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

Changes, page 4 of 52

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 (10273)

Page 4 of 52: changes 601 to 800 of 10273, newest first. Dates are when the change was recorded.

  1. Assessment publishedBy this siteEd Zitron

    OpenAI would need to grow ChatGPT revenue by 310% to reach $11.6B in 2025.

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

    Assessment wyea-oai-business-7d46c5-a2.

  2. HeldBy this siteEd Zitron

    Held record wyea-oai-business-3fa9f5

    The openai review reads the claim differently.

  3. HeldBy this siteEd Zitron

    Held record wyea-oai-business-2594ba

    hedged language is not scored, even on a plain reading

  4. Assessment publishedBy this siteEd Zitron

    ChatGPT's heaviest free users become less valuable over time and a cost burden, unlike free users of Facebook or Google.

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

    Assessment wyea-oai-business-1c126f-a1.

  5. Assessment publishedBy this siteEd Zitron

    Free ChatGPT is effectively identical to free Claude, Meta AI, Copilot and Grok; the product is extremely commoditized.

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

    Assessment wyea-oai-business-11d323-a1.

  6. Assessment publishedBy this siteEd Zitron

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

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

    Assessment wyea-nvidia-isnt-enron-so-what-is-it-ff3d18-a1.

  7. HeldBy this siteEd Zitron

    Held record wyea-nvidia-isnt-enron-so-what-is-it-ee1795

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

  8. Assessment publishedBy this siteEd Zitron

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

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

    Assessment wyea-nvidia-isnt-enron-so-what-is-it-ec6c1c-a1.

  9. Assessment publishedBy this siteEd Zitron

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

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

    Assessment wyea-nvidia-isnt-enron-so-what-is-it-bdd949-a2.

  10. HeldBy this siteEd Zitron

    Held record wyea-nvidia-isnt-enron-so-what-is-it-b9bcd2

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

  11. HeldBy this siteEd Zitron

    Held record wyea-nvidia-isnt-enron-so-what-is-it-b4c37c

    e3 is decisive for a financial claim but states no metric scope

  12. HeldBy this siteEd Zitron

    Held record wyea-nvidia-isnt-enron-so-what-is-it-aa0391

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

  13. Assessment publishedBy this siteEd Zitron

    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.

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

    Assessment wyea-nvidia-isnt-enron-so-what-is-it-9af629-a2.

  14. Assessment publishedBy this siteEd Zitron

    Data center GPU depreciation schedules are based at best on assumptions and at worst on hope.

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

    Assessment wyea-nvidia-isnt-enron-so-what-is-it-58c63d-a2.

  15. Assessment publishedBy this siteEd Zitron

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

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

    Assessment wyea-nvidia-isnt-enron-so-what-is-it-1bfead-a1.

  16. HeldBy this siteEd Zitron

    Held record wyea-news-openai-had-a-negative-122-operating-margin-in-q1-2026-and-dcffa979509f-bd9beb

    a observation claim cannot be assessed on the causal_assessment axis

  17. HeldBy this siteEd Zitron

    Held record wyea-news-openai-had-a-negative-122-operating-margin-in-q1-2026-and-dcffa979509f-bd5eb5

    hedged language is not scored, even on a plain reading

  18. HeldBy this siteEd Zitron

    Held record wyea-news-openai-had-a-negative-122-operating-margin-in-q1-2026-and-dcffa979509f-607888

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

  21. Assessment publishedBy this siteEd Zitron

    Bizarre AI-generated slop is dominating Facebook feeds.

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

    Assessment wyea-never-forgive-them-4f61de-a1.

  22. HeldBy this siteEd Zitron

    Held record wyea-never-forget-what-theyve-done-f3a9f7

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

  23. HeldBy this siteEd Zitron

    Held record wyea-never-forget-what-theyve-done-b8d81a

    The final-text check failed: citations_correct.

  24. Assessment publishedBy this siteEd Zitron

    LLMs are famous for their inconsistency rather than their efficacy, and even ardent AI supporters warn people not to trust their output.

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

    Assessment wyea-never-forget-what-theyve-done-822d24-a1.

  25. HeldBy this siteEd Zitron

    Held record wyea-never-forget-what-theyve-done-5fee70

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

  26. Assessment publishedBy this siteEd Zitron

    ChatGPT currently cannot be relied on to do what it's asked and its capabilities are unclear; a better ChatGPT would be a different product.

