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

Changes, page 15 of 55

Every substantive change to the register, newest first. Speaker events record what the people covered said or changed. Site events record what this site published, corrected, flagged, or withdrew. The two never share a label.

Substantive changes (10984)

Page 15 of 55: changes 2801 to 3000 of 10984, newest first. Dates are when the change was recorded.

  1. HeldBy this siteEd Zitron

    Held record wyea-exit-liquidity-3c87b5

    The openai review found the criteria unfair to the speaker's words.

  3. HeldBy this siteEd Zitron

    Held record wyea-exit-liquidity-1b19e2

    e2 is dated 2026-10, which is not inside the period described (2026-01-01 to 2026-10-28); it can only be context

  4. Assessment publishedBy this siteEd Zitron

    Anthropic's financials will never look better than in its S-1, and its IPO timing is deliberate.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-exit-liquidity-135dfd-a1.

  5. HeldBy this siteEd Zitron

    Held record wyea-exclusive-openai-financials-cc8ff1

    The openai review found the criteria unfair to the speaker's words.

  6. Assessment publishedBy this siteEd Zitron

    According to audited financial documents, OpenAI lost around $38.5 billion in 2025.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-exclusive-openai-financials-cae693-a1.

  7. HeldBy this siteEd Zitron

    Held record wyea-exclusive-openai-financials-bcdfd8

    The two reviews differ: anthropic historical_observation/not_applicable/unresolved, openai historical_observation/not_applicable/supported.

  8. Assessment publishedBy this siteEd Zitron

    In 2025 OpenAI paid Microsoft $10.59 billion for research and development expenses.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-exclusive-openai-financials-bc5580-a1.

  10. HeldBy this siteEd Zitron

    Held record wyea-exclusive-openai-financials-b1159f

    The final-text check failed: proposition_faithful.

  11. Assessment publishedBy this siteEd Zitron

    In 2025 SoftBank paid OpenAI $867 million and Microsoft paid OpenAI $303 million.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-exclusive-openai-financials-b0474e-a2.

  12. HeldBy this siteEd Zitron

    Held record wyea-exclusive-openai-financials-a0c028

    The final-text check failed: proposition_faithful.

  14. Assessment publishedBy this siteEd Zitron

    OpenAI's 2025 conversion to a for-profit entity led to a $41.55 billion loss from changes in fair value of convertible interests and warrant liability.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-exclusive-openai-financials-8c7f4b-a2.

  15. Assessment publishedBy this siteEd Zitron

    At the end of 2025 OpenAI had $3.64 billion in liabilities to Microsoft plus $21 million in accrued expenses and other current liabilities.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-exclusive-openai-financials-54e972-a1.

  16. Assessment publishedBy this siteEd Zitron

    In 2024 OpenAI had $3.7 billion in revenue, about $12.4 billion in costs and expenses, and a net loss attributable to the company of $5.09 billion.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-exclusive-openai-financials-46f2f0-a1.

  17. HeldBy this siteEd Zitron

    Held record wyea-exclusive-microsoft-moving-all-github-copilot-subscribers-to-t-164552e31c81-93a4bf

    The final-text check failed: citations_correct.

  18. HeldBy this siteEd Zitron

    Held record wyea-exclusive-microsoft-moving-all-github-copilot-subscribers-to-t-164552e31c81-7d162c

    The openai review found the criteria unfair to the speaker's words.

  20. HeldBy this siteEd Zitron

    Held record wyea-exclusive-microsoft-moving-all-github-copilot-subscribers-to-t-164552e31c81-570572

    The openai review found the criteria unfair to the speaker's words.

  21. HeldBy this siteEd Zitron

    Held record wyea-exclusive-microsoft-moving-all-github-copilot-subscribers-to-t-164552e31c81-43c3d7

    "likely" is hedged or probabilistic; one outcome does not score it; hedged language is not scored, even on a plain reading

  23. HeldBy this siteEd Zitron

    Held record wyea-exclusive-microsoft-moving-all-github-copilot-subscribers-to-t-164552e31c81-1758d1

    The final-text check failed: citations_correct.

