Changes, page 23 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 23 of 37: changes 4401 to 4600 of 7383, newest first. Dates are when the change was recorded.

2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-coding-agents-security-nightmare-2e4856
The openai screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
Even with manual approval, hurried developers could easily approve malicious changes suggested by coding agents.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-coding-agents-security-nightmare-270450-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-coding-agents-security-nightmare-216e97
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
An unthinking LLM, without meaning to do so, cons many of its users through a process that closely parallels a psychic's con.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-arent-very-bright-why-are-so-f9c798-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Gary Marcus thinks the whole world would be different if so many people were not so easily taken in by the parlor tricks of generative AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-arent-very-bright-why-are-so-ebf557-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Bjarnason's LLMentalist analysis is correct, and people who believe in LLMs have fooled themselves.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-arent-very-bright-why-are-so-de1b5f-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-arent-very-bright-why-are-so-ba092d
The final-text check failed: rationale_supported_by_evidence; overclaims: Only two sources were consulted, as context..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-e8167d
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
The rumor that LLMs are basically just like humans is false.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-e26eb8-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs do not sanity check their own work.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-db6a26-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-d11e59
The two reviews differ: anthropic not_applicable/not_applicable/not_adjudicable, openai causal_assessment/not_applicable/unresolved.
2026-10-29
Assessment publishedBy this siteGary Marcus
GPT-5 in thinking mode produced a list of Canadian Prime Ministers full of easily avoidable errors.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-cee75f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The AI field still has not put the time in on giving time, space, and causality a solid computational footing.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-c7d8c4-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs do not induce abstract world models they can reliably use, even when given the rules explicitly.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-b5e141-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
People should never trust LLMs and should treat them as weird function approximators.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-a1195c-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-a05902
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs do not think like humans, however well they mimic human language.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-906722-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-8d5e2c
The openai review reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-81cbbd
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-60473d
The final-text check failed: citations_correct.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-5f0c70
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT gets the British prime minister tasks wrong because it is munging words together rather than genuinely processing an elementary model of British PM history.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-3c190c-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-356cd4
The openai review reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs never induce proper world models, which causes them to be unable to play chess reliably and to make frequent stupid errors.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-14c423-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
A big company's added python code for chess legality is a special-purpose hack, the LLM still doesn't induce the rules, and it would likely fail on chess variants.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-llms-are-not-like-you-and-meand-never-14b5bb-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-are-not-like-you-and-meand-never-031167
The final-text check failed: rationale_supported_by_evidence; overclaims: The evidence search found no source on reasoning about inflation or changing prices..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-amplify-existing-security-risks-a21771
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-amplify-existing-security-risks-a21771-2
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-llms-amplify-existing-security-risks-9d3c4e
The openai screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
A large fraction of what LLMs do is mostly memorization, and Hinton was wrong on this point.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-lets-be-honest-generative-ai-isnt-ee2903-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Scaling of LLMs is no longer going all that well.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-lets-be-honest-generative-ai-isnt-d57727-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Scaling will probably not cure LLMs' problems of untrustworthiness, memorization, and limited value.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-lets-be-honest-generative-ai-isnt-d57727-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs still cannot be trusted.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-lets-be-honest-generative-ai-isnt-c5b210-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
The economy and geopolitical policy should not be oriented around generative AI, especially on unproven hopes that it will dramatically improve.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-lets-be-honest-generative-ai-isnt-55a992-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs still are not adding much quantifiable value to the world.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-lets-be-honest-generative-ai-isnt-17889c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Per the Remote Labor Index, AI can do only about 2.5% of jobs, consistent with LLMs adding little quantifiable value.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-lets-be-honest-generative-ai-isnt-0e7b76-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
To make progress on language acquisition (human or machine), researchers must first recognize how hard the problem is.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-learning-language-is-harder-than-fe9de6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Even if systems like GPT-3 did work, there would be no guarantee that they would work in a fashion similar to humans.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-learning-language-is-harder-than-fa6d18-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
No existing machine has acquired language in the sense of mapping sentences to meanings and meanings to sentences.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-learning-language-is-harder-than-bf4476-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Current machine approaches to language acquisition are not really working.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-learning-language-is-harder-than-b73ac2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Current AI systems that, roughly speaking, do a lot of cutting and pasting without grammar rules and trees aren't really all that close to understanding human language.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-learning-language-is-harder-than-b076d7-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-learning-language-is-harder-than-325d22
