Changes, page 34 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 34 of 37: changes 6601 to 6800 of 7383, newest first. Dates are when the change was recorded.

2026-10-29
Assessment publishedBy this siteGary Marcus
Much of what SB-1047 would require is something AI companies already voluntarily committed to at the White House and in Seoul.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-open-letter-to-fei-fei-li-concerning-d1f609-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Asking for standards is not unique to AI: in many industries, such as pharmaceuticals, aviation and automobiles, it is common to ask companies to evaluate the safety of their products against set standards.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-open-letter-to-fei-fei-li-concerning-cee48a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-open-letter-to-fei-fei-li-concerning-927ff2
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus finds it reasonable for SB-1047 to focus on specific critical harms, and he does not understand what would justify an exemption from those provisions.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-open-letter-to-fei-fei-li-concerning-90cf7a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-open-letter-to-fei-fei-li-concerning-77feb2
The anthropic review does not accept e1 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
With no other serious proposal on offer, Marcus personally favours SB-1047, though he would welcome discussion of alternatives.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-open-letter-to-fei-fei-li-concerning-521f92-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-open-letter-to-fei-fei-li-concerning-3e417e
The final-text check failed: citations_correct.
2026-10-29
Assessment publishedBy this siteGary Marcus
The US should at least consider pivoting toward working with China on AI for global good instead of starting a new AI cold war.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-open-letter-to-david-sacks-8195b4-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If the US starts a new AI cold war with China, it will probably end uncomfortably for everyone, with no clear winners.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-open-letter-to-david-sacks-8195b4-2-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-epistemic-clusterfuck-in-the-making-8959c9
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Automatically amplifying the already inherent tendency towards extremism without fact-checking or reasoned analysis is a recipe for chaos.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epistemic-clusterfuck-in-the-making-80c124-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The epistemic situation may actually get worse than social media, which Marcus says might itself be an example of an epistemic clusterfuck.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epistemic-clusterfuck-in-the-making-4daef4-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs whose synthesis is confined to X tweets will be even less grounded in reality than LLMs in general.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epistemic-clusterfuck-in-the-making-176594-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
In Marcus's opinion, the investor audience at the forum was being lied to: they were being assured that everything in AI is under control and well plotted out, when in fact science is hard and the field is in its infancy.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-f885e5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Many people will not be able to afford or want humanoid robots, and there will not be six humanoid robots per car anytime soon.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-dc0df2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Quantitative vibes, meaning gut feelings expressed as numbers, are still just vibes.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-ccc5d2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Humanoid robots will not sell 200 times Roomba's roughly 50 million units in the near term.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-b29a88-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-epidemic-of-weirdly-precise-ai-b29a88-2
The final-text check failed: citations_correct.
2026-10-29
Assessment publishedBy this siteGary Marcus
Even if one believed scaling was infinite and that useful data would not run out, there is just no way to predict all this, including how efficient future models would be in energy and chips, out to 2040 with any certainty.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-a20132-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Because investors working with other people's money get paid large fees even when investments do not pan out, so many of them love plausible stories with big numbers.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-97593f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI will not be able to do anything any human can do within the next one to two years, and Marcus would bet a million dollars on it.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-8edf78-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Masayoshi Son's figure that artificial superintelligence will be 10,000 times smarter than humans is a made-up round number with no real meaning; pure hype.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-88be50-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
By OpenAI's own data, o1 is certainly not 100 times better than GPT-4.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-6dd8f9-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
People should not take Musk's or Son's AI predictions from the Riyadh forum seriously.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-6ab6b1-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI uses so much power that it keeps fossil fuels in demand, which benefits oil-rich countries whether or not GenAI ever becomes reliable.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-68ed3f-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Those making these outlandish claims get free press for them, and that drives up the whole thing, at least for a while.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-68db25-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-epidemic-of-weirdly-precise-ai-5283b4
