Changes, page 59 of 60

Everything we published, changed, or held back, newest first.

All changes (11,844)

Changes 11,601 to 11,800 of 11,844, newest first. Page 59 of 60.

2026-10-29
PublishedBy usGary Marcus
Scaling laws aren't really laws anymore, if they ever were, and the meaning of the word 'scaling' has been greatly devalued.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
In recent weeks, much of the AI field has quietly acknowledged that recent unreleased large models are not as powerful as scaling laws predicted.
A piece of evidence came from the wrong dates.
2026-10-29
PublishedBy usGary Marcus
OpenClaw devotees should beware the robustness of AI-generated code even more than other coders, by what Marcus puts as '10x, if not 100x'.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
OpenAI President Greg Brockman sorta kinda acknowledged Marcus's point that code which compiles and passes the tests it was given is not the same as correct, secure, maintainable, well-architected software, and Marcus calls this a rare note of realism from OpenAI.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
Next-word prediction gets us a surprisingly long way in writing code, but less far in making sure that code is robust.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Coders, especially inexperienced vibe coders, should be wary of the robustness of AI-generated code.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
If a billion-dollar AI system cannot be used to solve a problem (Tower of Hanoi) that Herb Simon solved with AI in 1957 and that first-semester AI students solve routinely, then the chances that models like Claude or o3 will reach AGI seem truly remote.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
LLMs may succeed on easy test sets and mislead users into thinking they have a generalizable solution when they do not.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Models like o3 hallucinate much more than attentive humans and struggle with reliable diagrams.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Marcus sarcastically suggests that getting to 'alignment' or 'safety' without reliability is a long shot.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
LLMs writing Python code to supplement their weaknesses is not reliable.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Marcus is not interested in an 'AGI' that can't do arithmetic, and he certainly would not want to entrust global infrastructure or the future of humanity to such a system.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
AGI should not be about perfectly replicating a human but should be about combining the best of both worlds: human adaptiveness with computational brute force and reliability.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Anybody who thinks LLMs are a direct route to the sort of AGI that could fundamentally transform society for the good is kidding themselves.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
LLMs cannot reliably solve the Tower of Hanoi.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Apple's new paper is pretty devastating to LLMs and is a powerful followup to a paper from many of the same authors last year.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Other forms of deep learning, especially those that work well with symbols, may eventually thrive.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
We are not going to 'extract the light cone' of the earth or 'solve physics' (Altman claims whose meaning Marcus questions) with systems that can't play Tower of Hanoi on a tower of 8 discs.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
LLMs can't play chess, fold proteins, or run databases as well as conventional or special-purpose systems.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Neural networks of various kinds can generalize within the training distribution of data they are exposed to, but their generalizations tend to break down outside that distribution.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
The LLM approach has limits that are becoming clearer every day.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Claude gets under 80% accuracy on 7-disc Tower of Hanoi and essentially cannot solve 8 discs.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
A knockout blow for LLMs?
One of the AIs read the claim differently from our summary.
2026-10-29
PublishedBy usGary Marcus
o3 or Claude cannot simply be dropped into complex problems and expected to work reliably.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
For at least the next decade LLMs will continue to have uses, especially for coding, brainstorming and writing.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
LLMs are no substitute for good well-specified conventional algorithms.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
High school students will love Galactica and will use it to fool and intimidate some of their teachers.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
A Few Words About Bullshit
One of the AIs read the claim differently from our summary.
2026-10-29
Held backBy usGary Marcus
A Few Words About Bullshit
One AI wasn’t sure who said it.
2026-10-29
PublishedBy usGary Marcus
The Galactica examples Andrew Sundstrom shared are pitch-perfect but utterly bogus imitations of science and math, presented as the real thing.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
A Few Words About Bullshit
The final check failed: a cited source didn’t say what the write-up said.
2026-10-29
PublishedBy usGary Marcus
Like Galactica, every other large language model the author has seen prevaricates (fabricates content) a lot.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
A few dark words about chatbots and death
One of the AIs read the claim differently from our summary.
2026-10-29
Held backBy usGary Marcus
OpenAI faces at least eight wrongful death lawsuits related to ChatGPT users.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
By now there have been several reports of chatbot-associated deaths, both suicides and murders, enough that Wikipedia has started keeping track of them.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Chatbots apparently inducing delusions in humans is a growing problem.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Society should especially worry about chatbots influencing children, as teen deaths have shown.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
The harms Marcus warned about years ago (enshittification of the web, mass misinformation, more cyberscams) are now all happening.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
Marcus is increasingly convinced that, although generative AI has its uses, on balance it is not a step forward for humanity.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
Marcus's December 2022 prediction that the first chatbot-related death would occur in 2023 proved correct.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
The problem of chatbot-associated deaths is getting worse.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
By April 2023 a chatbot-related death had occurred, though causality was hard to prove.