  24. HeldBy this siteEd Zitron

    Held record wyea-exclusive-microsoft-moving-all-github-copilot-subscribers-to-t-164552e31c81-06953f

    The final-text check failed: citations_correct.

  25. HeldBy this siteEd Zitron

    Held record wyea-empty-laughter-8c61be

    the conclusive event e1 is not outcome evidence in the packet

  27. HeldBy this siteEd Zitron

    Held record wyea-empty-laughter-2a6d63

    e1 is dated 2024-01, which is not inside the period described (… to 2024-01-16); it can only be context

  28. Assessment publishedBy this siteEd Zitron

    Launching Grok to all Premium+ subscribers would make X less profitable as usage grows.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-elons-gordian-knot-e8d31d-a1.

  29. HeldBy this siteEd Zitron

    Held record wyea-elons-gordian-knot-b69574

    The final-text check failed: citations_correct.

  30. HeldBy this siteEd Zitron

    Held record wyea-elons-gordian-knot-6efd71

    The two reviews differ: anthropic historical_observation/not_applicable/supported, openai historical_observation/not_applicable/unresolved.

  31. Assessment publishedBy this siteEd Zitron

    If Grok is offered to every X Premium+ subscriber at $16/month, it will not be profitable.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-elons-gordian-knot-58c002-a1.

  32. Assessment publishedBy this siteEd Zitron

    LLMs, even well-resourced ones like GPT-4 trained on curated data, habitually hallucinate facts, figures, and events.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-elons-gordian-knot-31dd8b-a1.

  33. HeldBy this siteEd Zitron

    Held record wyea-elons-gordian-knot-10d5e2

    The openai review reads the claim differently.

  35. Assessment publishedBy this siteEd Zitron

    Grok has already fabricated a story about a British MP being expelled from Parliament.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-elons-gordian-knot-00bd77-a1.

  36. Assessment publishedBy this siteEd Zitron

    Musk's open-air beta test of self-driving vehicles has harmed and killed people.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-elon-musk-is-dangerous-to-society-0c165b-a1.

  38. Assessment publishedBy this siteEd Zitron

    Oracle is finished unless it sharply cuts back or cancels its capex.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-e8ad19-a1.

  39. Assessment publishedBy this siteEd Zitron

    Large-scale AI compute demand does not exist outside Anthropic and OpenAI, and he sees none emerging.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-e38d91-a1.

  40. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-dc73f4

    e1 is decisive for a financial claim but states no metric scope; e2 is decisive for a financial claim but states no metric scope

  41. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-d95f93

    The openai review found the criteria unfair to the speaker's words.

  42. Assessment publishedBy this siteEd Zitron

    Many commentators will keep hyping AI until NVIDIA or a hyperscaler misses analyst expectations or the AI labs begin to fail.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-d1e33d-a1.

  43. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-d16876

    The anthropic review found the criteria unfair to the speaker's words.

  45. Assessment publishedBy this siteEd Zitron

    NVIDIA's $500B financing arrangement with asset managers is only a theoretical memorandum of understanding, not a deal.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-b69b59-a1.

  48. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-967024

    The final-text check failed: proposition_faithful.

  49. Assessment publishedBy this siteEd Zitron

    OpenAI and Anthropic will each lose tens of billions of dollars in 2026 and more in 2027.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-94f999-a1.

  50. Assessment publishedBy this siteEd Zitron

    Hyperscalers spent huge sums on AI without a massive return; the strategy didn't work.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-92e901-a2.

  51. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-8f5039

    the horizon on 2026-10-28 is open_horizon, not not_due

  52. Assessment publishedBy this siteEd Zitron

    There are few signs that VC-backed AI companies can be sold to anyone or taken public, so every VC invested in AI needs a miracle.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-8b3466-a1.

  53. Assessment publishedBy this siteEd Zitron

    Apart from compute spend by Anthropic and OpenAI, AI is not why Google, Microsoft and Amazon are growing.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-8a5439-a1.