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT's errors were inevitable because it is trained to produce plausible next words rather than engage with meaning.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-large-language-models-like-chatgpt-ed0205-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-large-language-models-like-chatgpt-cbb2db
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT could not reliably count to four or do one-digit arithmetic in simple word problems.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-large-language-models-like-chatgpt-abae3d-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
In the authors' view, it is past time to have a community-facing corpus of ChatGPT and other LLM errors, given the systems' inherent unpredictability and wide range.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-large-language-models-like-chatgpt-a64de6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT made things up and, at release, exhibited sexist and racist biases.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-large-language-models-like-chatgpt-97a7b3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT made all the same kinds of mistakes that earlier GPT-3 versions did.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-large-language-models-like-chatgpt-97132a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-large-language-models-like-chatgpt-49ee95
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT could sometimes produce outputs that were correct and acceptable in the regards listed (counting, arithmetic, event order, physical reasoning, human psychology, fabrication, bias), but not reliably.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-large-language-models-like-chatgpt-17d031-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-large-language-models-like-chatgpt-00eb05
The final-text check failed: proposition_faithful.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-kara-swisher-sam-altman-and-the-openai-de18e0
The openai review reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
The CEO of a potentially powerful AI company like OpenAI should not be allowed to behave deceptively toward his board.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-kara-swisher-sam-altman-and-the-openai-dae366-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Sam Altman may eventually end up holding direct equity in OpenAI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-kara-swisher-sam-altman-and-the-openai-d91ccc-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-kara-swisher-sam-altman-and-the-openai-ce3379
The two reviews differ: anthropic historical_observation/not_applicable/mixed, openai historical_observation/not_applicable/unresolved.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman consolidating more and more power while seeming less and less on the level should deeply concern us all, particularly as long as the US relies essentially on self-regulation and Altman aspires to build ever more powerful models with less and less independent oversight.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-kara-swisher-sam-altman-and-the-openai-bd7e4f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI not informing its board in advance of ChatGPT's release shows the company has strayed from its nonprofit mission.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-kara-swisher-sam-altman-and-the-openai-55b094-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman giving the board inaccurate information on multiple occasions about OpenAI's few formal safety processes is pretty scary, given what OpenAI aspires to build.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-kara-swisher-sam-altman-and-the-openai-3a09d2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman probably divested from the OpenAI Startup fund because of Dan Primack's reporting and would otherwise still own it.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-kara-swisher-sam-altman-and-the-openai-157c4c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If CATAI can produce the right package of tools and expertise for auditing, prerelease examination, and so on, and address questions of bias, reliability, misinformation and so forth, working with existing agencies that have been aiming to craft related standards, adoption may be swift.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-jumpstarting-ai-governance-d534ef-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Major tech company executives might get governments to adopt AI regulation that does more harm than good by keeping out smaller players and entrenching Big Tech, while doing little about risks like misinformation and cybercrime.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-jumpstarting-ai-governance-862826-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-jumpstarting-ai-governance-58f50b
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
International AI governance should include scientists and civil society at the table.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-jumpstarting-ai-governance-4500cc-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-italian-ban-on-chatgpt-5fc817
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-it-remains-very-difficult-to-find-df0989
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
Character.AI has filed for bankruptcy.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-it-remains-very-difficult-to-find-8cedde-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-it-remains-very-difficult-to-find-8cedde-2
e1 measures something else (period 2026-10-13..2026-10-13 is not 2026-10-13..2026-10-14); it can only be context unless it is derived with stated inputs; e2 measures something else (period 2026-10-13..2026-10-13 is not 2026-10-13..2026-10-14); it can only be context unless it is derived with stated inputs
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-it-remains-very-difficult-to-find-16ece3
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-vibe-coding-dying-df74e0
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-is-vibe-coding-dying-dec746
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-vibe-coding-dying-c30b57
The openai review reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Vibe coding by amateurs using LLMs to build products that previously required engineering teams will never be remotely reliable.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-vibe-coding-dying-b1629b-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
The problem with vibe coding lies, as always, in generalizing outside the training distribution.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-vibe-coding-dying-43dbf9-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-vibe-coding-dying-3e2c01
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-vibe-coding-dying-09de22
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-vibe-coding-dying-00a6cc
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
The technical foundation of LLMs has never been strong enough to support the hype.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-this-the-moment-when-the-generative-f23ba6-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-this-the-moment-when-the-generative-d785c2
The final-text check failed: rationale_supported_by_evidence; overclaims: The early-resolution rule needs Nvidia data center revenue or combined hyperscaler AI capital spending to fall more than 30% year over year for two or more quarters in a row. | The packet has nothing on the period after that quarter..
