.”

    What we tested: It is entirely possible that everything NVIDIA is doing is above board.

  9. Assessment: SupportedSaid

    “NVIDIA is a company that sells all sorts of stuff, but the only reason you're hearing about it as a normal person is that NVIDIA's stock has become a load-bearing entity in the US stock market.”

    What we tested: NVIDIA's stock has become a load-bearing entity in the US stock market.

    What happened: These weights give one company substantial mechanical influence on both indexes. They support the ordinary reading of unusually large market importance. (Reuters, republished by Kitco, )

  10. Assessment: UnresolvedSaid

    “Because nobody else has really caught up with CUDA, NVIDIA has a functional monopoly, and yes, you can have a situation where a market has a monopoly, even if there is, at least in theory, competition.”

    What we tested: Because no competitor has caught up with CUDA, NVIDIA has a functional monopoly on AI GPUs.

  11. Assessment: Can’t be checkedSaid

    “In simple terms, big tech — Amazon, Google, Microsoft and Meta, but also a number of other companies — no longer has the “next big thing,” and jumped on AI out of an abundance of desperation.”

    What we tested: Big tech jumped on AI out of desperation because it had run out of a next big thing.

  12. Assessment: SupportedSaid

    “Did it matter that there was compelling evidence in 2022 (Gary Marcus was right!) that there were limits to scaling laws, and that we would hit the point of diminishing returns?”

    What we tested: There was compelling evidence in 2022 that LLM scaling laws had limits and would hit diminishing returns, as Gary Marcus argued.

    What happened: The study compared 14 models across 11 scientific tasks and reported little performance difference despite major size and data differences. This is direct evidence of limits in a specified language-model setting, not all LLMs. (Zhi Hong and colleagues, arXiv, )

  13. Assessment: MixedSaid

    “Large Language Models are limited, AI agents are a pipedream and simply do not work, AI-powered products are unreliable and coding LLMs make developers slower, and the cost of inference — the way in which a model produces its output — keeps going up.”

    What we tested: LLMs are limited, AI agents do not work, AI products are unreliable, coding LLMs make developers slower, and inference costs keep rising.

    What held: The February–June experiment found slowdown among 16 experienced developers completing 246 tasks in familiar open-source repositories. It supports that setting, and the authors explicitly reject generalizing it to most software development. (METR, )

    What did not: Agents successfully completed a measurable share of multi-step software and reasoning tasks, contradicting a categorical denial of task-level usefulness. A 50% success threshold also leaves substantial reliability limitations and does not establish dependable autonomous employment. (METR / arXiv, )

  14. Assessment: UnresolvedSaid

    “Any kind of depreciation schedule is based on, at best, assumptions, and at worst, hope.”

    What we tested: Data center GPU depreciation schedules are based at best on assumptions and at worst on hope.

  15. Assessment: UnresolvedSaid

    “The same fate awaits H100 GPUs too.”

    What we tested: H100 GPUs will, like A100s, fall to low rental prices as they age.

  16. Assessment: UnresolvedSaid

    “I do not know where these six million Blackwell GPUs have gone, but they certainly haven’t gone into data centers that are powered and turned on.”

    What we tested: The six million Blackwell GPUs NVIDIA says it shipped have not gone into powered, operating data centers.

  17. Assessment: UnresolvedSaid

    “In really simple terms: there isn’t enough power or built data centers for those six million Blackwell GPUs, in part because the data centers aren’t built, and in part because there isn’t enough power for the ones that are.”

    What we tested: There is not enough built data center capacity or power to deploy six million Blackwell GPUs.

  18. Assessment: UnresolvedSaid

    “Nobody has made a profit from AI, and those making revenue aren’t really making much.”

    What we tested: Nobody has made a profit from AI, and AI revenues are small.

  19. Assessment: UnresolvedSaid

    “In the same period, it spent $8.67 billion on inference (the process in which an LLM creates an output).”

    What we tested: In the same period as the OpenAI figures he cites, OpenAI spent $8.67 billion on inference.

  20. Assessment: UnresolvedSaid

    “Both companies burn billions of dollars, and require an infinite amount of venture capital to keep alive at a time when the Saudi Public Investment Fund is struggling and the US venture capital system is set to run out of cash in the next year and a half.”

    What we tested: OpenAI and Anthropic burn billions and depend on venture capital, while the US venture capital system is set to run out of cash within about 18 months.

  21. Assessment: Can’t be checkedSaid

    “CoreWeave made $1.36 billion in revenue (and lost $110 million doing so) in its last quarter — and if that’s indicative of the kind of actual, real demand for AI compute, I think it’s time to start panicking about whether all of this was for nothing.”

    What we tested: If CoreWeave's last quarter ($1.36 billion revenue, $110 million loss) is indicative of real demand for AI compute, Zitron thinks it is time to start panicking about whether all of this was for nothing.

  22. Assessment: Can’t be checkedSaid

    “Blackwell does not appear to be “more efficient” in a way that actually makes anybody a profit, and we’re potentially years from seeing these GPUs in operation in data centers at the scale they’re being shipped — so why would anybody be buying more?”

    What we tested: Blackwell does not appear to be more efficient in a way that makes anybody a profit, and we are potentially years from seeing Blackwell GPUs in operation in data centers at the scale they are being shipped.

  23. Assessment: UnresolvedSaid

    “In simple terms, NVIDIA’s customers rely on debt to buy its GPUs, and NVIDIA’s customers’ customers rely on debt to pay to rent them.”

    What we tested: NVIDIA's customers rely on debt to buy GPUs, and their customers rely on debt to rent them.

  24. Assessment: UnresolvedSaid

    “At some point, a link in this debt-backed chain breaks, because very little cashflow exists to prop it up.”

    What we tested: A link in the debt-backed AI compute chain will break because little cashflow supports it.

  25. Assessment: Can’t be checkedSaid

    “In fact, I’m not even saying it goes tits up. Hell, it might even have another good quarter or two.”

    What we tested: NVIDIA might have another good quarter or two.

  26. Assessment: UnresolvedSaid

    “Once this pops — and it will pop, because there is simply not enough money to do this forever — there must be a referendum on those that chose to ignore the naked instability of this era, and the endless lies that inflated the AI bubble.”

    What we tested: The AI bubble will pop because there is not enough money to sustain it forever.

  27. Assessment: Can’t be checkedSaid

    “Once this pops — and it will pop, because there is simply not enough money to do this forever — there must be a referendum on those that chose to ignore the naked instability of this era, and the endless lies that inflated the AI bubble.”

    What we tested: Once the AI bubble pops, there must be a referendum on those who chose to ignore the naked instability of this era and the endless lies that inflated the AI bubble.

Held from this post (15)

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