a observation claim cannot be assessed on the forecast_outcome axis; the horizon on 2026-10-28 is open_horizon, not not_due
2026-10-29
Assessment publishedBy this siteGary Marcus
When Musk extrapolated that AI would be 10,000 times better in 4 years, he did not consider potential bottlenecks around data or compute, or inherent problems with LLMs.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-1ed26f-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Nobody can predict AI's future power, chip, and capital needs with the precision Son claimed.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epidemic-of-weirdly-precise-ai-10b750-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI, even in its current form, is about to have a huge impact on society.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epic-ai-debateand-why-everyone-d8f88e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI should pursue hybrid approaches combining neural networks with reasoning and structured representations.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epic-ai-debateand-why-everyone-b91461-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Current approaches to getting machines to reason about causality are not adequate, at least not yet.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epic-ai-debateand-why-everyone-90c73e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The remaining years of the 2020s will be wild for AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epic-ai-debateand-why-everyone-68bc3b-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Recent advances in AI-generated content may lead to an avalanche of misinformation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epic-ai-debateand-why-everyone-5dd265-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Contemporary approaches to AI may never solve reasoning, abstraction, compositionality, and factuality, which any intelligent machine should have.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epic-ai-debateand-why-everyone-50d38d-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI research should place greater emphasis on understanding human intelligence.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-epic-ai-debateand-why-everyone-4eff9e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
SpaceX will not return 2,538x its IPO investment as Amazon did, which would imply a market cap of about $4,442 trillion.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-ed4c8f-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
If AI boosts productivity, the boost may appear first in coding.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-cd9fec-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-cd9fec-2
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-a03b89
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI is in trouble.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-9c00d2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI may not survive.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-9c00d2-2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-75e72f
The anthropic review reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-6b8b55
The final-text check failed: rationale_supported_by_evidence; overclaims: The 2,538x Amazon figure comes from Marcus's post, which relies on a figure posted on X..
2026-10-29
Assessment publishedBy this siteGary Marcus
Buying the SpaceX IPO reflects poor understanding of math or history.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-2bdc18-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The costs of developing AI data centers, including OpenAI's Stargate, are rising rapidly.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-26bf80-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI is a risky bet that could pay off or could well tank the economy.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-22b948-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-16b927
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-0b096d
The anthropic review does not accept e3 as contradicts.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI may never become profitable.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-0b096d-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Anthropic has a chance of becoming profitable.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-entire-industry-is-being-propped-0837e5-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-0837e5-2
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-entire-industry-is-being-propped-00cf36
The openai review reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
To Marcus, Midjourney's output closely resembling copyrighted film stills sure looks like automated, digital plagiarism.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-artist-fights-back-and-midjourney-c80263-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-artist-fights-back-and-midjourney-b260d0
The final-text check failed: citations_correct; overclaims: e3 explanation: "The preserved December 23 report independently describes close resemblance in the examples credited to Southen." The quoted paragraph is about 'several examples on X' and is not explicitly credited to Southen in the text..
2026-10-29
Assessment publishedBy this siteGary Marcus
The evidence suggests Midjourney was trained on high-resolution copyrighted images, which it may or may not have a license for.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-artist-fights-back-and-midjourney-ab891a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-artist-fights-back-and-midjourney-35a55f
hedged language is not scored, even on a plain reading
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-artist-fights-back-and-midjourney-2644ae
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-artist-fights-back-and-midjourney-25b426
The two reviews differ: anthropic historical_observation/not_applicable/supported, openai historical_observation/not_applicable/unresolved.