2026-10-29
Assessment publishedBy this siteGary Marcus
Once the markets really understand the point made in Marcus's cartoon, enthusiasm for generative AI may collapse fairly quickly.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-this-the-moment-when-the-generative-c7c663-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-this-the-moment-when-the-generative-c2af0d
The openai review reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-this-the-moment-when-the-generative-8d3ad0
e3 is dated 2025-08, which is not inside the period described (2023-03 to 2025-08-20); it can only be context; e4 is dated 2025-08, which is not inside the period described (2023-03 to 2025-08-20); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
Sam Altman is not deterred by a failure; in his mind, a failure is just an excuse to ask for money.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-this-the-moment-when-the-generative-82953b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The technical and economic fundamentals behind AI startup valuations are shaky.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-this-the-moment-when-the-generative-3c547d-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-white-houses-ai-policy-coherent-fef772
The openai review reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
The only excuse Marcus says he can see for the White House's escalating efforts to leave AI all but entirely deregulated, unlike virtually any other industry, is some sort of race with China.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-white-houses-ai-policy-coherent-fca3e2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-white-houses-ai-policy-coherent-ee784e-2
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Both federal preemption of state AI laws and H200 chip sales to China face strong bipartisan resistance.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-white-houses-ai-policy-coherent-ab2e2d-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-white-houses-ai-policy-coherent-8c42e8
The anthropic review does not accept e6 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
The expected Executive Order banning states from regulating AI may well be illegal.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-white-houses-ai-policy-coherent-8ae989-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-white-houses-ai-policy-coherent-3a1410
The final-text check failed: citations_correct; overclaims: e1 public excerpt "No federal legislation" cuts off "establishing broad regulatory authorities for the development or use of AI" and reads as if no federal AI law exists at all. The same CRS summary says targeted AI provisions have been enacted. | e2 explanation "It also identifies uncertainty and barriers to remedies": the retrieved NTIA text supports uncertainty, but nothing in it mentions barriers to remedies..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-white-houses-ai-policy-coherent-0f84e3
The anthropic screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
With his AI policies, Trump risks fracturing MAGA and alienating the public.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-white-houses-ai-policy-coherent-03d335-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Claude's constitution is basically a set of built-in guardrails, and OpenAI's models have guardrails too.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-f4d323-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-us-military-actually-afraid-ea3f1c
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs mimic human speech, so their statements should not be treated as accurate descriptions of their internal states.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-e6d41c-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-us-military-actually-afraid-d4c99b
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
What LLMs say should not be taken as true just because they say it; treating their statements as truth misunderstands them.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-c9f3fa-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
There is no more reason to think Claude is sentient than to think Google's LaMDA was.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-ac1600-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs should never be used without guardrails, so anyone worried about guardrails should not use LLMs at all.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-700854-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs hallucinate.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-6e3132-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-us-military-actually-afraid-693624
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-us-military-actually-afraid-66fe71
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
People should not put specific probabilities on AI model consciousness before defining what consciousness means.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-366b36-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
No hallucinating LLM should be used to choose military targets autonomously.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-30dd9a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-us-military-actually-afraid-270422
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Claude is a next-word predictor and not an authority on its own internal states, so it is silly to take its estimates of its own consciousness seriously.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-0dc0e9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anyone who actually believed that Claude was sentient and that sentience was de facto a supply chain risk would have to think that essentially every LLM is a supply chain risk, in which case the right choice would be to stop using LLMs altogether rather than selectively punish Anthropic.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-us-military-actually-afraid-04994c-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-great-ai-meltdown-imminent-ef92d5
hedged language is not scored, even on a plain reading
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-great-ai-meltdown-imminent-c152b7
e2 is decisive for a financial claim but states no metric scope
2026-10-29
Assessment publishedBy this siteGary Marcus
If the Nvidia–OpenAI deal collapses, the anticipated multitrillion-dollar AI expansion might well not happen at all.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-great-ai-meltdown-imminent-21b05e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If the Nvidia–OpenAI deal collapses, it could possibly bring down OpenAI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-the-great-ai-meltdown-imminent-21b05e-2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-great-ai-meltdown-imminent-215aaf
The openai review reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-the-great-ai-meltdown-imminent-197875
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-openai-more-like-wework-or-theranos-fdd3e7
a observation claim cannot be assessed on the causal_assessment axis
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's products have always come with lots of hype, and none of the robotics promised in its Rubik's cube campaign ever shipped.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-ede670-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
By all accounts the difference between GPT-3 and GPT-4 was massive, while by all accounts the difference between GPT-4 and o1 is modest.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-d8b385-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's products have not by a longshot lived up to everyone's expectations, but they do actually ship and customers can try them, so it is hard to call OpenAI smoke and mirrors.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-d7a151-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The AI industry should take a hard look at its own hype, as biomedicine had to after Theranos.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-c9f2e3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI will find it hard to live up to its $150B valuation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-c54e29-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If things with OpenAI do fall apart, society as well as investors stands to lose, and immense resources may end up being wasted in vain because of hype.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-c24753-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-openai-more-like-wework-or-theranos-be0992
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-openai-more-like-wework-or-theranos-b0c761
e8 is decisive for a financial claim but states no metric scope; e9 is decisive for a financial claim but states no metric scope
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's success is far from assured.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-9577ba-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-openai-more-like-wework-or-theranos-6dec7a
The openai screen doubts the attribution.