The final check failed: the write-up said more than the evidence showed.
2026-10-29
PublishedBy usGary Marcus
Because everybody is following the same playbook of ever more massive data and servers fed into ever-larger large language models, nobody has a real technical advantage in the generative AI race.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
For all intents and purposes, the US-China generative AI race is basically a tie, with both countries serving many of their own customers using largely similar products.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Trump has come to believe that generative AI is a winner-take-all race, probably because of self-interested whisperings from Silicon Valley investors who often lean on FOMO about China to manipulate the government.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
The core misconception that is driving American AI policy
The final check of the write-up failed.
2026-10-29
PublishedBy usGary Marcus
Marcus does not see any plausible scenario in which either the US or China outright wins the generative AI race to the exclusion of the other.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
The US is not far ahead of China in the generative AI race.
A piece of evidence came from the wrong dates.
2026-10-29
PublishedBy usGary Marcus
Over the next few years both the US and China will build a lot of generative AI infrastructure, quite possibly significantly more than they need.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
The core misconception that is driving American AI policy
The final check failed: the write-up said more than the evidence showed.
2026-10-29
PublishedBy usGary Marcus
Nothing China can realistically do, short of full out nuclear war, is going to stop companies like Google, Microsoft, and Amazon from serving generative AI on the cloud.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
If LLMs turn out to be a dud, or are replaced by smaller, more efficient systems that don't demand such immense infrastructure, it may be all the more the case that the real winner is whichever country doesn't overextend itself to the point of financial ruin.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Generative AI models are commodities.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Building US AI policy around a fantasy that the US is somehow going to crush China in an LLM war, or that China will crush the US, is misguided.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Nothing the United States can realistically do, short of full out nuclear war, is going to stop Chinese companies from serving generative AI on China-based cloud infrastructure.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
China is currently at most a few months behind the US in generative AI.
A piece of evidence came from the wrong dates.
2026-10-29
PublishedBy usGary Marcus
The US-China lead in generative AI will be slight and flip back and forth, with neither retaining a sizeable lead for long.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Especially because chips like GPUs depreciate quickly, it may be that the real winner is whichever country doesn't overextend itself to the point of financial ruin in a foolish effort to win a race that can't be won.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
A CERN for AI and the Global Governance of AI
One of the AIs read the claim differently from our summary.
2026-10-29
PublishedBy usGary Marcus
There is no way to guarantee LLMs won't be jailbroken, and current LLMs can be jailbroken quickly and automatically.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
It is likely that at least one of the 60+ elections worldwide in 2024 will be swayed by deepfakes.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
We lack substantial controls over current AI, especially as it gets connected to real-world services and APIs.
A piece of evidence came from the wrong dates.
2026-10-29
PublishedBy usGary Marcus
Current AI cannot reliably anticipate the consequences or harms of actions it recommends.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
A CERN for AI safety and global AI safety governance should be created, provided they address a balanced range of risks.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
There is no way to guarantee that large language models reliably stick to facts.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
The world can likely benefit enormously from AI, in ways extending far beyond the productivity gains seen in recent months, though there are clearly risks too.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
AI risk discussion should focus much more on misuse of LLMs by criminals and terrorists, not just on AI takeover scenarios.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
AI need not be superintelligent to create serious problems; the real issue is control.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
We need to consider both current AI risk and AGI risk.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
People should not listen to those who deny there is any AI risk, such as Marc Andreessen, because enormous money is at stake.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Some AI problems may be too complex for individual labs and of insufficient financial interest to profit-driven large AI companies, which still holds true.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
A CERN for AI and the Global Governance of AI
The final check failed: a cited source didn’t say what the write-up said, and our summary didn’t match their words.
2026-10-29
PublishedBy usGary Marcus
It seems to Marcus that Anthropic's Constitutional AI idea is clever and the best of the current AI safety ideas, but still inadequate, riddled with the usual LLM problems such as unreliability, excessive dependence on similarity to training data, hallucinations and unreliable reasoning.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
The combined AI safety efforts of large tech companies have not yielded much.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
SoftBank has sold its entire Nvidia position, which is an ominous sign for generative AI.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Meta's financing games have been revealed, presented as one of five recent ominous signs for generative AI.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
5 recent, ominous signs for Generative AI
One of the AIs read the claim differently from our summary.
