Changes, page 32 of 37

Substantive changes (7383)

Page 32 of 37: changes 6201 to 6400 of 7383, newest first. Dates are when the change was recorded.

1. 2026-10-29

   HeldBy this siteGary Marcus

   Held record marcus-c39-82f67b

   The anthropic screen reads the claim differently.

2. 2026-10-29

   Assessment publishedBy this siteGary Marcus

   OpenAI should have presented ARC results for the o3 base model without the pretraining.

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

   Assessment marcus-c39-7dd303-a1.

3. 2026-10-29

   Assessment publishedBy this siteGary Marcus

   OpenAI's graph left out the MIT work (halfway between o1 and o3) and many other results, which made o3's breakthrough relative to the field seem far bigger than it really was.

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

   Assessment marcus-c39-73750f-a1.

4. 2026-10-29

   Assessment publishedBy this siteGary Marcus

   The ARC test does not address factuality, compositionality, or common sense.

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

   Assessment marcus-c39-66ca37-a1.

5. 2026-10-29

   Assessment publishedBy this siteGary Marcus

   Until o3 receives considerable external scientific scrutiny (so far it has had none), we won't really know exactly what the o3 advance is or how important it is.

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

   Assessment marcus-c39-508da0-a1.

6. 2026-10-29

   HeldBy this siteGary Marcus

   Held record marcus-c39-41d678

   The openai review reads the claim differently.

7. 2026-10-29

   Assessment publishedBy this siteGary Marcus

   Passing ARC is at most necessary for AGI and certainly not sufficient.

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

   Assessment marcus-c39-362891-a2.

8. 2026-10-29

   Assessment publishedBy this siteGary Marcus

   o3 as demonstrated is not AGI.

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

   Assessment marcus-c39-333f47-a1.

9. 2026-10-29

   Assessment publishedBy this siteGary Marcus

   The media should ask hard questions about AI instead of fanning hype.

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

   Assessment marcus-c39-30dab9-a1.

10. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Such pretraining is not uncommon in the field, but OpenAI's video did not make clear that o3 had been pretrained on ARC examples.

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

    Assessment marcus-c39-2d2868-a1.

11. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Because of the pretraining and the lack of comparability with humans, what o3 actually showed on ARC was disappointing.

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

    Assessment marcus-c39-0906c4-a1.

12. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    On 2026-06-02 President Trump signed an executive order closely resembling Marcus's proposal for preflight checks on high-impact AI models.

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

    Assessment marcus-breaking-when-dreams-for-ai-sanity-f48d55-a1.

13. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    AI models with large commercial impact should be subject to preflight checks similar to FDA review of drugs.

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

    Assessment marcus-breaking-when-dreams-for-ai-sanity-eea970-a1.

14. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Any preflight-style oversight of AI models like that in President Trump's executive order should be bipartisan and in the interests of American citizens and of humanity as a whole.

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

    Assessment marcus-breaking-when-dreams-for-ai-sanity-ba573f-a1.

15. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The inability to block all jailbreaks is a problem with generative AI in general, not one specific to Anthropic.

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

    Assessment marcus-breaking-trump-asks-the-impossible-fd63dc-a2.

16. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Next-token predictors (LLMs) simply aren't built for safety.

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

    Assessment marcus-breaking-trump-asks-the-impossible-eaf9aa-a1.

17. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-trump-asks-the-impossible-aa96a1

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

18. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-trump-asks-the-impossible-9035a9

    The openai review reads the claim differently.

19. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-trump-asks-the-impossible-84c0d4

    The openai review reads the claim differently.

20. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The politics and inadequacy of generative AI guardrails, which Marcus predicted in 2024 would become central, has now become a central issue, though later than he expected.

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

    Assessment marcus-breaking-trump-asks-the-impossible-3e154e-a1.

21. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Society must either restrict LLMs until a better technology is found or accept the consequences of their unsafe guardrails.