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

    Assessment wyea-never-forget-what-theyve-done-1a44f4-a1.

  27. HeldBy this siteEd Zitron

    Held record wyea-negative-space-94f92f

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

  28. HeldBy this siteEd Zitron

    Held record wyea-measures-d06bc5

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

  29. HeldBy this siteEd Zitron

    Held record wyea-measures-412374

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

  30. Assessment publishedBy this siteEd Zitron

    As of May 2025, no real information exists about the AI device OpenAI and Jony Ive are developing.

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

    Assessment wyea-measures-064422-a2.

  31. HeldBy this siteEd Zitron

    Held record wyea-managing-up-993ea0

    e1 is dated 2024-04, which is not inside the period described (2024-04-01 to 2024-04-29); it can only be context; e2 is dated 2024-04, which is not inside the period described (2024-04-01 to 2024-04-29); it can only be context; e4 is dated 2024-04, which is not inside the period described (2024-04-01 to 2024-04-29); it can only be context

  32. Assessment publishedBy this siteEd Zitron

    Improvement in LLM-based services like ChatGPT has slowed and even regressed.

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

    Assessment wyea-managing-up-7dc921-a1.

  33. Assessment publishedBy this siteEd Zitron

    Generative AI doesn't really help anyone other than those who want to sell something.

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

    Assessment wyea-managing-up-19cf1d-a1.

  34. HeldBy this siteEd Zitron

    Held record wyea-make-fun-of-them-e326e2

    hedged language is not scored, even on a plain reading

  35. HeldBy this siteEd Zitron

    Held record wyea-make-fun-of-them-deb642

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

  36. Assessment publishedBy this siteEd Zitron

    Nearly $500 billion in capex and VC has produced a generative AI industry with revenue comparable to the smartwatch industry.

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

    Assessment wyea-make-fun-of-them-ca5a7c-a1.

  37. Assessment publishedBy this siteEd Zitron

    OpenAI is an unstable monolith with a vague product that can only make $10 billion a year in revenue, not profit.

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

    Assessment wyea-make-fun-of-them-c63e74-a1.

  38. Assessment publishedBy this siteEd Zitron

    OpenAI plans to lose over $100 billion before becoming profitable.

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

    Assessment wyea-make-fun-of-them-9222ca-2-a1.

  39. HeldBy this siteEd Zitron

    Held record wyea-make-fun-of-them-79a5d7

    The two reviews differ: anthropic forecast_outcome/open_horizon/not_adjudicable, openai forecast_outcome/open_horizon/unresolved.

  40. Assessment publishedBy this siteEd Zitron

    Microsoft bloats Word and Excel with odd interface choices and makes users fight with popups to use an AI-powered Copilot that most of them hate.

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

    Assessment wyea-make-fun-of-them-4c96cd-a1.

  41. HeldBy this siteEd Zitron

    Held record wyea-make-fun-of-them-3ecb07

    The final-text check failed: citations_correct.

  42. HeldBy this siteEd Zitron

    Held record wyea-make-fun-of-them-388601

    hedged language is not scored, even on a plain reading

  43. Assessment publishedBy this siteEd Zitron

    By his tally of run rates and projections, the entire generative AI industry has less than $39 billion in total revenue.

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

    Assessment wyea-make-fun-of-them-1b88fb-a1.

  44. Assessment publishedBy this siteEd Zitron

    The generative AI industry cannot generate even a tenth of the $370.8 billion in investment and capex it requires.

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

    Assessment wyea-make-fun-of-them-140c13-a1.

  45. HeldBy this siteEd Zitron

    Held record wyea-make-fun-of-them-13bc19

    The final-text check failed: proposition_faithful.

  46. HeldBy this siteEd Zitron

    Held record wyea-lost-in-the-future-ba5f2a

    The final-text check failed: citations_correct.

  47. HeldBy this siteEd Zitron

    Held record wyea-lost-in-the-future-b684cb

    The anthropic review does not accept e1 as contradicts.

  48. HeldBy this siteEd Zitron

    Held record wyea-lost-in-the-future-94b06e

    The anthropic review does not accept e2 as supports.

