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

Ed Zitron Where's Your Ed At Published

AI's Economics Don't Make Sense [Ad Free]

We recorded 28 claims from this post. 20 were checked against evidence. 8 can’t be checked. 17 more are held. Read the post at its source.

Claims from this post (28)

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

    “And that day has finally arrived, because every single AI service you use subsidized compute, and every single service is losing money as a result:”

    What we tested: Every single AI service you use subsidized compute, and every single one is losing money as a result.

  2. Assessment: UnresolvedSaid

    “On the back end, AI startups are annihilating cash, with up until recently Anthropic allowing you to burn upwards of $8 in compute for every dollar of your subscription.”

    What we tested: Until recently Anthropic let subscribers burn more than $8 in compute per dollar of subscription, and AI startups are burning cash.

  3. Assessment: UnresolvedSaid

    “while prices for some models might have come down, newer “reasoning” models burn way more tokens, which means the cost of inference has, somehow, gotten higher over time.”

    What we tested: Because reasoning models consume far more tokens, the cost of inference has risen over time even if some per-model prices fell.

  4. Assessment: UnresolvedSaid

    “Both assumptions were wrong, because the monthly subscription model does not make sense for any service connected to a Large Language Model.”

    What we tested: The monthly subscription model does not make economic sense for any LLM-based service.

  5. Assessment: UnresolvedSaid

    “There has never been — and never will be — an economically-feasible way to offer services powered by LLMs without charging the actual token burn of each user, and in the process of deceiving said users, these companies have created products with illusory benefits and questionable return on investment.”

    What we tested: There has never been, and never will be, an economically feasible way to offer LLM-powered services without charging each user's actual token burn, and in deceiving users these companies have created products with illusory benefits and questionable return on investment.

  6. Assessment: Can’t be checkedSaid

    “Had users been forced to pay their actual rates, I imagine many would’ve bounced off the product immediately, as it’s very, very easy to burn through $5 of tokens if you’re fucking around and exploring what an LLM can do.”

    What we tested: If users had paid actual token costs, many likely would have abandoned AI products immediately.

  7. Assessment: UnresolvedSaid

    “I think it’s inevitable that the majority of AI subscriptions move to token-based billing, especially as both Anthropic and OpenAI have now done so with their enterprise customers.”

    What we tested: The majority of AI subscriptions will inevitably move to token-based billing.

  8. Assessment: Can’t be checkedSaid

    “The real thing to look out for — a true pale horse — will be a major AI lab like Anthropic or OpenAI moving all of its subscribers to token-based billing. Once that happens, you’ll know it’s closing time.”

    What we tested: If a major AI lab like Anthropic or OpenAI moves all of its subscribers to token-based billing, that will be a 'true pale horse' signalling that it is 'closing time'.

  9. Assessment: Can’t be checkedSaid

    “In reality, that 10-person, $1250-a-month Teams subscription likely burns anywhere from $5000 to $10,000 a month in API calls, if not more.”

    What we tested: A 10-person $1,250/month Claude Teams subscription likely burns $5,000–$10,000 or more a month in API-equivalent costs.

  10. Assessment: UnresolvedSaid

    “These are astonishing numbers, made more so by the fact that there is no way you’re only spending $30 a day if you use any of Anthropic’s more-recent models.”

    What we tested: Using Anthropic's recent models with Claude Code costs more than $30 per day.

  11. Assessment: Can’t be checkedSaid

    “While this might be possible within the slush fund mindset of a well-funded startup or a banana republic like Meta, any business that actually cares about its costs will have a great deal of trouble justifying spending five or six figures in extra costs on a service that “increases productivity” in a way that nobody can seem to measure.”

    What we tested: Any business that actually cares about its costs will have a great deal of trouble justifying five or six figures in extra costs on a service whose claimed productivity increase nobody can seem to measure, though this might be possible for well-funded startups or companies like Meta.

  12. Assessment: UnresolvedSaid

    “At some point investors — even the braindead fuckwits who have been inflating the AI bubble — will begin to question mounting R&D costs (which is where AI token burn is usually hidden) when the company’s revenue growth fails to follow.”

    What we tested: Investors will eventually question mounting R&D costs from AI token spending when revenue growth fails to follow.

  13. Assessment: Can’t be checkedSaid

    “And once everybody switches to token-based billing, I’m not sure we’ll see quite as much hype around generative AI.”

    What we tested: Once token-based billing becomes universal, hype around generative AI may diminish.

  14. Assessment: Can’t be checkedSaid

    “As a result, every AI data center starts millions of dollars in the hole, and even with six-year-long depreciation schedules, takes years to pay off… and with NVIDIA’s yearly upgrade cycle, those GPUs are unlikely to make that much money once you’re done with your first customer contract.”

    What we tested: AI data centers take years to pay off, and their GPUs are unlikely to earn much after the first customer contract due to NVIDIA's yearly upgrade cycle.

  15. Assessment: UnresolvedSaid

    “It’s also unclear whether the customer base for AI compute exists outside of OpenAI and Anthropic, whose demand accounts for 50% of AI data centers under construction, creating a massive systemic weakness if either of them lacks the money to pay.”

