Changes, page 15 of 37

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 (7383)

Page 15 of 37: changes 2801 to 3000 of 7383, newest first. Dates are when the change was recorded.

2026-10-29
Assessment publishedBy this siteGary Marcus
Amodei and Anthropic are rapidly following Altman's fall from grace.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-fe6b60-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-f89798
the horizon on 2026-10-28 is open_horizon, not not_due
2026-10-29
Assessment publishedBy this siteGary Marcus
Amodei and Altman have repeatedly hinted AGI is near and quietly moved the deadline back when it didn't arrive.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-edd129-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
There is zero chance that human life span is doubled in the next decade.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-e3f509-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Dario Amodei is not all that different from Sam Altman.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-dea9b2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The results of AI-assisted targeting in Iran have probably already been disastrous, even supposing humans were in the loop.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-c6bad8-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI trained on practically anything publicly available, whether public domain or not.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-c0916d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Amodei's claims about AI are ludicrous at best and dishonest at worst.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-a12aa9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Overtrust in AI, already seen in law and journalism, is likely now being seen in war.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-8a85e5-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-865247
e4 is decisive for a financial claim but states no metric scope
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-7bb04c
The final-text check failed: rationale_supported_by_evidence, citations_correct; overclaims: The September 24, 2026 account describes compressed review and expectations that Maven would detect intelligence problems in the February 28 Minab strike.
2026-10-29
Assessment publishedBy this siteGary Marcus
Aside from his courageous stand on mass surveillance and fully autonomous weapons, Amodei is more like Altman than almost anyone Marcus can think of.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-769bee-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
In many other ways, Amodei pretty much follows Altman's playbook of constant hype and overpromising, frequently moving deadlines back and rarely taking responsibility.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-761128-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-63a87b
The anthropic review does not accept e1 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
If the AI is selecting 80 targets an hour, as has been reported, the humans in the loop probably aren't going to do a good job verifying those targets.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-5a3fae-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman and Amodei both repeatedly use apocalyptic talk, a special case of hype, which basically amounts to saying their tech is so scary it might ruin society so they should be given more money to build it ever faster.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-52c377-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Claude may well have selected the Iranian elementary school that was struck.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-448fb2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-3e166b
hedged language is not scored, even on a plain reading
2026-10-29
Assessment publishedBy this siteGary Marcus
Claude is, or at least was and may again be, part of the war machine, and even with humans putatively in the loop it may have been responsible for many deaths.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-there-are-no-heroes-in-commercial-392cc5-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-2e5c34
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-23ef2c
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-1299d0
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-there-are-no-heroes-in-commercial-0c0168
The final-text check failed: proposition_faithful, citations_correct.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI is currently damaging science.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-f16310-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
GenAI-generated content in papers is a potential major crisis for journal publishers because their value rests on reputation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-ecfc12-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs tend to hallucinate in plausible ways that may escape notice, so the lithium battery paper may contain other errors.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-c8975c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLM developers are unlikely to be held accountable for GenAI damage to scientific publishing.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-b94d5c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI's damage to science may get a lot worse.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-a643f5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
In 2024 so far, a now-retracted paper with an AI-generated rat image and a lithium battery paper with an LLM-written opening sentence are the clearest examples of GenAI contaminating science.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-4c8f11-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Although using LLMs for writing and illustration has utility, particularly for nonnative speakers, the long-term consequences of LLM use on science may be severe.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-1cf6ab-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
GenAI-created garbage submissions could quickly overwhelm journals' reviewing processes and cause a precipitous drop in their reputation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-year-of-the-big-balled-rat-1b1afc-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic ripped off a lot of writers' work, as shown by its $1.5B settlement.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-whole-thing-was-scam-c984d6-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-whole-thing-was-scam-808a57
The openai review does not accept e1 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