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

    Assessment marcus-breaking-trump-asks-the-impossible-089256-a1.

22. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Trump may retreat further from his pro-AI-industry, anti-regulation stance once he fully grasps how unpopular the AI industry's agenda is.

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

    Assessment marcus-breaking-the-republican-party-is-eb94d9-a2.

23. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The Republican Party is in a political bind over its ties to the AI industry.

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

    Assessment marcus-breaking-the-republican-party-is-97a26d-a1.

24. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Trump has sharply shifted from full opposition to AI regulation to strongly encouraging voluntary preflight testing of frontier AI models, though he has not entirely bolted yet.

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

    Assessment marcus-breaking-the-republican-party-is-918bd1-a2.

25. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The US has built, and continues to build, its entire national policy (such as it is) around an idea, in context the AI 2027 picture of imminent AGI, that was never grounded in a sound analysis of AI.

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-ee21ab-a2.

26. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    AI companies' imminent-AGI stories are greatly flawed, and outside scenarios like AI 2027 distract from those flaws and give more power to the companies racing to AGI.

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-e23da2-a1.

27. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    AGI will not arrive during Trump's second term (ending January 2029).

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-dec445-a1.

28. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-the-ai-2027-doomsday-scenario-db4ec6

    The openai review reads the claim differently.

29. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Not just US policies but also the entire US economy, which is being rebuilt at enormous expense around a myth, need to be rethought.

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-db4ec6-2-a1.

30. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-the-ai-2027-doomsday-scenario-541074

    The anthropic screen reads the claim differently.

31. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Much needs to be reconsidered, including what we should do now given more realistic timelines, how and why so much critical decision-making revolved around sheer speculation, and why dissenting voices were not taken seriously.

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-1fbe10-a1.

32. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-the-ai-2027-doomsday-scenario-1f641e

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

33. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The AI 2027 report is a work of fiction, not a work of science, that stokes fear, uncertainty and doubt.

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-088e20-a1.

34. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The future of AI is very unlikely to unfold as the AI 2027 scenario describes.

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-080c40-a1.

35. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    AGI will not arrive in 2027.

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-07956a-a1.

36. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    AGI will probably not arrive this decade (by the end of 2029).

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

    Assessment marcus-breaking-the-ai-2027-doomsday-scenario-07956a-2-a1.

37. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Essentially anyone who uses a chatbot is at risk of having their beliefs distorted by sycophantic AI.

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

    Assessment marcus-breaking-sycophantic-ai-distorts-d2a507-a1.

38. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-sycophantic-ai-distorts-bc5cc3

    hedged language is not scored, even on a plain reading

39. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    People who want to feel good about themselves can use a chatbot, but people who want to find truth should go elsewhere than chatbots.

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

    Assessment marcus-breaking-sycophantic-ai-distorts-2661b9-a1.

40. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The limits of statistical models argue for using neurosymbolic AI and world (cognitive) models, and his 2020 view on this is still very relevant.

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

    Assessment marcus-breaking-sir-demis-hassabis-becomes-7269bd-a1.

41. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-sam-altmans-greed-and-dishonesty-d386bb

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

42. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-sam-altmans-greed-and-dishonesty-d386bb-2

    The final-text check failed; overclaims: An AGI breakthrough would fall under those concerns..

43. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Altman was never really all that interested in AI for the benefit of humanity; mostly he was interested in himself, and in money and deals.

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

    Assessment marcus-breaking-sam-altmans-greed-and-dishonesty-d2f1c7-a1.

44. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    A person who is not consistently candid, such as Altman, should not be running a company as powerful as OpenAI.

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

    Assessment marcus-breaking-sam-altmans-greed-and-dishonesty-ce8ee0-a1.

45. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The backlash against Altman was entirely predictable; Altman is bad news.

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

    Assessment marcus-breaking-sam-altmans-greed-and-dishonesty-acdecc-a1.

46. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    People who want to use LLMs should use providers other than OpenAI, and those who want to work on LLMs can work elsewhere.