  49. HeldBy this siteEd Zitron

    Held record wyea-lost-in-the-future-322333

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

  50. Assessment publishedBy this siteEd Zitron

    Apart from perhaps Facebook, no startup has had as much media coverage so early as ChatGPT/OpenAI.

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

    Assessment wyea-longcon-e92d49-a2.

  52. Assessment publishedBy this siteEd Zitron

    The generative AI boom will fall apart, leading to a very public reckoning for the tech industry.

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

    Assessment wyea-longcon-cc3598-a1.

  53. Assessment publishedBy this siteEd Zitron

    Microsoft's $13bn AI run-rate figure is not a distinct reported segment but a relabelling of anything with AI.

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

    Assessment wyea-longcon-aba9b6-a2.

  54. Assessment publishedBy this siteEd Zitron

    Deep Research queries cost significantly more than regular ChatGPT queries.

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

    Assessment wyea-longcon-aa8385-a1.

  55. HeldBy this siteEd Zitron

    Held record wyea-longcon-9d1c36

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

  57. HeldBy this siteEd Zitron

    Held record wyea-longcon-61855a

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

  58. Assessment publishedBy this siteEd Zitron

    OpenAI spent about $5 billion on compute alone in 2024.

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

    Assessment wyea-longcon-501d82-a2.

  59. HeldBy this siteEd Zitron

    Held record wyea-longcon-4c221d

    hedged language is not scored, even on a plain reading

  60. HeldBy this siteEd Zitron

    Held record wyea-longcon-2fbe46

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

  61. Assessment publishedBy this siteEd Zitron

    Anthropic's $60bn valuation fundraise rests primarily on inflated marketing claims.

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

    Assessment wyea-longcon-159db0-a1.

  62. Assessment publishedBy this siteEd Zitron

    Nothing about Deep Research moves OpenAI toward profitability, and he thinks it moves OpenAI in the opposite direction.

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

    Assessment wyea-longcon-14c7aa-a2.

  63. Assessment publishedBy this siteEd Zitron

    OpenAI burned more than $5 billion in 2024.

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

    Assessment wyea-longcon-04441f-a1.

  64. Assessment publishedBy this siteEd Zitron

    The AI industry is not building autonomous intelligence and never will.

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

    Assessment wyea-let-ai-burn-efdf5f-a2.

  65. Assessment publishedBy this siteEd Zitron

    Anthropic's and OpenAI's services are unsustainable and unprofitable, and buying more capacity makes them lose more money.

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

    Assessment wyea-i-will-never-respect-a-website-fe768d-a1.

  66. HeldBy this siteEd Zitron

    Held record wyea-i-will-never-respect-a-website-f73f38

    The anthropic review does not accept e2 as contradicts.

  69. HeldBy this siteEd Zitron

    Held record wyea-i-will-never-respect-a-website-9b8bb8

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

  70. Assessment publishedBy this siteEd Zitron

    Because OpenAI and Anthropic must show revenue growth and keep capacity for new customers, existing customers will always be punished first.

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

    Assessment wyea-i-will-never-respect-a-website-93fc65-a1.

  72. HeldBy this siteEd Zitron

    Held record wyea-i-will-never-respect-a-website-6889e6

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

  74. HeldBy this siteEd Zitron

    Held record wyea-i-will-never-respect-a-website-248e74

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

  75. Assessment publishedBy this siteEd Zitron

    AI companies are inherently unsustainable because of unreliable LLM outputs and high service costs.

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

    Assessment wyea-i-will-never-respect-a-website-155fdb-a1.

  76. HeldBy this siteEd Zitron

    Held record wyea-i-will-never-respect-a-website-077569

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

  77. HeldBy this siteEd Zitron

    Held record wyea-hyperscale-normalization-cd57cf

    The final-text check failed: citations_correct.

  78. HeldBy this siteEd Zitron

    Held record wyea-hyperscale-normalization-b972ae

    The openai review reads the claim differently.

  79. HeldBy this siteEd Zitron

    Held record wyea-hyperscale-normalization-9c2e9d

    The final-text check failed: rationale_supported_by_evidence, citations_correct; overclaims: The report says this included an approximately $30 billion nonrecurring accounting charge [e1]..