The whole episode, including Altman's public support for Amodei while he was secretly working on his own deal, was all theatre, and Dario never had a chance.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-whole-thing-was-scam-73b1fb-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The US government permanently banished Anthropic and absurdly and punitively declared it a supply chain risk, going beyond merely rejecting its terms.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-whole-thing-was-scam-5eebea-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-whole-thing-was-scam-5ea829
The final-text check failed: citations_correct.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-whole-thing-was-scam-5a6159
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
The government acted in favor of someone else (the OpenAI/Altman side) who took pretty similar terms to Anthropic's and happened to have given more campaign contributions.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-whole-thing-was-scam-4560d6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic should be given a chance at exactly the same government terms as OpenAI; anything else would look corrupt.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-whole-thing-was-scam-37bb02-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-white-houses-shambolic-ai-policy-fe8f17
hedged language is not scored, even on a plain reading
2026-10-29
Assessment publishedBy this siteGary Marcus
What was just identified about Fable should not have been addressed in such a hasty, arbitrary way, with such little attention to downstream consequences.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-fe8f17-2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-white-houses-shambolic-ai-policy-f638b0
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-white-houses-shambolic-ai-policy-f562db
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
The Commerce Department's June 2026 export control order has basically led to the possibly temporary shutdown of Anthropic's Fable and Mythos.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-ec7232-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The Commerce Department export control order may spark a crisis in the US AI industry.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-ec7232-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
US AI policy is a mess.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-bf8faa-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Leaders should provide strong but measured AI oversight that is bipartisan, technically well-informed, calm, and proactive rather than reactive.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-a8c28f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
US states must step in on AI oversight if they are to protect their citizens.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-a1e52c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic brought the export control order on itself to a considerable degree, by overselling doom and its own products and by literally and repeatedly calling for export controls itself.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-91f520-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
The Trump administration should take a carefully considered, nuanced approach to AI policy rather than shooting first.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-884a52-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-white-houses-shambolic-ai-policy-86e558
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-white-houses-shambolic-ai-policy-8510a4
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-white-houses-shambolic-ai-policy-792fff
The anthropic review reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Government must comprehensively review, across all known risks, new AI models likely to be massively deployed before they are released.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-792fff-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The US government should take a stronger hand in forcing AI companies to address chatbot harms.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-301167-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-white-houses-shambolic-ai-policy-1ce7f1
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
The Commerce Department's export control order was both bad policy and bad politics.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-18b7b6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The harms alleged in Florida's lawsuit against OpenAI were all foreseeable.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-white-houses-shambolic-ai-policy-0c084b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The AI surveillance tools the US government is apparently using are likely not reliable.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-us-and-china-want-different-things-cb6884-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The AI tools the US is apparently using to surveil its citizens are likely full of insidious bias, as so many previous AI systems have shown themselves to be.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-us-and-china-want-different-things-cb6884-3-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-us-and-china-want-different-things-cb6884-2
hedged language is not scored, even on a plain reading
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-us-and-china-want-different-things-c2ccf4
The final-text check failed: proposition_faithful.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-us-and-china-want-different-things-5f01d1
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-uk-ai-summit-41b96d
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-uk-ai-summit-3d3953
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus suggests, tentatively ('if I am not mistaken, I guess'), that Suleyman's definition would make cell phones and Microsoft Word 'superintelligent'.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-two-wildest-stories-today-in-a862b2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus's only guess is that OpenAI's $250 million purchase of TBPN is a desperation move: an effort to control the narrative and to distract from the billion or so dollars it is losing each month this year and from rumors that its stock is losing favor on the secondary market.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-two-wildest-stories-today-in-8f7123-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-two-wildest-stories-today-in-8f7123-2
The two reviews differ: anthropic historical_observation/not_applicable/unresolved, openai historical_observation/not_applicable/contradicted.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-two-wildest-stories-today-in-70bcfe
The final-text check failed: rationale_supported_by_evidence, citations_correct; overclaims: His November 7, 2025 essay described superintelligence as going far beyond human-level performance, providing context for the narrower April framing [e2]..