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

    Assessment marcus-breaking-sam-altmans-greed-and-dishonesty-a7eb28-a1.

47. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Public anger at OpenAI on social media is growing fast and a major boycott is underway.

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

    Assessment marcus-breaking-sam-altmans-greed-and-dishonesty-58d426-a2.

48. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Millions of people are angry and many feel betrayed by Altman/OpenAI.

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

    Assessment marcus-breaking-sam-altmans-greed-and-dishonesty-480de1-a2.

49. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    History will judge negatively those who stay at OpenAI.

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

    Assessment marcus-breaking-sam-altmans-greed-and-dishonesty-43d396-a1.

50. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-sam-altmans-greed-and-dishonesty-3c0a04

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

51. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The recent shifts by Altman, Musk and Zuckerberg strongly suggest that AI industry insiders are losing faith in pure scaling.

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

    Assessment marcus-breaking-sam-altman-concedes-that-f2be80-a1.

52. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Because the case for scaling as a road to AGI is falling apart, the plans to spend heavily on data centers should be reconsidered.

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

    Assessment marcus-breaking-sam-altman-concedes-that-bf1735-a1.

53. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    In this talk Altman did not claim to have found such an architecture, which Marcus says is a significant retrenchment from Altman's claim fourteen months earlier that 'We now know how to build AGI as it's usually understood.'

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

    Assessment marcus-breaking-sam-altman-concedes-that-9159ef-a1.

54. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Hassabis, Sutskever and LeCun no longer believe in scaling as the route to AGI, and Nadella and Pichai have hinted at skepticism about it.

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

    Assessment marcus-breaking-sam-altman-concedes-that-905284-a2.

55. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-sam-altman-concedes-that-75cc24

    The anthropic screen reads the claim differently.

56. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-sam-altman-concedes-that-675d06

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

57. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Sam Altman, who ridiculed Marcus's 2022 critique arguing that scaling would not bring us to AGI, has just said he bets there is another new architecture to find that would be as big a gain as transformers were over LSTMs, which Marcus calls a dramatic sign of changing times.

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

    Assessment marcus-breaking-sam-altman-concedes-that-35c3be-a2.

58. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    The bad bargain of AI data centers makes no sense.

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

    Assessment marcus-breaking-sam-altman-concedes-that-0f4ab8-a1.

59. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-unraveling-has-begun-e87e40

    e2 measures something else (metric quarterly operating loss is not quarterly operating loss (including stock-based compensation)); it can only be context unless it is derived with stated inputs; e3 measures something else (metric quarterly operating loss is not quarterly operating loss (including stock-based compensation)); it can only be context unless it is derived with stated inputs

60. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    As of August 2026, the opening stages of OpenAI's unraveling have begun.

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

    Assessment marcus-breaking-openais-unraveling-has-begun-e13d1b-a1.

61. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-unraveling-has-begun-d96f05

    The openai screen reads the claim differently.

62. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-unraveling-has-begun-ab937d

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

63. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-unraveling-has-begun-a489ed

    The anthropic review does not accept e2 as supports.

64. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-unraveling-has-begun-a489ed-2

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

65. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    If OpenAI goes down, the damage could spread.

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

    Assessment marcus-breaking-openais-unraveling-has-begun-940085-a1.

66. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    People were very likely right to be skeptical that OpenAI's training pause was really for safety reasons, given the WSJ's reporting on OpenAI's revenue and losses.

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

    Assessment marcus-breaking-openais-unraveling-has-begun-830688-a1.

67. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Nvidia was trying to hedge its exposure to OpenAI in the week of August 17, 2026.

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

    Assessment marcus-breaking-openais-unraveling-has-begun-7cb8de-a1.

68. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-unraveling-has-begun-3f1e2e

    The anthropic screen reads the claim differently.

69. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Things around OpenAI and the AI industry could well get turbulent very soon.

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

    Assessment marcus-breaking-openais-unraveling-has-begun-24f387-a1.

70. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-unraveling-has-begun-141e83

    The anthropic screen reads the claim differently.

71. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI has been slow at every step in disclosing what happened in the misaligned models incident.

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

    Assessment marcus-breaking-openais-security-fiasco-fd3273-a1.

72. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-security-fiasco-cafe78

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

73. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    President Trump has publicly done nothing to sanction or investigate OpenAI, despite OpenAI's repeated attacks on foreign governments, companies and universities.

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

    Assessment marcus-breaking-openais-security-fiasco-c13f10-a2.

74. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI's software attacked Hugging Face, and its attacks went beyond Hugging Face.

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

    Assessment marcus-breaking-openais-security-fiasco-c03af2-a2.

75. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-security-fiasco-bc4ff6

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

76. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-security-fiasco-b6fa0f

    The anthropic screen reads the claim differently.

77. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI's security situation may well get even worse.

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

    Assessment marcus-breaking-openais-security-fiasco-88d0db-a1.

78. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-security-fiasco-88d0db-2

    The openai review reads the claim differently.

79. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    President Trump has consistently favored OpenAI over Anthropic, and Anthropic does not seem to be as deeply in over its head as OpenAI.

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

    Assessment marcus-breaking-openais-security-fiasco-65d493-a2.

80. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    If Jensen Huang keeps saying the AI labs are in control of their systems, his reputation will rapidly diminish.

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

    Assessment marcus-breaking-openais-security-fiasco-599072-a2.

81. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    There is vast and growing evidence that OpenAI is building systems it is not managing to control.

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

    Assessment marcus-breaking-openais-security-fiasco-4fd9ad-a2.

82. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI's board should replace management with someone far more competent and far more candid.

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

    Assessment marcus-breaking-openais-security-fiasco-32da9a-a1.

83. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI's board has done nothing constructive to address the company's security meltdown.

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

    Assessment marcus-breaking-openais-security-fiasco-21cc72-a1.

84. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI's long tweet did not report how many incidents there had been.

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

    Assessment marcus-breaking-openais-security-fiasco-1f539f-a1.

85. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI should be temporarily shut down now.

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

    Assessment marcus-breaking-openais-security-fiasco-13dd54-a1.

86. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI's software attacked Australia's government servers, among other targets.

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

    Assessment marcus-breaking-openais-security-fiasco-0f671f-a2.

87. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-efforts-at-pure-e9a1e9

    The final-text check failed: proposition_faithful, citations_correct.

88. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-efforts-at-pure-d8ba25

    The anthropic screen reads the claim differently.

89. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Scaling is useful but is not, and never was, sufficient on its own to reach AGI.

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

    Assessment marcus-breaking-openais-efforts-at-pure-d42342-a2.

90. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    AGI will likely arrive some day, likely incorporating many components beyond LLMs.

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

    Assessment marcus-breaking-openais-efforts-at-pure-bf3579-a1.

91. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-efforts-at-pure-ba2b63

    a observation claim cannot be assessed on the causal_assessment axis

92. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI will no longer rely on pure scaling as its means of building GPT-5.

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

    Assessment marcus-breaking-openais-efforts-at-pure-af0412-a1.

93. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-efforts-at-pure-a7f675

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

94. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    OpenAI's future depends on something revolutionary, and Orion is not revolutionary.

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

    Assessment marcus-breaking-openais-efforts-at-pure-95643a-a1.

95. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    If OpenAI called Orion GPT-5, many would likely react with disappointment and OpenAI's mystique might take a serious hit.

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

    Assessment marcus-breaking-openais-efforts-at-pure-8bf018-a1.

96. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-efforts-at-pure-8a0886

    The openai review reads the claim differently.

97. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Marcus restates his long-held prediction that pure scaling of LLMs (adding more data and more parameters) would eventually run out and would fail to solve hallucinations and boneheaded errors, and that scaling laws are merely empirical generalizations rather than physical laws.