  54. Assessment publishedBy this siteEd Zitron

    Hyperscalers spent hundreds of billions on GPUs to earn only tens of billions in AI revenue.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-843cef-a1.

  55. Assessment publishedBy this siteEd Zitron

    NVIDIA's future depends wholly on its customers persuading credit markets and on its circular financing continuing and expanding.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-81d999-a1.

  56. Assessment publishedBy this siteEd Zitron

    Monetizing the data center capacity under construction requires hundreds of billions of dollars of demand in the next few years.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-6dd293-a1.

  57. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-66f94b

    The final-text check failed: proposition_faithful.

  59. Assessment publishedBy this siteEd Zitron

    Amazon and Google are doing as poorly on AI revenue as Microsoft, which is why nobody wants to talk about their actual AI revenues.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-56e7ca-a1.

  60. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-4ec1fb

    hedged language is not scored, even on a plain reading

  61. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-4bf44f

    the contract's accounting basis is unknown; a financial claim is scored only on a stated basis; supported needs at least one supporting item of outcome evidence

  64. Assessment publishedBy this siteEd Zitron

    AI has not paid off for Microsoft, Google, Amazon or Meta, and their AI capex was not well spent.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-32eb9a-a1.

  65. Assessment publishedBy this siteEd Zitron

    Hyperscaler GPU purchases did little for revenue until OpenAI and Anthropic grew large, but they inflated stock values.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dont-look-up-207d84-a2.

  66. HeldBy this siteEd Zitron

    Held record wyea-degenerative-ai-f9efee

    The anthropic review does not accept e1 as supports.

  67. HeldBy this siteEd Zitron

    Held record wyea-degenerative-ai-a4c530

    The anthropic review found the criteria unfair to the speaker's words.

  68. Assessment publishedBy this siteEd Zitron

    Presenting fallible AI as an authoritative information source will cause genuine human suffering, including people being inspired to harm themselves or others.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-degenerative-ai-81cd4e-a1.

  69. Assessment publishedBy this siteEd Zitron

    Generative AI products are extremely expensive to run; ChatGPT burns millions of dollars a day in compute.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-degenerative-ai-72e357-a1.

  71. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-ef5f60

    The openai review reads the claim differently.

  72. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-e7dfba

    The openai review found the criteria unfair to the speaker's words.

  73. Assessment publishedBy this siteEd Zitron

    LLM developers are running out of training data, and model capabilities had been peaking since March 2024.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-dfa5aa-a1.

  74. Assessment publishedBy this siteEd Zitron

    Others will build their own models from DeepSeek's work or run their own DeepSeek instances on rented cloud GPUs.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-d930ab-a1.

  75. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-d58a7f

    The two reviews differ: anthropic historical_observation/not_applicable/supported, openai historical_observation/not_applicable/unresolved.

  77. Assessment publishedBy this siteEd Zitron

    In 2024 OpenAI lost $5 billion on $3.7 billion revenue and Anthropic lost just under $3 billion.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-cbda85-a1.

  78. Assessment publishedBy this siteEd Zitron

    If OpenAI cuts prices to compete with DeepSeek, its losses will deepen.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-c8938a-a2.

  79. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-c3d8de

    The two reviews differ: anthropic historical_observation/not_applicable/mixed, openai historical_observation/not_applicable/unresolved.

  80. Assessment publishedBy this siteEd Zitron

    LLMs and reasoning models are niche products.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-c2afdb-a1.

  82. Assessment publishedBy this siteEd Zitron

    OpenAI's Operator agent is barely functional and does not seem to work.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-bcdae1-a1.

  83. Assessment publishedBy this siteEd Zitron

    AGI is not coming.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-bb3868-a1.

  84. Assessment publishedBy this siteEd Zitron

    OpenAI's o1 was not particularly reliable at launch, failing to count the r's in 'strawberry'.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-a1a8da-a1.

  86. Assessment publishedBy this siteEd Zitron

    LLMs run by OpenAI, Anthropic, Google and Meta are unprofitable and unsustainable.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-9b9bed-a2.

  87. Assessment publishedBy this siteEd Zitron

    The transformer-based architecture underlying LLMs has peaked.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-9b9bed-2-a1.