2026-10-29
Assessment publishedBy this siteGary Marcus
Midjourney's output was not labelled as being directly based on copyrighted work.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-artist-fights-back-and-midjourney-050eae-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-an-artist-fights-back-and-midjourney-043f60
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
Scaling laws are merely empirical generalizations that held for a limited period while enough fresh data and compute were available.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-ai-rumor-you-wont-want-to-miss-db6cf3-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The era when simply adding data and compute produced large gains may well be over.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-ai-rumor-you-wont-want-to-miss-d707e6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If the rumor that a second major AI company is seeing diminishing returns from scaling is verified, the AI industry could see the equivalent of a bank run.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-ai-rumor-you-wont-want-to-miss-c0ab84-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
There has never been a principled argument that scaling would solve hallucinations, reasoning, or open-world edge cases, or that synthetic data would suffice indefinitely.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-ai-rumor-you-wont-want-to-miss-9ec92f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus strongly suspects that pure scaling of AI models has reached a point of diminishing returns, as of November 2024.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-ai-rumor-you-wont-want-to-miss-7e7d49-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Scaling (improving AI by adding more data and compute) was always a hypothesis.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-an-ai-rumor-you-wont-want-to-miss-1734b2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The American people should have a say in AI policy, especially AI policy of this magnitude, and Congress should deliberate.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-f400be-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Hegseth's demand would in principle allow Anthropic's software to be applied to nuclear weapons without humans in the loop.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-db7386-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Hegseth's deadline to Anthropic is a power grab meant to circumvent Congress and cut others out of the decision.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-d84d43-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Nothing Marcus has read convinces him that Secretary Hegseth has a nuanced understanding of the strengths and limits of current AI, or that Hegseth will show restraint in how he applies it.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-91efca-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus endorses the reader's view that whether AI may be used for surveillance of Americans and lethal strikes without humans in the loop must be decided publicly by Congress, not forced by the Pentagon.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-88ea22-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
We are headed exactly toward a situation where one individual, a Cabinet official, decides at gunpoint on mass surveillance and AI-fueled weapons, possibly nuclear, without humans in the loop.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-7725fe-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Mass surveillance and AI-fueled weapons, possibly nuclear, without humans in the loop are categorically not things that one individual, even one in the Cabinet, should be allowed to decide at gunpoint.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-71c2f2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-america-and-probably-the-world-stands-5b4015
The final-text check failed: conditions_shown, citations_correct; overclaims: Hegseth promoted fast military AI adoption in a high-profile public speech, which shows it was a personal, public priority..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-america-and-probably-the-world-stands-524ea5
The anthropic review does not accept e4 as contradicts.
2026-10-29
Assessment publishedBy this siteGary Marcus
Because the Block Nuclear Launch by Autonomous AI Act failed, there might be no human like Stanislav Petrov to stop a mistaken nuclear launch next time.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-3467d5-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-america-and-probably-the-world-stands-2bfba4
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Unrestricted military use of Anthropic's AI could well extend to nuclear weapons.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-america-and-probably-the-world-stands-266558-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-america-and-probably-the-world-stands-1e1f18
The two reviews differ: anthropic forecast_outcome/open_horizon/not_adjudicable, openai forecast_outcome/open_horizon/unresolved.
2026-10-29
Assessment publishedBy this siteGary Marcus
The real reason for Sam Altman's move to cut into Elon Musk's turf is to get more data and to pivot OpenAI into surveillance.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-altman-vs-musk-cage-match-37688e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's move into Musk's turf is part of an effort to pivot OpenAI into surveillance.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-altman-vs-musk-cage-match-37688e-2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-ff00da
e1 is dated 2024-07, which is not inside the period described (2024-07-01 to 2024-07-28); it can only be context
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-e5f129
the conclusive event's date differs from its evidence
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-ddb3ee
e6 is dated 2024, which is not inside the period described (… to 2024-07-28); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs should either be abandoned or used only as components within larger systems that reason and plan better.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-ce9adc-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI's reliability failings can never be solved, given its inherent nature.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-cab6c8-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI has no killer app.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-c58ed5-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus says that, for the most part, all the major tech companies are building the same thing (bigger LLMs on more data) and getting almost the same results (GPT-4-level models that barely differ and all struggle with hallucinations and boneheaded errors), with hardly any investment in anything else.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-bb6581-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-b9230d
The final-text check failed; overclaims: No decisive event contradicting the forecast had occurred, but this does not confirm 'never'..