2026-10-29
PublishedBy usGary Marcus
CoreWeave's stock fell 20% in one week, which is an ominous sign for generative AI.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
Sam Altman could not explain how OpenAI would meet its $1.4 trillion in obligations, which is an ominous sign for generative AI.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
Oracle's debt has been downgraded, which is an ominous sign for generative AI.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Knowing more about exactly what it is, beyond brain size, that makes the human mind unique would tell us an immense amount both about ourselves and about how we might want to build AI.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
What Marcus said to Joscha Bach 26 months earlier still holds, 100%.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
The lords of deep learning have failed to resolve what ails their AI.
The final check failed: the write-up said more than the evidence showed, and a condition they stated was left out.
2026-10-29
PublishedBy usGary Marcus
Despite tens of billions of dollars invested since 2022, deep learning systems are still unreliable and still produce toxicity and hallucinations.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
There has been no progress toward solving the AI alignment problem between 2022 and mid-2024.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Current AI technology, which Musk is now wildly overhyping, remains exactly as incorrigible as it ever was.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
In 2025 AI power consumption will rise and remain a major problem, with few generative AI companies transparent about usage.
The two reviewers reached different verdicts.
2026-10-29
Held backBy usGary Marcus
25 AI Predictions for 2025, from Marcus on AI
The final check of the write-up failed.
2026-10-29
Held backBy usGary Marcus
No single AI system will solve more than 4 of the Marcus-Brundage AI 2027 tasks by end of 2025.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
AI will presumably eventually solve compositionality, factuality and reasoning, but current techniques have not been adequate.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
His predictions about generative AI's technical and economic limits were largely correct, but he was wrong about investor behavior.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
There could be no GPT-5 level model, meaning a huge across-the-board leap by community consensus, throughout 2025.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Marcus suspects, without having tested it, that changes of state and the persistence of objects will still cause problems for Google's Veo 2, though it seems better than Sora.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Possibly none of the Marcus-Brundage tasks will be reliably solved by end of 2025.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Commercial artists and voiceover actors have perhaps been hardest hit by AI so far.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
Available AI models remain essentially GPT-4 level with incremental changes, and OpenAI has nothing it deems worthy of the GPT-5 label.
A reviewer didn’t accept one of the pieces of evidence.
2026-10-29
Held backBy usGary Marcus
Through 2025, profits from AI models will remain modest or nonexistent while chipmakers continue to do well.
A piece of evidence didn’t measure what the claim was about.
2026-10-29
Held backBy usGary Marcus
In 2025 AI Safety Institutes will offer guidance but have little legal authority to stop dangerous models.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
In 2025 there may be a large-scale cyberattack in which generative AI plays an important causal role.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
25 AI Predictions for 2025, from Marcus on AI
One of the AIs read the claim differently from our summary.
2026-10-29
PublishedBy usGary Marcus
By the end of 2025 truly driverless cars, in which no human is required to attend to traffic, will continue to see usage limited to a modest number of cities, mainly in the West and mainly in good weather, and human drivers will still make up a large part of the economy.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Corporate adoption of AI is far more limited than most people expected, and total profits across all companies, except hardware companies like Nvidia that profit from chips rather than models, have been modest at best.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
OpenAI's o3 will likely be quite expensive when released.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
With medium confidence, by the end of 2025 neurosymbolic AI will become much more prominent.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
By the end of 2025 OpenAI will have continued to preview products months, and perhaps years, before they are solid and widely available at an accessible price.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
Most AI model companies have lost money, and at least three have been quasi-acquired with little profit for investors.
A piece of evidence didn’t measure what the claim was about.
2026-10-29
Held backBy usGary Marcus
LLMs did not markedly improve commercial assistants like Amazon Alexa during 2024.
The two reviewers reached different verdicts.
2026-10-29
Held backBy usGary Marcus
In 2025 much of the world will increasingly look to Europe for AI regulatory guidance.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
With medium confidence, by the end of 2025 a technical moat will continue to be elusive, with more convergence on broadly similar models across the US and China, and some European systems catching up to roughly the same place.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
25 AI Predictions for 2025, from Marcus on AI
The final check of the write-up failed.
2026-10-29
PublishedBy usGary Marcus
AI agents will be heavily hyped in 2025 but far from reliable except possibly in very narrow use cases.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
There are many GPT-4 level models, a price war, and very little moat among LLM providers.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
In 2025 the US will continue to have very little regulation protecting consumers from generative AI risks.