  88. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-9a8961

    a observation claim cannot be assessed on the causal_assessment axis

  89. Assessment publishedBy this siteEd Zitron

    Hyperscalers will struggle to justify spending billions more on GPUs to markets now aware near-identical models can be built cheaply on older hardware.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-869848-a1.

  90. Assessment publishedBy this siteEd Zitron

    OpenAI loses money on every transaction, including the $200/month ChatGPT Pro subscription.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-83d8a8-a1.

  91. Assessment publishedBy this siteEd Zitron

    The API market for OpenAI and Anthropic will eventually collapse, with little else changing.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-80baf9-a1.

  92. Assessment publishedBy this siteEd Zitron

    OpenAI and Anthropic will have to cut API prices, and possibly subscription prices.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-79d680-a1.

  93. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-6f4bcd

    The anthropic review does not accept e1 as supports.

  94. Assessment publishedBy this siteEd Zitron

    OpenAI and Anthropic no longer have moats with which to raise capital.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-670506-a1.

  95. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-49a598

    The openai review reads the claim differently.

  96. Assessment publishedBy this siteEd Zitron

    Once DeepSeek's models are hosted elsewhere and its methods recreated (soon), its low prices will prove to reflect genuinely low running costs.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-44a68b-a1.

  97. Assessment publishedBy this siteEd Zitron

    OpenAI's training and inference costs in 2024 were around $7 billion.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-3d6502-a1.

  98. Assessment publishedBy this siteEd Zitron

    LLM companies like OpenAI have no moat; their infrastructure advantage and talent no longer matter.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-3b26d1-a1.

  100. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-2a094a

    hedged language is not scored, even on a plain reading

  101. Assessment publishedBy this siteEd Zitron

    AGI, which Altman has always portrayed as close on the horizon, has yet to materialize or even come close to materializing.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-2770c3-a1.

  102. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-2744dc

    The final-text check failed: citations_correct.

  103. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-25067b

    The openai review reads the claim differently.

  105. Assessment publishedBy this siteEd Zitron

    DeepSeek's models do not require massive new data centers or ever-bigger NVIDIA GPUs to progress.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-deep-impact-117a45-a2.

  106. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-110800

    e1 is decisive for a financial claim but states no metric scope; e2 is decisive for a financial claim but states no metric scope

  107. Assessment publishedBy this siteEd Zitron

    Repricing CoreWeave's bonds at today's rates would raise their cost 35.8%, adding $1.2 billion.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-fd5ee0-a1.

  108. HeldBy this siteEd Zitron

    Held record wyea-dead-money-f8f550

    hedged language is not scored, even on a plain reading

  109. Assessment publishedBy this siteEd Zitron

    AI compute demand outside Anthropic and OpenAI is roughly $22 billion.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-f32ea5-a1.

  110. HeldBy this siteEd Zitron

    Held record wyea-dead-money-ec2ec3

    The openai review found the criteria unfair to the speaker's words.

  111. HeldBy this siteEd Zitron

    Held record wyea-dead-money-ea8d8d

    The anthropic review reads the claim differently.

  112. HeldBy this siteEd Zitron

    Held record wyea-dead-money-ea79f0

    e1 measures something else (company Anthropic is not Anthropic and OpenAI); it can only be context unless it is derived with stated inputs; e2 measures something else (metric Comparative annual operating expenses per revenue dollar is not Anthropic operating-expense-to-revenue trend and 2025 comparison with OpenAI); it can only be context unless it is derived with stated inputs

  113. HeldBy this siteEd Zitron

    Held record wyea-dead-money-ce018b

    The final-text check failed: rationale_supported_by_evidence; overclaims: Bloomberg reported on October 28, 2026 that 18 data-center financings carried double-digit yields and that Volta’s $5 billion loan required higher interest payments [e1]..

  114. Assessment publishedBy this siteEd Zitron

    AI data center debt is now extremely expensive.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-cd1e3f-a1.