2026-10-29
Assessment publishedBy this siteGary Marcus
The current generative AI approach has reached a plateau.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-ad371e-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus says that, as far as he can tell, hallucinations, boneheaded reasoning errors and 'discomprehensions' are an inevitable byproduct of LLMs, and that they certainly have not yet disappeared despite many promises to the contrary.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-a8a13f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI, especially ChatGPT, will soon flame out.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-a214e8-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Google DeepMind's work shows neural networks and symbolic systems can be profitably combined.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-a113d8-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
No generative AI company has a competitive moat.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-958ae5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AGI cannot be achieved without neurosymbolic AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-88edd6-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The generative AI bubble will begin to burst by July 28, 2025.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-75be04-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-6b90d8
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
The current AI investment climate is hugely biased against genuine novelty.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-5fd6c8-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
GenAI is vastly overrated, and its collapse may well lead to an AI winter of sorts.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-502de0-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-3b7617
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
The impending generative AI collapse will not lead to the absolute death of AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-3694e4-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-313fcd
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
AlphaProof and AlphaGeometry 2 rely on human coders translating Olympiad problems into formal form, a serious shortcoming.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-22368d-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-21594b
e1 is dated 2024-07, which is not inside the period described (2024-07-01 to 2024-07-28); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
People are starting to realize generative AI has plateaued, lacks a killer app, still hallucinates and errs, and has no moat.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-1d76b9-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-1c26e3
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus says any reasoning or planning in LLMs is hit or miss: viable if a situation's particulars are close enough to the training set, but highly vulnerable if they are not.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphaproof-alphageometry-chatgpt-19c5a2-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-1141c0
The two reviews differ: anthropic historical_observation/not_applicable/mixed, openai historical_observation/not_applicable/contradicted.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphaproof-alphageometry-chatgpt-0bf994
The final-text check failed: proposition_faithful; overclaims: LLMs cannot be counted on, which is a big part of why Fortune 500 companies have lost confidence in them..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphageometry2-impressive-accomplishment-fbc5b6
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
Greater modularity will be a necessary part of future AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-ce59af-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
It is disappointing that a top research team confused success on a prestigious competition with mastery of a foundational subject, and tiresome that the media overhype things as ever.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-cd2c14-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphageometry2-impressive-accomplishment-c4a43f
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
DeepMind's 'gold-medalist performance' claim seems too strong to the authors, because preconverting each problem into formal notation means the system sidesteps understanding the English problem statements, which in a sense gives it an unfair start.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-c32be7-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Combining the best general AIs with AlphaGeometry2 would still fall far short of a talented high school math student's geometric understanding and problem-solving.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-a7916f-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
AlphaGeometry2 is a specialized system, nowhere close to a single general-purpose learning architecture.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-9b8c0f-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphageometry2-impressive-accomplishment-9afe42
The anthropic screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Benchmarks are often narrow, and strong benchmark results should not be taken as full conceptual understanding.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-907018-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphageometry2-impressive-accomplishment-7ecc0d
e2 is dated 2025-02, which is not inside the period described (2025-02-01 to 2025-02-17); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
The TechCrunch headline's claim about AlphaGeometry2 was not made by DeepMind and is not true.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-7cc07a-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Of the 50 IMO geometry problems from 2000 to 2024, a composite AI system starting from the English problem statements could solve at most 60%.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-7c1011-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The authors see no reason at all to call AlphaGeometry2's solutions 'superhuman'.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-7a3f81-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The claim that AlphaGeometry2 'just figured out geometry' is false because it does not cover many basic aspects of geometry.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-507dca-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphageometry2-impressive-accomplishment-4af1b1
e2 is dated 2025-02, which is not inside the period described (2024-01-01 to 2025-02-17); it can only be context
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphageometry2-impressive-accomplishment-43163c
The final-text check failed: citations_correct.
2026-10-29
Assessment publishedBy this siteGary Marcus
AlphaGeometry2 solving 42 of 50 IMO geometry problems is, without question, a very impressive accomplishment.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-alphageometry2-impressive-accomplishment-2def63-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-alphageometry2-impressive-accomplishment-188db4
The openai review reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's reported effort to launch a social media network is presumably motivated in part by a wish to collect more private data.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-all-your-data-belong-to-us-b05376-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's only financially viable path may be to move into surveillance, and it has already taken steps in that direction such as adding Paul Nakasone to its board and investing in a webcam company.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-all-your-data-belong-to-us-884b92-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
People should speak up now against the growing AI-driven surveillance.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-all-your-data-belong-to-us-68fce2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Two reports make clear that the era of Big Brother is about to enter the United States in massive new ways, not just via OpenAI but through multiple vectors.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-all-your-data-belong-to-us-460087-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-jurassic-park-moment-d6dfe4
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
If Stack Overflow is flooded with worthless generated code, programmers will leave, its database will become untrustworthy, and the site will die.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-c8fb27-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
If Stack Overflow is flooded with worthless ChatGPT-generated code examples and dies, then, because it is one of the most central resources the world's programmers rely on, the consequences for software quality and developer productivity could be immense.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-c8abdc-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI systems like ChatGPT, Dall-E 2, and Lensa can easily be automated to generate misinformation at unprecedented scale.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-c6fe15-a3.