A reviewer didn’t accept one of the pieces of evidence.
2026-10-29
PublishedBy usGary Marcus
Through 2025 hallucinations will continue to plague generative AI.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
AGI will not be achieved in 2025.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
There is still no robust solution to LLM hallucinations.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
OpenAI has no profits, does not expect any until at least 2029, burns billions, and lacks a clear technical moat.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
OpenAI's long-term security, especially at high valuations, is far from assured.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
AI-generated misinformation became increasingly widespread, but its effects have thus far been less than Marcus feared; he remains quite concerned and thinks vigilance continues to be required.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
With medium confidence, by the end of 2025 few companies, and even fewer consumers, will adopt o3 at wide scale because of concerns about price and robustness relative to that price.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
By end of 2025 less than 10%, probably less than 5%, of the workforce will be replaced by AI.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
2024 was a year of AI disillusionment for some but not all; his prediction was overstated.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
With medium confidence, by the end of 2025 Sora will continue to have trouble with physics.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
25 AI Predictions for 2025, from Marcus on AI
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
Held backBy usGary Marcus
GPT-5 was not released in 2024 despite repeated predictions.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
His 2023 prediction that OpenAI might someday be seen as the WeWork of AI has not been confirmed, but OpenAI is far from out of the woods.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
By end of 2025 few if any radiologists will be replaced by AI.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Sora's troubles with physics have not gone away since its February 2024 announcement.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
In 2025 we may see o1-like models good at synthetic-data-friendly tasks but only incrementally better than GPT-4 elsewhere.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Through 2025 reasoning errors will continue to plague generative AI.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
With medium confidence, by the end of 2025 companies will continue to experiment with AI, but adoption of production-grade systems scaled out in the real world will continue to be tentative.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
Marcus's 2024 AI predictions have largely held, with some caveats.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
Valuations of major AI companies could start to fall in 2025.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
In 2025 humanoid robotics will be hyped but no one will release a robot remotely as capable as Rosie; situational awareness and flexibility will remain poor.
The two reviewers reached different verdicts.
2026-10-29
PublishedBy usGary Marcus
Copyright lawsuits over generative AI will continue throughout 2025.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
His seven December 2022 predictions about GPT-4 largely held true for every model released in 2024, with prediction number seven still pending.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
His predictions since 2022 that pure scaling of LLMs would eventually start to run out appear to have been confirmed in November and December 2024, based on comments from many industry insiders.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
AI sweatshops result from substituting shallow data for deep understanding in AI design.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Genuine AI would not require vast amounts of training data.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
CNET was the first casualty of putting too much faith in ChatGPT, publishing AI-written stories with many mistakes.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
A bunch of the human filtering work behind ChatGPT was done by poorly paid workers in Kenya, paid less than $2/hour to evaluate graphic material such as descriptions of sexual situations involving children and animals.
The final check failed: the write-up said more than the evidence showed.
2026-10-29
Held backBy usGary Marcus
24 Seriously Embarrassing Hours for AI
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
In choosing what kind of AI to build, sometimes the slower road is the better road.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Anthropic's Claude, despite its alignment focus, still quickly hallucinates like ChatGPT.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
24 Seriously Embarrassing Hours for AI
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
Genuine AI would not require sweatshop labor.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
As of January 2023, no car has completed an intervention-free autonomous drive from LA to NYC.
The two reviewers reached different verdicts.
2026-10-29
Held backBy usGary Marcus
Court testimony indicates some fairly senior driverless-car employees were apparently unaware of human factors engineering, which is essential when humans are in the loop.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
Held backBy usGary Marcus
OpenAI used sweatshop labor behind the scenes.
A reviewer didn’t accept one of the pieces of evidence.
2026-10-29
Held backBy usGary Marcus
24 Seriously Embarrassing Hours for AI
The two reviewers read the claim differently.
2026-10-29
PublishedBy usGary Marcus
Tesla FSD is getting better, and he raises the question of whether it is getting enough better.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
The generative AI investment bubble may not last forever, as signalled by some late-stage VC firms sitting out.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
Generative AI is starting to get a tiny bit of a 'Scarlet Letter' feel, which he presents as possible foreshadowing.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
You could argue that almost every LLM, Grok included, has a slight liberal bias.
The claim is hedged, so it isn’t scored.
2026-10-29
Held backBy usGary Marcus
1984, but with LLM’s
The final check of the write-up failed.