2026-10-29
Assessment publishedBy this siteGary Marcus
Actual AGI is nowhere in sight as of April 2026.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-two-wildest-stories-today-in-0e1277-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-two-wildest-stories-today-in-0e1277-2
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
People should not be fooled by Google's PR about Gemini.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-truth-about-googles-amazing-new-df2f04-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's Rubik's cube robot demo relied on a sensor-instrumented cube and had little real-world application.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-truth-about-googles-amazing-new-a77836-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-truth-about-googles-amazing-new-8a82a4
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-truth-about-googles-amazing-new-67ab22
The anthropic review does not accept e2 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
Gemini will be shown to still have dicey, unpredictable reasoning and a very limited understanding of the world.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-truth-about-googles-amazing-new-67ab22-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Tech demo videos almost never tell the true story.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-truth-about-googles-amazing-new-225142-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Filip Piekniewski's test of Gemini, which tried something not in the vast training corpus that requires real reasoning and checked how it ran, was the right test.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-truth-about-googles-amazing-new-1bc3bb-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI quietly dropped its Rubik's cube robotics project a year or two after the demo.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-truth-about-googles-amazing-new-055a8a-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The potential for AI to automatically generate misinformation at scale is increasing.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-threat-of-automated-misinformation-ec878c-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-threat-of-automated-misinformation-854758
hedged language is not scored, even on a plain reading
2026-10-29
Assessment publishedBy this siteGary Marcus
In Oakley's reported example, Bing makes the false narrative it generates look as if it is referenced.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-threat-of-automated-misinformation-6477c1-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus does not really think GPT-4 has much to do with AGI, as he has said before.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sparks-of-agi-or-the-end-of-science-debba1-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The public must demand transparency from Microsoft and OpenAI, and if it is not provided, consider shutting their projects down.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sparks-of-agi-or-the-end-of-science-c0e4dd-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sparks-of-agi-or-the-end-of-science-b5bd34
The openai review abstained.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sparks-of-agi-or-the-end-of-science-868b0c
e4 is dated 2023-03, which is not inside the period described (2023-03-01 to 2023-03-24); it can only be context; e5 is dated 2023-03, which is not inside the period described (2023-03-01 to 2023-03-24); it can only be context
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sparks-of-agi-or-the-end-of-science-81fda6
The final-text check failed: citations_correct.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sparks-of-agi-or-the-end-of-science-76c58f
The two reviews differ: anthropic historical_observation/not_applicable/mixed, openai historical_observation/not_applicable/contradicted.
2026-10-29
Assessment publishedBy this siteGary Marcus
The Sparks of AGI paper's claim that GPT-4 attains a form of general intelligence cannot be seriously tested because scientists have no access to its training data.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sparks-of-agi-or-the-end-of-science-3a62a4-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Microsoft's claim that GPT-4 is an early (yet still incomplete) version of an AGI system would never survive serious scientific peer review.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sparks-of-agi-or-the-end-of-science-342473-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Microsoft and OpenAI's opaque deployment of risky AI systems should not be allowed to continue.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sparks-of-agi-or-the-end-of-science-230400-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
GPT-4 has not solved hallucinations, reliability, or planning on complex tasks.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sparks-of-agi-or-the-end-of-science-16c15a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-seven-most-shambolic-things-that-ddbaf8
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-seven-most-shambolic-things-that-dbea3a
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-seven-most-shambolic-things-that-94e034
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
MIT Technology Review reported new jailbreaking adventures, which Marcus calls disturbing.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-seven-most-shambolic-things-that-91268f-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-seven-most-shambolic-things-that-2fd8db
e1 measures something else (metric Financial distress and forced disposal of public-equity holdings is not Financial distress and forced disposal of a heavily leveraged public-equity portfolio); it can only be context unless it is derived with stated inputs; e2 measures something else (metric Public-equity portfolio leverage before forced disposal is not Financial distress and forced disposal of a heavily leveraged public-equity portfolio); it can only be context unless it is derived with stated inputs
2026-10-29
Assessment publishedBy this siteGary Marcus
Competitive pressure drove OpenAI to cut its prices by 80%.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-seven-most-shambolic-things-that-0ab8d6-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-secret-to-agi-in-4-pages-6c6b92
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
Zuckerberg keeps making the same "mistakes" over and over.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-secret-behind-metas-muse-e8cb99-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
No more than 30% of Facebook M's requests were handled by AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-secret-behind-metas-muse-e606fc-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
It is uncertain whether Meta Muse will work where Facebook M failed.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-secret-behind-metas-muse-cf4d56-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Meta Muse is basically trying to do exactly what Facebook M tried to do, such as restaurant reservations and theatre tickets.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-secret-behind-metas-muse-7a5dd9-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-secret-behind-metas-muse-4b48c6
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-secret-behind-metas-muse-21a178
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-secret-behind-metas-muse-0ca3ae
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
Before the end of the decade, another $100B AI investment (Microsoft/OpenAI's Stargate) may join driverless cars as a poor investment.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-ff73cc-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
On most problems, customers probably would not pay the premium required by a $100B LLM.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-df6ff3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