  115. HeldBy this siteEd Zitron

    Held record wyea-dead-money-c74a0e

    The openai review found the criteria unfair to the speaker's words.

  116. HeldBy this siteEd Zitron

    Held record wyea-dead-money-bfb79d

    The openai review reads the claim differently.

  117. HeldBy this siteEd Zitron

    Held record wyea-dead-money-baa6f8

    The openai review found the criteria unfair to the speaker's words.

  118. HeldBy this siteEd Zitron

    Held record wyea-dead-money-ba6b18

    hedged language is not scored, even on a plain reading; e1 measures something else (metric Reported Q3 adjusted operating-profit expectation is not Q3 adjusted profitability presentation and its reconciliation to unadjusted profitability); it can only be context unless it is derived with stated inputs

  119. Assessment publishedBy this siteEd Zitron

    About $800 billion in VC has gone into AI companies since 2023, at least $266 billion to Anthropic and OpenAI.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-b7b1eb-a1.

  120. Assessment publishedBy this siteEd Zitron

    AI data center customers are brittle, unprofitable and dependent on near-perpetual funding.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-b65924-a2.

  121. HeldBy this siteEd Zitron

    Held record wyea-dead-money-b10e40

    The openai review found the criteria unfair to the speaker's words.

  122. HeldBy this siteEd Zitron

    Held record wyea-dead-money-ab5d19

    The openai review found the criteria unfair to the speaker's words.

  123. Assessment publishedBy this siteEd Zitron

    The AI data center situation will collapse, made worse by widespread ignorance or optimism.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-a7fbfe-a1.

  124. HeldBy this siteEd Zitron

    Held record wyea-dead-money-a60375

    The openai review found the criteria unfair to the speaker's words.

  125. HeldBy this siteEd Zitron

    Held record wyea-dead-money-a54502

    The anthropic review reads the claim differently.

  126. Assessment publishedBy this siteEd Zitron

    In 2025 Anthropic spent $12.65 billion in operating expenses to make $4.6 billion in revenue ($2.75 per $1).

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-9eca1d-a1.

  127. Assessment publishedBy this siteEd Zitron

    The talking point that NVIDIA's GPU sales prove actual demand for AI services, or that hyperscaler growth is a result of their capital expenditures, is a complete lie.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-9dadb2-a1.

  128. HeldBy this siteEd Zitron

    Held record wyea-dead-money-9c116d

    The openai review reads the claim differently.

  129. Assessment publishedBy this siteEd Zitron

    Long-term B300 rental prices are less than half of SiliconData's hyperscale pricing for the older H100.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-9182e2-a1.

  130. Assessment publishedBy this siteEd Zitron

    AI data center projects will at some point become too risky for lenders to fund, sooner if Treasury rates rise or projects hit cashflow or construction problems.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-8db40f-a1.

  132. Assessment publishedBy this siteEd Zitron

    The author thinks it is fair to say that at least 50% of Oracle's AI revenues come from OpenAI.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-88966b-a1.

  133. HeldBy this siteEd Zitron

    Held record wyea-dead-money-87afc7

    The openai review reads the claim differently.

  134. HeldBy this siteEd Zitron

    Held record wyea-dead-money-7876dc

    The openai review reads the claim differently.

  135. HeldBy this siteEd Zitron

    Held record wyea-dead-money-7462a4

    The anthropic review found the criteria unfair to the speaker's words.

  136. Assessment publishedBy this siteEd Zitron

    On a strictly cash basis, Oracle's 2025 bonds discussed in the post would, if issued today, cost it another $6.89 billion in interest.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-72967a-a1.

  138. Assessment publishedBy this siteEd Zitron

    Issuing more AI data center debt makes that debt more expensive.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-6a879b-a1.

  139. Assessment publishedBy this siteEd Zitron

    Every AI data center SPV is funded by customer payments that only arrive if the data center is completed.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-6a15de-a2.

  140. Assessment publishedBy this siteEd Zitron

    Cognition, like many AI startups, will eventually run out of money and die or become much smaller.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-691dfe-a1.