  80. HeldBy this siteEd Zitron

    Held record wyea-hyperscale-normalization-94a81f

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

  81. Assessment publishedBy this siteEd Zitron

    A finished AI data center creates roughly 100–200 jobs and brings little or no economic value.

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

    Assessment wyea-hyperscale-normalization-7a0f9b-a1.

  82. Assessment publishedBy this siteEd Zitron

    Industries created by AI, such as companies wrapping existing models, employ a negligible number of people.

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

    Assessment wyea-hyperscale-normalization-702ad9-a2.

  83. Assessment publishedBy this siteEd Zitron

    Anthropic's actual revenue in Q2 2026 was $11.6 billion, and AI companies do not disclose actual revenues or underlying economics.

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

    Assessment wyea-hyperscale-normalization-4cbdce-a1.

  84. Assessment publishedBy this siteEd Zitron

    Anthropic's reported $65 billion annualized run rate is an undefined figure that was reported uncritically.

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

    Assessment wyea-hyperscale-normalization-36ffb9-a1.

  85. HeldBy this siteEd Zitron

    Held record wyea-how-to-argue-with-an-ai-booster-dd7d5f

    hedged language is not scored, even on a plain reading

  86. HeldBy this siteEd Zitron

    Held record wyea-how-to-argue-with-an-ai-booster-c54af0

    The openai review reads the claim differently.

  87. Assessment publishedBy this siteEd Zitron

    OpenAI's growth will eventually slow because businesses are not getting real returns from AI.

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

    Assessment wyea-how-to-argue-with-an-ai-booster-a3e27b-a1.

  88. HeldBy this siteEd Zitron

    Held record wyea-how-to-argue-with-an-ai-booster-9fffa9

    The openai review reads the claim differently.

  89. Assessment publishedBy this siteEd Zitron

    Someone who cannot code cannot build a fully functional, secure, useful app entirely by vibe coding.

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

    Assessment wyea-how-to-argue-with-an-ai-booster-965aba-a1.

  90. Assessment publishedBy this siteEd Zitron

    He doesn't believe OpenAI will continue growing forever, and OpenAI's poor conversion rate isn't going to change anytime soon.

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

    Assessment wyea-how-to-argue-with-an-ai-booster-937bc5-a2.

  91. HeldBy this siteEd Zitron

    Held record wyea-how-to-argue-with-an-ai-booster-9274fa

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

  92. Assessment publishedBy this siteEd Zitron

    ChatGPT is essentially the same product as Claude, Grok and other LLMs.

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

    Assessment wyea-how-to-argue-with-an-ai-booster-5a16d7-a1.

  93. Assessment publishedBy this siteEd Zitron

    There is no proof that AI is why young people are struggling to find jobs.

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

    Assessment wyea-how-to-argue-with-an-ai-booster-37780a-a1.

  94. Assessment publishedBy this siteEd Zitron

    AI foundation models will not do scientific research.

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

    Assessment wyea-how-to-argue-with-an-ai-booster-2883f4-a2.

  95. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-f53615

    hedged language is not scored, even on a plain reading

  97. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-ddd093

    The final-text check failed: citations_correct.

  98. Assessment publishedBy this siteEd Zitron

    Anthropic is on course to lose more than $2.7 billion in 2024.

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

    Assessment wyea-godot-isnt-making-it-ddd093-2-a2.

  99. Assessment publishedBy this siteEd Zitron

    Generative AI products have not meaningfully progressed for over a year.

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

    Assessment wyea-godot-isnt-making-it-d9b083-a1.

  100. Assessment publishedBy this siteEd Zitron

    Generative AI is a dead end because probabilistic models hallucinate, stating false things authoritatively.

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

    Assessment wyea-godot-isnt-making-it-d80c97-a1.

  101. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-d210c7

    hedged language is not scored, even on a plain reading

  102. Assessment publishedBy this siteEd Zitron

    If NVIDIA does not solve Blackwell's cooling problem well, customers will encounter thermal throttling.

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

    Assessment wyea-godot-isnt-making-it-bedfc0-a1.

  103. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-a00824

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

  104. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-9d1910

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

  105. Assessment publishedBy this siteEd Zitron

    Big tech's LLMs are increasingly expensive, near-identical iterations because they all use the same training data, which is running out.