On many tasks, a $100B LLM's results might be only slightly better than those of a $10B or $1B model, while its operating costs would presumably be much larger.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-dd78f2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Even a $100B LLM will likely still have flawed reasoning and planning in unexpected scenarios.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-cabe29-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs, even when scaled, may well remain often troubled by unexpected cases, as driverless cars were by the long tail of outliers.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-b2c8e1-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The generality of Waymo's approach (to conditions such as bad weather, poorly mapped roads and different driving patterns) is very much still in question, and Marcus is not sure anyone else is still seriously in the full autonomy game.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-ac5323-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
As of March 2024, investment in driverless cars has not produced much payoff.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-85774e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
By any reasonable account, the over $100 billion invested in driverless cars has to be the worst investment in the history of AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-57fc6d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
A $100B LLM is still an LLM.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-53af09-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Most of those who invested in driverless cars lost money, unless they invested early and cashed out before patience began to run out.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-2eb154-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If the Stargate premises are wrong, the project will delay AGI by siphoning talent away from other approaches.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-1d1af7-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-second-worst-100b-investment-1d1af7-2
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
A $100B LLM investment is unlikely to produce AGI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-00ad4d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
It is doubtful that customers would pay literally anything for AGI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-second-worst-100b-investment-00ad4d-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Whether generative AI proves to be AGI is very much still up for grabs, and on any reasonable account there is a long way to go.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-f038db-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Training GPT-5 may cost an order of magnitude more than GPT-4, and GPT-6 another order of magnitude more.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-deaa61-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Meta will likely release a free GPT-4-level model within months of May 2024.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-de61c2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
It is not a sure thing that AI systems will become dramatically more capable every year for many years.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-c3bb47-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Google DeepMind, Anthropic, Meta or someone else could conceivably beat OpenAI to a GPT-5-level model.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-bbaa6a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sam-altman-playbook-b071a1
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman is selling hope rather than a new product.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-aaeee0-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sam-altman-playbook-56b19c
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Almost all AI companies follow the same overpromising, speculative playbook as Altman.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-46b088-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sam-altman-playbook-2e999d
The openai screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
If OpenAI lost a copyright lawsuit setting precedent against training on copyrighted materials, its entire enterprise could be in jeopardy.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-2c4be2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-sam-altman-playbook-1a8ea4
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
In seeking vast resources for a deeply flawed approach, Altman is diverting resources from alternative AI approaches that might be more trustworthy, reliable and interpretable, is potentially incurring considerable environmental costs, and is distracting from better ways of using capital to help humanity more generally.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-sam-altman-playbook-0eec79-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-roi-on-genai-might-not-be-so-d2b2e8
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI is among the most hyped things the author has ever seen, with the possible exception of LK-99.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-roi-on-genai-might-not-be-so-2235c2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-rise-and-fall-of-microsofts-new-8b2fae
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
As of late August 2023, it is not obvious that Marcus's bet that ChatGPT will turn out to be a dud is wrong, though it may be too soon to declare victory.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rise-and-fall-of-chatgpt-a05028-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-rise-and-fall-of-chatgpt-92c972
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT has suffered a stunning reversal of fortune, though the story is not over.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rise-and-fall-of-chatgpt-23b3c1-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The potential New York Times lawsuit could be particularly devastating for OpenAI because, unlike classical databases, LLMs cannot have individual pieces of data removed at will; as far as Marcus understands it, removing any content means retraining the entire model at great expense.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rise-and-fall-of-chatgpt-072f6c-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-revenge-of-claude-mythos-b12fa3
The final-text check failed: rationale_supported_by_evidence, citations_correct; overclaims: Specified harmful requests were routed to Opus 4.8, while Mythos 5 without those limits remained restricted to vetted customers [e1]..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-revenge-of-claude-mythos-a2d516
The two reviews differ: anthropic historical_observation/not_applicable/mixed, openai historical_observation/not_applicable/unresolved.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-revenge-of-claude-mythos-7b735d
The two reviews differ: anthropic historical_observation/not_applicable/supported, openai historical_observation/not_applicable/unresolved.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic once again played the media like a fiddle over Mythos, and so many people ate it all up.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-revenge-of-claude-mythos-65c4a7-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-revenge-of-claude-mythos-33e1f9
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-revenge-of-claude-mythos-33e1f9-2
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI might turn out to be the WeWork of AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-ece0f1-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Apple was apparently disappointed with its 2024 pilot with OpenAI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-eafe83-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Given OpenAI's setbacks, it is time to reconsider whether Sam Altman should lead OpenAI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-e46c4b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Sam Altman's act is starting to wear thin.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-d72777-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-rapid-rise-and-slow-decline-of-d521cc