  141. HeldBy this siteEd Zitron

    Held record wyea-dead-money-65e8ea

    The final-text check failed: rationale_supported_by_evidence, citations_correct; overclaims: The September 4, 2026 report's 80% concentration figure concerns enterprise revenue, not startup shares of each lab's total revenue [e1]..

  142. HeldBy this siteEd Zitron

    Held record wyea-dead-money-61d327

    The openai review reads the claim differently.

  143. Assessment publishedBy this siteEd Zitron

    AI data centers are extremely expensive.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-6054f8-a1.

  144. HeldBy this siteEd Zitron

    Held record wyea-dead-money-57bc73

    The anthropic review found the criteria unfair to the speaker's words.

  145. HeldBy this siteEd Zitron

    Held record wyea-dead-money-55a225

    The final-text check failed: citations_correct.

  146. HeldBy this siteEd Zitron

    Held record wyea-dead-money-514ad4

    The openai review reads the claim differently.

  147. Assessment publishedBy this siteEd Zitron

    AI data centers take years to build.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-4e807b-a2.

  148. HeldBy this siteEd Zitron

    Held record wyea-dead-money-4c136c

    hedged language is not scored, even on a plain reading

  149. HeldBy this siteEd Zitron

    Held record wyea-dead-money-4b64da

    The final-text check failed: citations_correct.

  150. HeldBy this siteEd Zitron

    Held record wyea-dead-money-4966f2

    The final-text check failed: citations_correct.

  151. HeldBy this siteEd Zitron

    Held record wyea-dead-money-46c25a

    hedged language is not scored, even on a plain reading

  153. HeldBy this siteEd Zitron

    Held record wyea-dead-money-419e3a

    The openai review found the criteria unfair to the speaker's words.

  154. Assessment publishedBy this siteEd Zitron

    AI data center debt will eventually become untenable because 11%+ rates on thin margins don't work.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-376eff-a1.

  155. HeldBy this siteEd Zitron

    Held record wyea-dead-money-36854f

    The openai review found the criteria unfair to the speaker's words.

  156. HeldBy this siteEd Zitron

    Held record wyea-dead-money-35e74d

    hedged language is not scored, even on a plain reading

  157. HeldBy this siteEd Zitron

    Held record wyea-dead-money-2d673f

    The anthropic review reads the claim differently.

  158. Assessment publishedBy this siteEd Zitron

    AI industry revenue is nowhere near the scale needed to justify hyperscaler capex.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-24ca09-a1.

  159. Assessment publishedBy this siteEd Zitron

    Nearly every AI data center's customers are hyperscalers or companies that cannot pay for compute without outside money.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-229b6a-a2.

  161. Assessment publishedBy this siteEd Zitron

    If Anthropic and OpenAI don't pay their $1.3 trillion in compute bills, AI data center demand will collapse.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-1828d5-a2.

  162. Assessment publishedBy this siteEd Zitron

    The planned AI data center debt will run into rising US Treasury yields, raising borrowing costs.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-122d48-a1.

  163. Assessment publishedBy this siteEd Zitron

    Google, Microsoft, Amazon, Oracle and SpaceX have about $183 billion in annual AI revenue, at least 64.3% ($118.2 billion) from Anthropic and OpenAI.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-09c336-a1.

  164. Assessment publishedBy this siteEd Zitron

    Unless hyperscalers reach at least $2 trillion in annual AI revenue by 2030, their AI capex will have been wasted.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-047af1-a2.

  165. HeldBy this siteEd Zitron

    Held record wyea-dead-money-01827e

    hedged language is not scored, even on a plain reading

  166. Assessment publishedBy this siteEd Zitron

    CoreWeave must raise more debt to build its data centers.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-dead-money-014415-a1.

  167. Assessment publishedBy this siteEd Zitron

    As tech companies push engineers harder and shed staff, engineers will turn to AI-generated code that fills codebases with insecure and buggy code.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-crowdstruck-2-0b5d19-a2.

  168. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-feb35d

    The anthropic review reads the claim differently.