2026-10-29
Assessment publishedBy this siteGary Marcus
Because generative AI systems like ChatGPT, Dall-E 2, and Lensa cost almost nothing to operate, they are on a path to reducing the cost of generating disinformation to zero.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-b9a4db-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
ChatGPT is about as capable as Galactica of producing nonsense misinformation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-ace0c5-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-jurassic-park-moment-acb031
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
User accounts must be more strongly validated and anonymous bot-resistant authentication should become mandatory.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-9ad25e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Every social media company and search engine should support and extend Stack Overflow's ban, misleading automatically-generated content should not be welcome, and regularly posting it should be grounds for a user's removal.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-9567ff-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Countries may need to make misinformation legally actionable like libel when created with sufficient malice and volume.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-8d6126-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Propaganda-producing nation-states and bad actors are likely to use LLMs to attack social media and create fake websites at unprecedented volume.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-8c31a2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs lack mechanisms for verifying truth.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-811c3a-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
We need to integrate LLMs with classical AI tools such as databases, knowledge graphs, and reasoning to fight misinformation.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-811c3a-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Generative AI systems like chatGPT pose a real and imminent threat to the fabric of society.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-66039f-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Executives at Meta and OpenAI are as enthusiastic about their tools as the proprietors of Jurassic Park were about theirs.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-57e151-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-jurassic-park-moment-38e673
e2 measures something else (company MosaicML is not Custom LLM training providers (e.g., MosaicML)); it can only be context unless it is derived with stated inputs; e4 measures something else (company MosaicML is not Custom LLM training providers (e.g., MosaicML)); it can only be context unless it is derived with stated inputs
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-jurassic-park-moment-18dda3
a observation claim cannot be assessed on the causal_assessment axis
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-jurassic-park-moment-10605e
e5 is dated 2022, which is not inside the period described (2022-01-01 to 2022-12-10); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
Misinformation-capable LLM technology can no longer be contained or withdrawn.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-0aa544-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLMs make greatly escalated propaganda volume possible, and propagandists might succeed in creating a world where people cannot know what to trust.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-jurassic-park-moment-04b5a5-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The developments described (SpaceX leasing compute to Google and Anthropic, and firms buying hardware before there was demand for it) are not particularly great for Nvidia, CoreWeave, Oracle, Nebius, Micron, and a variety of other hardware and semiconductor providers.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-f92c43-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-black-friday-f3dca1
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-black-friday-c7975c
The openai review found the criteria unfair to the speaker's words.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-black-friday-bd2775
The openai screen reads the claim differently.
2026-10-29
Assessment publishedBy this siteGary Marcus
Meta may end up burning the most money on AI for the least results.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-ad8da7-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-black-friday-76d223
The final-text check failed: rationale_supported_by_evidence; overclaims: The cited evidence is dated before the June 5, 2026 Google agreement, and the terms of that lease are not among the cited items..
2026-10-29
Assessment publishedBy this siteGary Marcus
SpaceX's AI division is leasing out large numbers of GPUs in 2026 because it cannot figure out what to do with them.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-725a28-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
If an American AI company is partly owned by the US government, nobody is going to trust it.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-6f5d76-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
The market ought to keep falling on Monday, June 8, 2026, though Marcus says he does not know whether it will.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-50b5a3-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-black-friday-4364d8
The anthropic screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ais-black-friday-42dc64
The openai review does not accept e2 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
Bailing out OpenAI is crony socialism and corruption that prevents capitalism from taking its natural course.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-3b461a-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Marcus doubts that Friday, June 5, 2026 was as bad as the AI-related market downturn will get.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-2ff999-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Google might be wise not to accept Trump's offered government investment, because so much of its income is global.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-2402b9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI may be so financially desperate that it has no choice but to accept government investment.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-2402b9-2-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
The new cold war will be about AI.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ais-black-friday-1c94a0-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-fcb735
e5 is dated 2023-11, which is not inside the period described (2023-11-01 to 2023-11-25); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
AI should aim to supplement humans where they are cognitively weak and take on dull, dirty, and dangerous work, not replicate humans.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-versus-the-clumsy-human-mind-de0c1e-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-c2bc7d
The anthropic review does not accept e2 as supports.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-ad98e1
The final-text check failed: rationale_supported_by_evidence; overclaims: The two papers considered do not show whether any system has or lacks world models..
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-a2c5f9
The anthropic screen finds the claim checkable, so it must be graded.