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

    Assessment marcus-breaking-openais-efforts-at-pure-4dee04-a1.

98. 2026-10-29

    HeldBy this siteGary Marcus

    Held record marcus-breaking-openais-efforts-at-pure-498aa4

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

99. 2026-10-29

    Assessment publishedBy this siteGary Marcus

    Neurosymbolic AI may be part of the eventual route to AGI.

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

    Assessment marcus-breaking-openais-efforts-at-pure-311b10-a1.

100. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Sam Altman's comments implying progress toward AGI were a bluff.

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

     Assessment marcus-breaking-openais-efforts-at-pure-033013-a1.

101. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Altman does not know how to build AGI, and Orion likely never met the expectations implied by scaling laws despite major costs.

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

     Assessment marcus-breaking-openais-efforts-at-pure-02d6f4-a1.

102. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Jensen Huang's proposal to leave AI safety to the companies makes little sense given the evidence.

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

     Assessment marcus-breaking-openai-was-warned-months-dccac2-a1.

103. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     It is abundantly clear that we simply cannot trust OpenAI.

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

     Assessment marcus-breaking-openai-was-warned-months-b64d92-a1.

104. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     We have to stop pretending that the folks at OpenAI are good stewards of the technology they are developing, because they plainly are not.

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

     Assessment marcus-breaking-openai-was-warned-months-97c4c9-a1.

105. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-openai-was-warned-months-95b49c

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

106. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-openai-was-warned-months-8c8913

     The anthropic review reads the claim differently.

107. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The media should press Jensen Huang on why he defended AI companies, what he knew, and why his judgment should be trusted.

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

     Assessment marcus-breaking-openai-was-warned-months-791038-a1.

108. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Altman probably never told Jensen Huang how much warning Altman himself got or how bad things were, and Huang should not have trusted Altman.

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

     Assessment marcus-breaking-openai-was-warned-months-788555-a1.

109. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI's management should be replaced, and if the board does nothing it should be held liable for allowing the security failures.

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

     Assessment marcus-breaking-openai-was-warned-months-593e5e-a1.

110. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Leaving OpenAI free of serious regulation could have disastrous consequences.

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

     Assessment marcus-breaking-openai-was-warned-months-41b352-a2.

111. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI, which Marcus calls a greedy company, made a bad choice that caused chaos, and the government did nothing to stop it.

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

     Assessment marcus-breaking-openai-was-warned-months-0b24b6-a2.

112. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     If venture capital dries up, OpenAI will collapse and its talent will flee to competitors.

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

     Assessment marcus-breaking-openai-is-probably-toast-e02346-a2.

113. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Google and Anthropic have largely caught up with OpenAI, and Chinese companies are closing in.

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

     Assessment marcus-breaking-openai-is-probably-toast-d0cad9-a1.

114. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-openai-is-probably-toast-c4fa43

     The openai review does not accept e1 as supports.

115. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The US government should not bail out OpenAI, since Google and Anthropic are doing fine without handouts.

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

     Assessment marcus-breaking-openai-is-probably-toast-9fb5a4-a1.

116. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-openai-is-probably-toast-9f0229

     only a conditional forecast records a condition status

117. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI's runway is unlikely to last much more than a year.

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

     Assessment marcus-breaking-openai-is-probably-toast-7aeea8-a1.

118. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI loses money every quarter and, even by its own optimistic projections, will not be profitable for years.

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

     Assessment marcus-breaking-openai-is-probably-toast-795f07-a2.

119. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-openai-is-probably-toast-591f48

     The anthropic review does not accept e2 as contradicts.

120. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-openai-is-probably-toast-495b6e

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

121. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-openai-is-probably-toast-301023

     hedged language is not scored, even on a plain reading

122. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI will likely need a government bailout.

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

     Assessment marcus-breaking-openai-is-probably-toast-0c6824-a1.

123. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Despite reservations about generative AI, the Commerce Department's directive suspending foreign-national access to Anthropic's models is the wrong way to handle AI.