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

    Assessment wyea-godot-isnt-making-it-85796e-a1.

  106. Assessment publishedBy this siteEd Zitron

    Generative AI is a dead-end technology whose inessential products cost more to serve than anyone would pay.

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

    Assessment wyea-godot-isnt-making-it-738efb-a1.

  107. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-5f5628

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

  108. Assessment publishedBy this siteEd Zitron

    Capital expenditures on generative AI and infrastructure exceeded $200 billion in the last year.

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

    Assessment wyea-godot-isnt-making-it-5955f2-a2.

  109. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-48334a

    hedged language is not scored, even on a plain reading

  110. Assessment publishedBy this siteEd Zitron

    If transformer models are at diminishing returns, it will be extremely difficult to justify buying NVIDIA GPUs beyond Blackwell.

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

    Assessment wyea-godot-isnt-making-it-3f1df7-a2.

  111. HeldBy this siteEd Zitron

    Held record wyea-godot-isnt-making-it-36d73e

    The openai review reads the claim differently.

  112. Assessment publishedBy this siteEd Zitron

    OpenAI currently spends $2.35 for every $1 of revenue it makes.

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

    Assessment wyea-godot-isnt-making-it-254737-a1.

  113. Assessment publishedBy this siteEd Zitron

    Generative AI is incredibly unprofitable.

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

    Assessment wyea-godot-isnt-making-it-03878e-a2.

  114. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-f68e7b

    The final-text check failed: citations_correct.

  115. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-ea438f

    The two reviews differ: anthropic forecast_outcome/open_horizon/not_adjudicable, openai forecast_outcome/open_horizon/unresolved.

  116. Assessment publishedBy this siteEd Zitron

    Most of OpenAI's, Cursor's and Anthropic's revenue comes from subsidized software subscriptions.

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

    Assessment wyea-four-horsemen-of-the-aipocalypse-dc2623-a2.

  117. Assessment publishedBy this siteEd Zitron

    Anthropic cannot afford to provide a stable and reliable service to its customers.

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

    Assessment wyea-four-horsemen-of-the-aipocalypse-d94798-a2.

  118. Assessment publishedBy this siteEd Zitron

    Hundreds of billions of dollars of GPUs are sitting uninstalled in warehouses, with a quarter's GPU sales taking six months to install.

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

    Assessment wyea-four-horsemen-of-the-aipocalypse-d46332-a1.

  119. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-d20362

    The anthropic review does not accept e2 as contradicts.

  120. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-c887ca

    The final-text check failed: proposition_faithful.

  121. Assessment publishedBy this siteEd Zitron

    AI demand is inflated by token-maxxing and subsidized services.

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

    Assessment wyea-four-horsemen-of-the-aipocalypse-99734a-a2.

  122. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-8feffd

    a observation claim cannot be assessed on the causal_assessment axis

  123. Assessment publishedBy this siteEd Zitron

    Venture capital funding allows Anthropic to massively subsidize its services.

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

    Assessment wyea-four-horsemen-of-the-aipocalypse-8e3ee9-a2.

  124. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-81e0f4

    The final-text check failed: proposition_faithful; overclaims: Normalised proposition: '...below the four-nines standard.'.

  126. Assessment publishedBy this siteEd Zitron

    NVIDIA is selling more GPUs than can possibly be installed.

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

    Assessment wyea-four-horsemen-of-the-aipocalypse-6aa69e-a1.

  127. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-63d5b6

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

  128. Assessment publishedBy this siteEd Zitron

    Because Anthropic's capacity issues are not affecting its growth, adding more capacity simply makes the product work better for a much higher cost.

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

    Assessment wyea-four-horsemen-of-the-aipocalypse-5bf297-a1.

  129. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-36d665

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

  130. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-36a305

    The anthropic review reads the claim differently.

  132. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-118e35

    The two reviews differ: anthropic forecast_outcome/open_horizon/not_adjudicable, openai forecast_outcome/open_horizon/unresolved.

  133. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-0af810

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

  134. HeldBy this siteEd Zitron

    Held record wyea-four-horsemen-of-the-aipocalypse-0aad8a

    The final-text check failed: proposition_faithful.

  135. HeldBy this siteEd Zitron

    Held record wyea-expectations-versus-reality-e2773d

    The anthropic review reads the claim differently.