2026-10-29
Assessment publishedBy this siteGary Marcus
AGI should be a reliable, hallucination-free system that solves human problems rather than creating new ones.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-versus-the-clumsy-human-mind-a26a37-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
We can probably make machines better if we first spend time thinking about human minds, both their strengths and their weaknesses.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-versus-the-clumsy-human-mind-8fd334-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-759c54
The final-text check failed: citations_correct.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-6f5b7c
a causal claim cannot be assessed on the forecast_outcome axis; hedged language is not scored, even on a plain reading
2026-10-29
Assessment publishedBy this siteGary Marcus
As of November 2023, neither humans nor machines are more intelligent overall; they have different abilities.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-versus-the-clumsy-human-mind-6151f4-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-47d2eb
e1 is dated 2023-11, which is not inside the period described (2023-01 to 2023-11-25); it can only be context; e2 is dated 2023, which is not inside the period described (2023-01 to 2023-11-25); it can only be context; e3 is dated 2023-11, which is not inside the period described (2023-01 to 2023-11-25); it can only be context; e4 is dated 2023-11, which is not inside the period described (2023-01 to 2023-11-25); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
Eventually, we will be able to build machines that help us reason critically, overcome human bias, and reason objectively, giving just the facts without confabulation, the way calculators do.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-versus-the-clumsy-human-mind-43b7c4-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
People should recognize, as soon as possible, that humans and silicon machines are importantly different.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-versus-the-clumsy-human-mind-35227a-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-versus-the-clumsy-human-mind-22ee05
e1 is dated 2023-11, which is not inside the period described (2023-01 to 2023-11-25); it can only be context
2026-10-29
Assessment publishedBy this siteGary Marcus
AI should not aim to replicate the human mind in every detail.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-versus-the-clumsy-human-mind-023b6e-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
General-purpose household robots like Rosie the Robot are not doing our housework, and this is in part because the physical reasoning such a robot would need cannot be done purely by running simulations.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-ffeaae-a1.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-still-lacks-common-sense-70-years-fa6dd9
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
The hope that reliable commonsense will emerge from large text and video datasets has repeatedly proven unrealistic.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-f53ad1-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
AI agents, meaning AI systems that act in the world on users' behalf, are already here to a limited degree, even though general-purpose domestic robots are a long way away.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-de2cac-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
OpenAI's o1 often makes very basic spatial mistakes because it does not understand how objects are embedded and interact in space.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-da3ee2-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Recent LLMs no longer make GPT-3's specific mistakes but still lack a robust understanding of the real external world.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-d51afb-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-still-lacks-common-sense-70-years-c9a2f8
The openai review does not accept e2 as supports.
2026-10-29
Assessment publishedBy this siteGary Marcus
General-purpose AI will begin to mature only once researchers directly tackle commonsense reasoning about entities interacting in space and persisting in time.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-b137c9-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Much of the commonsense reasoning problem the authors raised in 2015 and 2019 remains unsolved in AI as of January 2025.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-b0c0ed-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Current AI approaches have consistently fallen short on reasoning about entities and their properties.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-a13f79-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Sora, now publicly available, struggles with physical reasoning.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-7c6de3-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
Current LLM agents (Claude, Gemini, GPT-4 etc.) are not close to ready for prime time on real-world workplace tasks.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-732925-a2.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-still-lacks-common-sense-70-years-72badf
The openai screen reads the claim differently.
2026-10-29
HeldBy this siteGary Marcus
Held record marcus-ai-still-lacks-common-sense-70-years-55dca2
The openai review found the criteria unfair to the speaker's words.
2026-10-29
Assessment publishedBy this siteGary Marcus
None of the three popular approaches to commonsense reasoning (simulation, LLM emergence, video generation) has proved fully satisfactory.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-4f1e61-a1.
2026-10-29
Assessment publishedBy this siteGary Marcus
Physical simulation has proven much less effective for AI commonsense reasoning than for scientific, engineering, and entertainment applications.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-4037bf-a2.
2026-10-29
Assessment publishedBy this siteGary Marcus
LLM commonsense failure examples keep appearing because no robust solution has been found, even though individual examples can be fixed.
First assessment published after both reviews and the final-text check passed.
Assessment marcus-ai-still-lacks-common-sense-70-years-34937c-a1.