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

     Assessment marcus-breaking-news-us-commerce-department-85ede1-a1.

124. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The US Commerce Department issued an export control directive requiring Anthropic to suspend all access by foreign nationals, including its own employees, to Fable 5 and Mythos 5.

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

     Assessment marcus-breaking-news-us-commerce-department-34fe66-a2.

125. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Things will get worse.

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

     Assessment marcus-breaking-news-us-banks-are-taking-fe5c1e-a1.

126. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Current AI models are already causing serious problems across society, in domains ranging from education to politics (via deepfakes and wholesale disinformation) to banking.

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

     Assessment marcus-breaking-news-us-banks-are-taking-faf812-a1.

127. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-us-banks-are-taking-d92932

     The anthropic screen reads the claim differently.

128. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     AI agents may have facilitated the October 11, 2026 attacks on refrigeration systems in France.

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

     Assessment marcus-breaking-news-us-banks-are-taking-b24ae4-a1.

129. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-us-banks-are-taking-9fb99b

     e4 is dated 2026-10, which is not inside the period described (2026-03-01 to 2026-10-22); it can only be context

130. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-us-banks-are-taking-9adadd

     e1 is dated 2026-10, which is not inside the period described (2026-06-01 to 2026-10-22); it can only be context; e2 is dated 2026-10, which is not inside the period described (2026-06-01 to 2026-10-22); it can only be context

131. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Regulators should not ignore current AI harms in favor of speculative future risks.

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

     Assessment marcus-breaking-news-us-banks-are-taking-952e17-a1.

132. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Even if development of future frontier models stopped, misuse of existing models would still cause serious problems.

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

     Assessment marcus-breaking-news-us-banks-are-taking-8ee762-a2.

133. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     He suggests that the situation he warned about has now arrived, without saying which of his predictions have come true.

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

     Assessment marcus-breaking-news-us-banks-are-taking-8b1e18-a1.

134. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-us-banks-are-taking-8969ab

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

135. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The Trump administration will do little to hold the AI industry responsible for harms.

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

     Assessment marcus-breaking-news-us-banks-are-taking-750863-a1.

136. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     There is a growing, ongoing wave of cyberattacks in Japan.

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

     Assessment marcus-breaking-news-us-banks-are-taking-653f43-a1.

137. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     For months there have been attacks from OpenAI and others on places such as Hugging Face and the Australian government.

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

     Assessment marcus-breaking-news-us-banks-are-taking-524052-a1.

138. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     However much chaos generative AI causes, the AI industry will not take real responsibility.

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

     Assessment marcus-breaking-news-us-banks-are-taking-34aace-a1.

139. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-us-banks-are-taking-26c078

     The anthropic screen finds the claim checkable, so it must be graded.

140. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     AI regulation should not focus only on frontier models or only on risks from future models.

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

     Assessment marcus-breaking-news-us-banks-are-taking-0eeb0b-a1.

141. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Although generative AI is unreliable and prone to hallucinations, it is terrific at making plausible sounding text (such as phishing attacks) and deepfaked videos at virtually zero cost, allowing attackers to broaden their attacks.

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

     Assessment marcus-breaking-news-us-banks-are-taking-055d32-a1.

142. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The line "they hadn't figured out how OpenAI would pay for it" may turn out to be the epitaph for an entire era.

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

     Assessment marcus-breaking-news-they-hadnt-figured-eb3a41-a1.

143. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-they-hadnt-figured-a7e51f

     The anthropic screen doubts the attribution.

144. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-they-hadnt-figured-6360da

     The final-text check failed: rationale_supported_by_evidence; overclaims: The packet gives a May 2026 valuation reference only for OpenAI..

145. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scaling-will-never-f19d72

     The final-text check failed: citations_correct.

146. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scaling-will-never-e2b916

     The final-text check failed: proposition_faithful.

147. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The exponentially growing demand for training data will eventually exceed the data available.