    What we tested: OpenAI and Anthropic account for 50% of AI data centers under construction, a systemic weakness if either cannot pay.

  16. Assessment: UnresolvedSaid

    “As of April 27 2026, only two buildings are operational and generating revenue, and the third barely has any IT gear in it. I estimate the total cost of Stargate Abilene to be around $52.8 billion.”

    What we tested: As of April 27, 2026, only two Stargate Abilene buildings are operational and generating revenue and the third barely has any IT gear in it; the author estimates the campus's total cost at around $52.8 billion.

  17. Assessment: UnresolvedSaid

    “I should also be clear that Oracle is mortgaging its entire fucking future on projects like Stargate Abilene, incurring billions of dollars of costs up front for a business that will take years to turn a profit even if OpenAI makes every single payment in a timely manner.”

    What we tested: Oracle's Stargate business will take years to become profitable even if OpenAI pays on time.

  18. Assessment: UnresolvedSaid

    “It is an incredibly unprofitable business, and even if it wasn’t, it would have to make so much more money than it currently does to pay Oracle on an ongoing basis.”

    What we tested: OpenAI is incredibly unprofitable and would need far more revenue than it has to pay Oracle on an ongoing basis.

  19. Assessment: UnresolvedSaid

    “And if OpenAI can’t pay for that compute, Oracle dies, because it’s taken on around $115 billion in debt just to build Stargate’s data centers, and needs another $150 billion to finish them:”

    What we tested: If OpenAI cannot pay for its Stargate compute, Oracle will collapse given its ~$115 billion of Stargate debt.

  20. Assessment: UnresolvedSaid

    “I really need to be clear that Oracle has no other path to making this revenue without OpenAI, and these projects are entirely financed and paid for using the projected cashflow of the data centers themselves.”

    What we tested: Oracle has no other path to making this Stargate revenue without OpenAI, and these projects are entirely financed and paid for using the data centers' own projected cash flow.

  21. Assessment: UnresolvedSaid

    “If we assume that the 15.2GW of data center capacity under construction (due by the end of 2028) has a PUE of around 1.35, that leaves us with roughly 11.2GW of critical IT. At $14 million per megawatt, that works out to around $156.8 billion in annual revenue in GPU rental revenue required to actually make these data centers worth it.”

    What we tested: Assuming a PUE of about 1.35 and $14 million per megawatt, the 15.2GW of data center capacity under construction (due by the end of 2028) would require around $156.8 billion a year in GPU rental revenue to be worth it.

  22. Assessment: UnresolvedSaid

    “Anthropic and OpenAI both intend to burn tens of billions of dollars in the next few years, and neither of them have a path to profitability.”

    What we tested: Neither OpenAI nor Anthropic has a path to profitability.

  23. Assessment: UnresolvedSaid

    “This means that a large part — if not the majority of — AI compute revenue is dependent on a continual flow of venture capital and debt, both of which are only made possible by investors that still believe that generative AI will be the biggest, most hugest thing in the world.”

    What we tested: A large part, possibly a majority, of AI compute revenue depends on continued venture capital and debt.

  24. Assessment: UnresolvedSaid

    “Even Cursor, a company that briefly (before its pseudo-acquisition by Musk’s SpaceX) claimed it had positive gross margins, actually had negative 23% gross margins as of January, or negative 31% if you include the cost of non-paying users, which you fucking should if you actually care about your accounting.”

    What we tested: Cursor had -23% gross margins as of January 2026, or -31% including non-paying users.

  25. Assessment: Can’t be checkedSaid

    “I realize it seems a little much, but I genuinely believe that subscription-based AI services were an act of deception tantamount to fraud, as they misrepresented the core unit economics and thus the possibilities of a Large Language Model.”

    What we tested: He genuinely believes subscription-based AI services were an act of deception tantamount to fraud, because they misrepresented the core unit economics and thus the possibilities of LLMs.

  26. Assessment: UnresolvedSaid

    “I truly believe that LLM-based subscriptions are going to go away entirely, at least at the scale of any product that generates code, and in doing so, Amodei and Altman will wrap up their con, or at least believe that they have.”

    What we tested: LLM-based subscriptions will disappear entirely, at least for code-generating products.

  27. Assessment: Can’t be checkedSaid

    “Generative AI is an insult. It is unreliable, its economics don’t make sense, its outcomes don’t justify its existence, and the perpetrators of its con are boring, oafish and avaricious men disconnected from society and anybody who would ever disagree with them.”

    What we tested: Generative AI is an insult: it is unreliable, its economics don't make sense, its outcomes don't justify its existence, and the perpetrators of its con are boring, oafish and avaricious men disconnected from society.

  28. Assessment: UnresolvedSaid

    “This is a deeply sad era. The people that have so aggressively worked together to hold up this industry have only delayed its inevitable fall.”

    What we tested: The AI industry's fall is inevitable; its supporters have only delayed it.

Held from this post (17)

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