2026-10-29
Assessment publishedBy this siteGary Marcus
Like WeWork, OpenAI has a questionable business plan and its valuation could deflate immensely if confidence wavers.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-54c81d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
As of October 2024, neither GPT-5 nor Sora has been released.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-317b7d-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-openai-more-like-wework-or-theranos-1b1d05
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
If bigger LLMs are not the answer, the field may have wasted half a decade.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-186115-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If OpenAI loses Microsoft's unwavering support, it could end up in deep trouble.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-openai-more-like-wework-or-theranos-15c5f3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus's best guess is that Altman does not know how to turn OpenAI into the juggernaut others believe it is.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-microsoft-about-to-get-the-deal-f3324a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-microsoft-about-to-get-the-deal-eb2c3d
The openai review reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-microsoft-about-to-get-the-deal-ea3b51
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
Maybe Altman is getting out of OpenAI while the getting is good.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-microsoft-about-to-get-the-deal-d2e784-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-microsoft-about-to-get-the-deal-c632ad
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-microsoft-about-to-get-the-deal-7cc6a9
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
Giving Microsoft 75% of OpenAI's profits until it recoups its investment is not something a rocketship startup would typically agree to.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-microsoft-about-to-get-the-deal-6f4b31-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The less rosy picture—that ChatGPT is very limited and not close to AGI—is closer to the truth than the view that OpenAI is near AGI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-microsoft-about-to-get-the-deal-60aa3f-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-microsoft-about-to-get-the-deal-576a9d
e2 is decisive for a financial claim but states no metric scope
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI preparing to take a lot of money off the table doesn't quite add up if Altman genuinely thinks OpenAI is close to AGI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-microsoft-about-to-get-the-deal-573479-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-microsoft-about-to-get-the-deal-0c09b1
The final-text check failed: citations_correct.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-it-time-to-hit-the-pause-button-f71829
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
It seems foolhardy for the government to play catchup yet again as we move towards technologies that may exceed human cognitive capabilities.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-c7e059-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Given that new AI systems have demonstrated the ability to manipulate humans, tech companies could be subjected to oversight similar to that for pharmaceutical trials and research ethics boards.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-bf5778-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
More transparency about how widespread AI product release decisions are made is probably needed.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-bcbfab-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
It is time for government to consider frameworks that allow AI research under rules providing ethical standards and safety, while pausing widespread public dissemination of potentially risky new AI technologies, with severe penalties for misuse, until we can be assured of the safety of these technologies.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-b470e4-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
At present, anybody can release AI at whatever scale they like, with virtually no oversight, literally overnight.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-93548b-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-it-time-to-hit-the-pause-button-4288be
The anthropic review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
There is a high risk that chatbots will be used to create misinformation at an unprecedented scale.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-378022-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Most governments lack the institutional nimbleness and knowledge to keep pace with AI development.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-357d0b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Events of recent weeks certainly suggest that the big tech companies have shown they do not yet have the AI situation under control.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-26e7e7-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
An outright ban on AI research could prevent humanity from developing technology that could revolutionize science, medicine and technology.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-it-time-to-hit-the-pause-button-2432be-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-chatgpt-really-a-code-red-for-ea00b4
e10 is dated 2022-12, which is not inside the period described (2022-12-01 to 2022-12-30); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT's mindless answers ('discomprehension') show it has no clue what it is talking about.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-e81909-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs are not databases; they combine bits of text that don't always belong together.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-e753d6-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT's knowledge is out of date; it doesn't know who owns Twitter.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-e39859-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs are very hard to update, usually needing full retraining that can take weeks or months.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-c9a962-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-chatgpt-really-a-code-red-for-b31faa
e1 is dated 2022-12, which is not inside the period described (2022-12-01 to 2022-12-30); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
The way truth and falsity are thoroughly and authoritatively mixed together in LLM-driven search answers like the ones he examined (such as the Neeva bio) is insidious, and he personally is not ready for 'post-truth information overlords'.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-a91eb3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Users should be cautious when relying on LLM-driven search.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-975c0f-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-chatgpt-really-a-code-red-for-93dc82
e1 is dated 2022-12, which is not inside the period described (2022-11-01 to 2022-12-30); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
YouChat, though tailored for search, produced fluent but hallucinated answers to his friend's queries.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-85eab8-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-chatgpt-really-a-code-red-for-80d635
The final-text check failed: proposition_faithful.