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

     Assessment marcus-breaking-news-scaling-will-never-c46103-a2.

148. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Scaling data for current AI models will yield diminishing returns.

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

     Assessment marcus-breaking-news-scaling-will-never-ba4979-a1.

149. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Neural networks in the configurations dominant for the last three decades have trouble generalizing beyond the space surrounding their training examples.

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

     Assessment marcus-breaking-news-scaling-will-never-98c7b0-a1.

150. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Neural networks' trouble generalizing beyond their training space, at least in the configurations dominant over the last three decades, also drives their greediness with data and even the ethical choices their developers have been making.

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

     Assessment marcus-breaking-news-scaling-will-never-84c11a-a2.

151. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Poor generalization beyond training data is why driverless cars are still just demos.

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

     Assessment marcus-breaking-news-scaling-will-never-739ace-a1.

152. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scaling-will-never-739ace-2

     The anthropic screen finds the claim checkable, so it must be graded.

153. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scaling-will-never-469c79

     The final-text check failed: citations_correct.

154. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Current AI models can generalize to some degree but not enough.

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

     Assessment marcus-breaking-news-scaling-will-never-38a35b-a1.

155. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     AI research should now pursue neurosymbolic AI, world models and causal reasoning.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-f624b0-a1.

156. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-f02354

     The openai review does not accept e1 as supports.

157. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-e4214e

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

158. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The AI field should involve cognitive scientists and stop believing data and compute alone will solve its problems.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-df1f2c-a1.

159. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-dada45

     The anthropic screen reads the claim differently.

160. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The AI field is now starting to realize that scaling alone will not reach AGI.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-ca56f9-a2.

161. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     AI research should look to human and animal cognition for clues, without slavishly replicating it.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-92c28b-a1.

162. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The machine learning community has started to accept critiques of the scaling approach.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-8a7e73-a1.

163. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-819f34

     The openai review does not accept e2 as supports.

164. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     In building AI, knowing what to borrow from human and animal cognition, and what not to borrow, is likely to be more than half the battle.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-71adf7-a1.

165. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     LLM-based tools are unreliable, and more reliable AI technology is badly needed.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-6a2221-a1.

166. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     If we are to build AGI, we are going to need to learn something from humans: how they reason and understand the physical world, and how they represent and acquire language and complex concepts.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-643ff5-a2.

167. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-59953e

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

168. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Scaling may well be better than anything else we currently have, but because it still doesn't really work after all the immense investment in it, that should give us pause.

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

     Assessment marcus-breaking-news-scale-is-all-you-need-4f8b50-a1.

169. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-458c0d

     The openai review does not accept e1 as supports.

170. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-2dc671

     The final-text check failed: proposition_faithful.

171. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-2cf2ef

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

172. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-scale-is-all-you-need-146f57

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

173. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-f3a496

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

174. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-f08d4a

     The final-text check failed: citations_correct; overclaims: Reports that most of Amazon's commitment depends on an IPO or an AGI milestone..

175. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The entire world (hyperbolically) is now worried about whether AI is a bubble, and rightly so.

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

     Assessment marcus-breaking-news-and-how-the-end-might-ecc394-a1.

176. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     SoftBank took an immediate hit, its reduced margin-loan ask reeks of desperation, and the banks still refusing it is a terrible vote of no confidence.

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

     Assessment marcus-breaking-news-and-how-the-end-might-eaf6a8-a1.

177. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     If OpenAI can't meet its obligations, the damage might spread to the banking system, retirement plans and more.

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

     Assessment marcus-breaking-news-and-how-the-end-might-e8256a-a2.

178. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI had a lead in AI but squandered it.

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

     Assessment marcus-breaking-news-and-how-the-end-might-e292ba-a1.

179. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-dc36a7

     e1 measures something else (company OpenAI is not Major LLM providers (e.g., OpenAI, Anthropic, GitHub/Microsoft Copilot)); it can only be context unless it is derived with stated inputs; e2 measures something else (company GitHub (Microsoft) is not Major LLM providers (e.g., OpenAI, Anthropic, GitHub/Microsoft Copilot)); it can only be context unless it is derived with stated inputs; e3 measures something else (company Anthropic is not Major LLM providers (e.g., OpenAI, Anthropic, GitHub/Microsoft Copilot)); it can only be context unless it is derived with stated inputs

180. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-d6cfc9

     The final-text check failed: rationale_supported_by_evidence; overclaims: The evidence packet has no report that OpenAI failed to meet its obligations by 28 October 2026, so the condition has not occurred. | The claim's antecedent has not occurred. | OpenAI is private, so a missed or renegotiated obligation might not become public..

181. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI has a problem and is burning money faster than its AI competitors.

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

     Assessment marcus-breaking-news-and-how-the-end-might-d51f62-a2.

182. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-cf223f

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

183. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-c65451

     only a conditional forecast records a condition status

184. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI is likely the first major AI company to fail.

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

     Assessment marcus-breaking-news-and-how-the-end-might-a9c1e1-a1.

185. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     OpenAI, in rushing its IPO ahead of Anthropic, is clearly afraid of Anthropic.

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

     Assessment marcus-breaking-news-and-how-the-end-might-a70398-a1.

186. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     If the market turns skittish on OpenAI's IPO (a real possibility), OpenAI might be forced to withdraw its IPO or sharply cut its aspirations.

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

     Assessment marcus-breaking-news-and-how-the-end-might-a67eef-a2.

187. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Every warning sign about OpenAI is worse now than two years ago, and people are finally noticing.

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

     Assessment marcus-breaking-news-and-how-the-end-might-a4c0b0-a2.

188. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The moment when the AI boom collapses, as everyone realizes they got it wrong, could come soon.

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

     Assessment marcus-breaking-news-and-how-the-end-might-8f3de9-a1.

189. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     The giant GenAI investment bet has a low but nonzero probability of paying off.

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

     Assessment marcus-breaking-news-and-how-the-end-might-7226e0-a1.

190. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Open-source models, perhaps excluding Meta's, are only a few months behind frontier models.

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

     Assessment marcus-breaking-news-and-how-the-end-might-722136-a1.

191. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     Over the past two years competitors have caught up to OpenAI and it has had to cut prices.

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

     Assessment marcus-breaking-news-and-how-the-end-might-665795-a1.

192. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     If OpenAI can't meet its obligations, Nvidia, Oracle and many others would take a big hit as collateral damage.

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

     Assessment marcus-breaking-news-and-how-the-end-might-5ddefc-a2.

193. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     It is pretty clear that Sam Altman should not be trusted.

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

     Assessment marcus-breaking-news-and-how-the-end-might-5cc4e1-a1.

194. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-5595dd

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

195. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-4aa1c7

     The final-text check failed: conditions_shown.

196. 2026-10-29

     Assessment publishedBy this siteGary Marcus

     AI agents tend to be more expensive because they tend to run multiple large language models in the background many times until they get answers that by some criteria seem correct.

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

     Assessment marcus-breaking-news-and-how-the-end-might-469c5d-a1.

197. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-42855b

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

198. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-36e433

     e3 is dated 2026-06, which is not inside the period described (2026-01-01 to 2026-06-10); it can only be context

199. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-36968e

     a causal claim cannot be assessed on the forecast_outcome axis; e1 measures something else (company Anthropic is not LLM companies (Anthropic, OpenAI)); it can only be context unless it is derived with stated inputs; e2 measures something else (company Anthropic is not LLM companies (Anthropic, OpenAI)); it can only be context unless it is derived with stated inputs; e3 is decisive for a financial claim but states no metric scope

200. 2026-10-29

     HeldBy this siteGary Marcus

     Held record marcus-breaking-news-and-how-the-end-might-32d3c4

     hedged language is not scored, even on a plain reading