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
DeepSeek's price war forced OpenAI to cut prices dramatically.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-b8a9d5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
GPT-5 was both late and short of expectations.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-b231c9-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman's only technical vision was scaling, and scaling was never going to be enough.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-982825-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
There is no clear explanation of how OpenAI will earn enough profit to justify its valuation and commitments.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-90bb1c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs have continued to have a truthfulness problem.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-7121a9-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-rapid-rise-and-slow-decline-of-693d07
The final-text check failed: proposition_faithful.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLM price wars occurred and user costs per token are roughly 100 times lower than two years earlier.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-4cb944-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic and Google have caught up with OpenAI, and xAI and Chinese companies are not far behind.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-48e740-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman's public credibility has steadily declined over the last couple of years.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-46a175-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic is taking many corporate customers away from OpenAI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-4454d9-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-rapid-rise-and-slow-decline-of-3da027
The final-text check failed: citations_correct.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI is mired in a lot of litigation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-3da027-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Many, many people were disappointed with GPT-5, which was arguably among the most overhyped products in history.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-3c125b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI has never made a profit.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-2bc3cf-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs are commodities, as DeepSeek demonstrated.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-17c5a6-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI lost the Apple Siri generative AI deal to Google.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-rapid-rise-and-slow-decline-of-0e3e11-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-rapid-rise-and-slow-decline-of-0af5f0
The final-text check failed: citations_correct.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-fc1bf5
The two reviews differ: anthropic historical_observation/not_applicable/supported, openai historical_observation/not_applicable/unresolved.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-f854ee
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
Nvidia could soon take a serious hit because LLMs may be trainable far more efficiently.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-e92456-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
In late December, DeepSeek, apparently first built for quantitative trading, produced a nearly state-of-the-art model that needed roughly 1/50th of the training costs of previous models, putting it in the big leagues with American companies such as OpenAI, Google and Anthropic in both performance and innovation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-e3e128-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-d76705
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
US export controls may have kneecapped Nvidia.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-cdc1fa-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-ccdc53
e2 is dated 2025-01, which is not inside the period described (2024-12-01 to 2025-01-26); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs are too opaque, unwieldy and hard to debug and verify, so betting the future on them is foolish.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-cbeeea-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-bc6d6d
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's revenue (on the order of $5B) has been modest compared with much greater outlays, and no actual killer app, relative to the costs, has emerged.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-b65bc9-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-b58d0f
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-b4f443
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-b453dc
e2 is dated 2025-01, which is not inside the period described (2025-01-01 to 2025-01-26); it can only be context
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-a8b1fb
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Building $500B worth of power and data centers to serve enormous collections of top Nvidia chips isn't looking so sensible either.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-a7d1e8-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-a388d9
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
If the LLM race stays unregulated in the US and LLMs remain detached from reality, rushed development cycles and intense global competition might well worsen the risks of misinformation, biased outputs, privacy breaches and misuse by malicious actors.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-a11e27-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-98d5c3
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-9692db
The anthropic review does not accept e1 as supports.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-951c0a
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-931d40
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
If all this is true, the CHIPS Act, which was designed to slow China in the AI race, may turn out to be one of the worst backfires in history.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-8544c0-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-7f6392
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-79a527
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-7917c2
The final-text check failed: citations_correct; overclaims: e4 explanation: "METR found substantially weaker performance than leading US models on some autonomous and AI research tasks." METR says R1 was slightly better than o1-preview and roughly at the September 2024 frontier on autonomy tasks. On RE-Bench it was only 'worse' than frontier models other than GPT-4o. 'Substantially' and the autonomy part are not supported..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-603fda
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
Because LLMs got much cheaper and need less special hardware, other players may catch up.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-5c116a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-58e915
The two reviews differ: anthropic historical_observation/not_applicable/mixed, openai historical_observation/not_applicable/unresolved.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-the-race-for-ai-supremacy-is-over-44f477
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
The country that best fosters true innovation, not the one with the most chips, will win the race to AGI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-4149b5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI may well become the WeWork of AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-32469f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus agrees with Musk that the US should not be so reliant on Altman and should not have given him such an eminent seal of approval based on so little.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-the-race-for-ai-supremacy-is-over-31da00-a1.