  169. Assessment publishedBy this siteEd Zitron

    Anthropic's 'Earnings Before Training, Interest and Taxes' metric does not matter to debt investors, who judge on cashflows.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-f7ef89-a1.

  170. Assessment publishedBy this siteEd Zitron

    Anyone raising AI-related debt will face punitive costs from Treasury prices, AI sentiment and their own financial situation.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-f4d725-a1.

  173. Assessment publishedBy this siteEd Zitron

    Every Stargate data center is heavily behind schedule, and Stargate Abilene has no more than half its capacity installed despite Oracle's '75% delivered' claim.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-e60eb6-a1.

  174. Assessment publishedBy this siteEd Zitron

    Any debt OpenAI and Anthropic issue will be priced like junk with yields of at least 8%, even if they get investment-grade ratings.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-d64d76-a1.

  175. Assessment publishedBy this siteEd Zitron

    OpenAI and Anthropic are unprofitable and unsustainable, and could raise significant debt only through catastrophic failures of regulators and ratings agencies.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-ba180d-a1.

  177. Assessment publishedBy this siteEd Zitron

    Because CoreWeave needs to raise $102 billion in debt through 2030 (per UBS analysts), it will have to keep raising debt on its own two feet.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-95d3bc-a2.

  178. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-8dc480

    The final-text check failed: proposition_faithful.

  179. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-8bd918

    The two reviews differ: anthropic historical_observation/not_applicable/contradicted, openai historical_observation/not_applicable/mixed.

  180. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-8a8075

    The final-text check failed: proposition_faithful.

  181. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-84ea13

    The anthropic review reads the claim differently.

  182. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-81e81e

    The final-text check failed: rationale_supported_by_evidence; overclaims: Its account draws on documents obtained by Zitron and describes separate Financial Times review; it is not an original financial statement retrieved for this assessment [e1]..

  183. Assessment publishedBy this siteEd Zitron

    50-60% of NVIDIA's revenue comes from hyperscalers in the AI boom, and without that boom nobody else will buy its chips at anything near today's scale.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-72791e-a2.

  184. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-6e26b6

    The openai review found the criteria unfair to the speaker's words.

  185. Assessment publishedBy this siteEd Zitron

    Despite looking everywhere, Zitron has found at the very best $22 billion of compute purchases by buyers other than OpenAI and Anthropic.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-68f889-a1.

  186. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-5ff073

    a observation claim cannot be assessed on the causal_assessment axis

  187. Assessment publishedBy this siteEd Zitron

    If the $444 billion OpenAI and Anthropic are expected to pay hyperscalers over three years fails to arrive, every connected hyperscaler will miss revenue massively.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-5c0b68-a2.

  189. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-4df8ff

    The final-text check failed: citations_correct.

  190. Assessment publishedBy this siteEd Zitron

    Anthropic's $413 billion in non-cancelable contracts over seven to ten years guarantee financial strain while its future cashflows are uncertain.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-4544b7-a2.

  192. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-3e4179

    The openai review found the criteria unfair to the speaker's words.

  193. Assessment publishedBy this siteEd Zitron

    Data center debt issuance has had some effect but is not the main reason ten-year Treasuries are selling off.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-3a2db5-a1.

  194. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-369409

    The openai review reads the claim differently.

  195. Assessment publishedBy this siteEd Zitron

    Because of hundreds of billions of dollars in non-cancellable commitments, neither OpenAI nor Anthropic can "cut their costs."

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-24a409-a1.

  196. Assessment publishedBy this siteEd Zitron

    Oracle's $18 billion September bond sale would now carry over $7 billion in added interest due to its debt sell-off and the Treasuries market.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-21dfb5-a2.

  198. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-14ee80

    The final-text check failed: excerpt_matches, proposition_faithful; overclaims: The report also does not establish that their purchases created an illusion of broad demand or lacked genuine downstream customers [e1]..

  199. Assessment publishedBy this siteEd Zitron

    AI data center projects not backstopped by hyperscalers will face 9-14% rates, making data center economics near-impossible to work.

    First assessment published after both reviews and the final-text check passed.

    Assessment wyea-credit-crunch-11d459-a2.