  136. HeldBy this siteEd Zitron

    Held record wyea-expectations-versus-reality-cecde8

    The final-text check failed: citations_correct.

  137. HeldBy this siteEd Zitron

    Held record wyea-expectations-versus-reality-6ac1ed

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

  138. Assessment publishedBy this siteEd Zitron

    Anthropic's compute costs, hundreds of billions due over the next few years, will increase dramatically.

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

    Assessment wyea-exit-liquidity-683c04-a2.

  139. HeldBy this siteEd Zitron

    Held record wyea-dont-look-up-0e8993

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

  140. HeldBy this siteEd Zitron

    Held record wyea-disruption-killed-innovation-d2bd26

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

  141. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-7b0c7a

    The final-text check failed: citations_correct.

  142. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-6c92cd

    e1 is dated 2025-01, which is not inside the period described (2025-01-01 to 2025-01-29); it can only be context

  144. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-491ea1

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

  145. HeldBy this siteEd Zitron

    Held record wyea-deep-impact-26b61f

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

  146. Assessment publishedBy this siteEd Zitron

    The roughly $800 billion invested in AI startups over the past four years has no real hope of being recovered.

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

    Assessment wyea-credit-crunch-bf0839-a1.

  147. HeldBy this siteEd Zitron

    Held record wyea-credit-crunch-54ad05

    The final-text check failed: citations_correct.

  150. HeldBy this siteEd Zitron

    Held record wyea-costs-d797fd

    hedged language is not scored, even on a plain reading

  151. HeldBy this siteEd Zitron

    Held record wyea-costs-c5038f

    hedged language is not scored, even on a plain reading

  152. HeldBy this siteEd Zitron

    Held record wyea-costs-b512d9

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

  154. HeldBy this siteEd Zitron

    Held record wyea-costs-ae9ac2

    hedged language is not scored, even on a plain reading

  155. HeldBy this siteEd Zitron

    Held record wyea-costs-9f9b2a

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

  156. HeldBy this siteEd Zitron

    Held record wyea-costs-7063fe

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

  157. Assessment publishedBy this siteEd Zitron

    The AWS billing data seen for Anthropic and Cursor shows an overall direction of travel in which compute costs only keep going up.

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

    Assessment wyea-costs-5ec17c-a1.

  158. Assessment publishedBy this siteEd Zitron

    AI-native companies will see multiple price increases from model providers in the next few months.

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

    Assessment wyea-costs-382388-a1.

  159. Assessment publishedBy this siteEd Zitron

    Through September 2025 Anthropic spent $2.66 billion on AWS against an estimated $2.55 billion in revenue, more than 100% of its revenue.

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

    Assessment wyea-costs-259ad1-a2.

  160. HeldBy this siteEd Zitron

    Held record wyea-costs-0d0c5d

    hedged language is not scored, even on a plain reading

  161. Assessment publishedBy this siteEd Zitron

    CoreWeave's financials suggest there is either no demand or no profit in providing generative AI services.

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

    Assessment wyea-core-incompetency-993664-a2.

  162. Assessment publishedBy this siteEd Zitron

    The market for by-the-hour AI compute sold to regular companies does not exist at meaningful revenue.

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

    Assessment wyea-core-incompetency-741de4-a1.

  163. HeldBy this siteEd Zitron

    Held record wyea-core-incompetency-6deed1

    The final-text check failed: rationale_supported_by_evidence.

  164. HeldBy this siteEd Zitron

    Held record wyea-cargo-culture-c1a498

    The final-text check failed: proposition_faithful, citations_correct.

  165. Assessment publishedBy this siteEd Zitron

    OpenAI spent $8.67 billion on inference in the first nine months of 2025, undermining claims that its inference is profitable.

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

    Assessment wyea-cargo-culture-874112-a1.

  166. HeldBy this siteEd Zitron

    Held record wyea-cargo-culture-785c33

    The final-text check failed: rationale_supported_by_evidence, citations_correct.

  167. HeldBy this siteEd Zitron

    Held record wyea-cargo-culture-149d3d

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

  168. Assessment publishedBy this siteEd Zitron

    Zitron fully believes some low-fi version of generative AI will plug along silently once the hyperscalers flee.