2026-10-29
Assessment publishedBy this siteGary Marcus
Driverless cars have taken years longer than promised, because edge cases have so far kept them from becoming widely available.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-7cd365-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Developing LLM-based search will probably be a difficult, drawn-out process, as with driverless cars.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-717af0-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Until the pieces of LLM-based search fit together in a way that can be trusted, he personally prefers to pass on it ('Thank U, Next').
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-63b82e-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-chatgpt-really-a-code-red-for-5f9b0b
The anthropic screen doubts the attribution.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-chatgpt-really-a-code-red-for-5c1f61
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
For search engines, hallucination is a further problem for LLM-based systems and arguably an even bigger concern than the shallowness, or discomprehension, he described just before.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-4eb5a3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
When LLMs paraphrase and combine text, they often lose track of which pieces truly belong together, producing errors.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-257f27-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT's guardrails don't really understand what they guard against and appear to rely mostly on keywords, which makes ChatGPT prone to producing nonsense.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-chatgpt-really-a-code-red-for-24d58c-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The public will very probably not get straight answers about AI's role in military incidents.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-f8f777-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The race to shove AI into everything is grossly premature because the technology fundamentally lacks reliability.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-ef475d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Current AI algorithms are probably not precise.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-e3c31d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Whether or not current AI algorithms are precise (probably they aren't), and whether or not humans are involved in the specific selection of targets, those who use military algorithms bear responsibility for the outcomes they produce.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-e3c31d-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
We should absolutely want any AI used for war to be as precise and reliable as possible, minimizing casualties, and we should never forget that those who use it are responsible for decisions about how many casualties are acceptable.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-d8c344-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-ai-already-killing-people-by-accident-ca4247
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI still has problems with common sense.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-c3d575-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-ai-already-killing-people-by-accident-bb1267
The anthropic review does not accept e4 as contradicts.
2026-10-29
Assessment publishedBy this siteGary Marcus
The public may get little detail about AI's role in future military incidents.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-babe01-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI-triggered military mistakes may lead to escalation and, in the worst case, could lead to nuclear war.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-b7c57e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Because of its inherent tendencies, generative AI's military performance may vary by task and may be worse in unfamiliar situations.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-af5ee9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Altman signed a military contract full of holes despite talking about red lines, and seems not to care.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-aabc17-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Military use of AI should be evaluated granularly, task by task.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-8cdd32-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Without empirical military studies on collateral damage, we won't know whether AI is helping or hurting in targeting.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-81da61-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-ai-already-killing-people-by-accident-653669
The final-text check failed; overclaims: Shows that attributing even the original incident to AI was disputed and unresolved for months. | This makes later attributions hard to establish..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-ai-already-killing-people-by-accident-49fc9b
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-ai-already-killing-people-by-accident-309b11
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
In Marcus's own view, the biggest problem with the algorithms targeting Gaza was not necessarily the algorithms themselves, about which not much may be public, but the decision to tolerate a large number of civilian casualties as part of the targeting.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-2ab46e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Those who use military AI should be obliged to understand its limitations and inaccuracies.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-ai-already-killing-people-by-accident-195adb-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI labs should devote more resources to building specialized tools for science, medicine, technology and education.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-f51311-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
To Marcus, definitions of AGI based on economic criteria such as profits or percentage of jobs displaced seem fundamentally misguided.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-ad5597-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus thinks LLMs are not the right path to AGI, either morally or technically.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-9a1b6c-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-is-agi-the-right-goal-for-ai-953b20
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
The AGI definition framework should be treated as evolving, with better benchmarks replacing existing ones, not as a static checklist.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-8a075a-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Hallucinations and errors may be one reason generative AI has not produced the predicted surge in profits and productivity.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-7e5ee7-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Right now, it feels as if Big Tech is throwing general-purpose AI at the wall like spaghetti and hoping nothing truly terrible sticks.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-7429da-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
At least for now, AGI is maybe not the right near-term goal for AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-5edb7e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
To harness AI's strengths, the tech industry should shift focus from general-purpose tools to narrow, specialized AI tools for particular problems.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-is-agi-the-right-goal-for-ai-010822-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
What has been released cannot be unreleased, and AI developers should take that as a lesson.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-irony-too-funny-for-words-9046cc-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-irony-too-funny-for-words-85aaf4
The openai screen reads the claim differently.