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

    Assessment wyea-burst-damage-95765f-a2.

  169. HeldBy this siteEd Zitron

    Held record wyea-bubble-trouble-a3ceda

    The openai review reads the claim differently.

  170. HeldBy this siteEd Zitron

    Held record wyea-bubble-trouble-3f6607

    The openai review reads the claim differently.

  171. HeldBy this siteEd Zitron

    Held record wyea-bubble-trouble-2b7401

    The final-text check failed: excerpt_matches.

  173. HeldBy this siteEd Zitron

    Held record wyea-are-we-watching-the-internet-die-b46ea1

    The final-text check failed: citations_correct.

  174. Assessment publishedBy this siteEd Zitron

    The NVIDIA GPUs big tech bought with ~$850B capex since 2023 will be nearly completely obsolete by 2030.

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

    Assessment wyea-am-i-meant-to-be-impressed-fc83f7-a1.

  175. Assessment publishedBy this siteEd Zitron

    Meta, Microsoft, Google and Amazon will not die as a result of the AI bubble bursting.

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

    Assessment wyea-am-i-meant-to-be-impressed-eaeb30-a2.

  176. HeldBy this siteEd Zitron

    Held record wyea-am-i-meant-to-be-impressed-e99ec4

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

  177. HeldBy this siteEd Zitron

    Held record wyea-am-i-meant-to-be-impressed-e89892

    The final-text check failed: proposition_faithful, citations_correct; overclaims: He personally thinks Anthropic's implausibly fast revenue numbers are explained by it doing something peculiar with its annualized revenue calculations..

  178. Assessment publishedBy this siteEd Zitron

    Outside OpenAI and Anthropic there is barely any AI demand at the scale needed to justify a trillion dollars in capex.

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

    Assessment wyea-am-i-meant-to-be-impressed-4fa453-a1.

  179. HeldBy this siteEd Zitron

    Held record wyea-ais-economics-dont-make-sense-f016a3

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

  180. HeldBy this siteEd Zitron

    Held record wyea-ais-economics-dont-make-sense-cde3ba

    a observation claim cannot be assessed on the forecast_outcome axis

  182. Assessment publishedBy this siteEd Zitron

    LLM-based subscriptions will go away entirely, at least for code-generating products.

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

    Assessment wyea-ais-economics-dont-make-sense-853e46-a1.

  183. HeldBy this siteEd Zitron

    Held record wyea-ais-economics-dont-make-sense-745a5a

    The final-text check failed: citations_correct.

  184. Assessment publishedBy this siteEd Zitron

    Hyperscalers will not recover their AI investments without a massive new product line or three or four more Anthropic/OpenAI-sized compute contracts.

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

    Assessment wyea-ai-is-too-expensive-ffe4e5-a2.

  185. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-d4f8aa

    hedged language is not scored, even on a plain reading

  187. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-9dba55

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

  188. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-8e234e

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

  189. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-834bfc

    The openai review reads the claim differently.

  190. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-7e93aa

    The final-text check failed: proposition_faithful.

  191. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-744eed

    The anthropic review reads the claim differently.

  192. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-54f878

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

  193. Assessment publishedBy this siteEd Zitron

    Every organization the author spoke to has exceeded or nearly exhausted its annual AI token budget five months into 2026.

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

    Assessment wyea-ai-is-too-expensive-51c48c-a2.

  194. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-39f0b3

    The openai review doubts the attribution.

  195. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-176c49

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

  196. HeldBy this siteEd Zitron

    Held record wyea-ai-is-too-expensive-098caa

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

  197. HeldBy this siteEd Zitron

    Held record wyea-ai-is-slowing-down-a128c7

    The final-text check failed: proposition_faithful, citations_correct.

  198. HeldBy this siteEd Zitron

    Held record wyea-ai-is-slowing-down-49f0ff

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

  199. Assessment publishedBy this siteEd Zitron

    TokenMaxxing spending is exactly the kind of revenue that disappears first.

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

    Assessment wyea-ai-is-really-weird-cc44bd-a1.

  200. Assessment publishedBy this siteEd Zitron

    Anthropic's revenue is roughly 75% API and 25% subscriptions, by the author's estimate.

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

    Assessment wyea-ai-is-really-weird-9c15c4-a3.