2026-10-29
Held backBy usGary Marcus
If you use the same data as everyone else, it is actually pretty hard to make an LLM that stands out from the pack, short of using some fine-tuning tricks that likely reduce accuracy.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
If a system's input is deliberately altered, it may wind up faring poorly on other benchmarks, a cost Musk may not wish to bear.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
No LLM has fully coherent beliefs, because LLMs repeat conflicting material from inconsistent training data instead of reasoning their way to a coherent view.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
After failing to make Grok match his personal beliefs, Musk will rewrite Grok's training data (history) so that it matches his views.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
PublishedBy usGary Marcus
Musk's plan to rewrite Grok's training data will probably be harder than he expects, because it cannot be done by hand and automating it to his satisfaction will be difficult.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
1984, but with LLM’s
A reviewer said our test wasn’t fair to what was actually said.
2026-10-29
Held backBy usGary Marcus
The Chinese government has already made conformity to its Party line a legal requirement; Putin is probably doing the same; and it is not inconceivable that the present US government might pressure companies to do the same.
A reviewer didn’t accept one of the pieces of evidence.
2026-10-29
Held backBy usGary Marcus
1984, but with LLM’s
The final check of the write-up failed.
2026-10-29
PublishedBy usGary Marcus
Even if LLMs are not AGI, they could easily become the most powerful form of mind control ever invented.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
1984, but with LLM’s
The final check failed: the write-up said more than the evidence showed.
2026-10-29
Held backBy usGary Marcus
AI that is smarter than any single human (human-superior AGI) will not arrive by the end of 2025, contrary to Elon Musk's prediction.
A reviewer didn’t accept one of the pieces of evidence.
2026-10-29
Held backBy usGary Marcus
Bad Actors are Grooming LLMs to Produce Falsehoods
One of the AIs read the claim differently from our summary.
2026-10-29
PublishedBy usGary Marcus
Generative AI chatbots' lack of reasoning can directly contribute to the harmful effects of LLM grooming by disinformation actors.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
Bad Actors are Grooming LLMs to Produce Falsehoods
One of the AIs read the claim differently from our summary.
2026-10-29
Held backBy usGary Marcus
Bad Actors are Grooming LLMs to Produce Falsehoods
One of the AIs read the claim differently from our summary.
2026-10-29
PublishedBy usGary Marcus
Current AIs, being naive mimicry systems, degrade their own futures and training data whenever they repeat propaganda.
Published after both AI reviews and the final check passed.
2026-10-29
PublishedBy usGary Marcus
If an LLM's training data or search-accessible data is tainted, the model may well regurgitate the tainted information.
Published after both AI reviews and the final check passed.
2026-10-29
Held backBy usGary Marcus
Bad Actors are Grooming LLMs to Produce Falsehoods
One of the AIs read the claim differently from our summary.
2026-10-29
PublishedBy usGary Marcus
To the extent that such systems now supply news to a large number of people, o3's failure to note that the Pravda network is a known disinformation network is a serious failure with potentially serious consequences.
Published after both AI reviews and the final check passed.
2026-10-28
CorrectionBy usGary Marcus
Marcus seriously doubts that, by the end of 2025, AI will be able to learn and master the basics of almost any new video game within a few minutes or hours and solve original puzzles in the alternate world of that video game.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
Offered as an unlikely scenario that could nonetheless come true: by the end of 2026, the giant AI infrastructure plays (like Project Stargate) that President Trump championed after his inauguration will look like an unprofitable and underused mistake.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
Filed under 'Medium confidence': at the end of 2025, OpenAI's Sora will still have trouble with physics.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
The current generation of AI agents should be removed from the market until they can be fixed.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
Filed under 'Medium confidence': by the end of 2025, few companies (and even fewer consumers) will adopt OpenAI's o3 at wide scale, because of concerns about price and robustness relative to that price.
A new assessment with the same outcome replaced the previous one.
2026-10-28
Held backBy usGary Marcus
By the end of 2025 there will be no AI system superior to the smartest individual humans, meaning one able to do any intellectual work that any human can do.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-28
Held backBy usGary Marcus
By about February 2025, one year after the post, there will be no perfectly reliable, hallucination-free chatbots.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-28
Held backBy usGary Marcus
In 2029, no AI system will be able to reliably build bug-free code of more than 10,000 lines from a natural-language specification or through interaction with a non-expert user, not counting code glued together from existing libraries.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-28
PublishedBy usGary Marcus
Filed under 'Medium confidence': through the end of 2025, a technical moat will continue to be elusive; instead there will be more convergence on broadly similar models across both the US and China, and some systems in Europe will catch up to roughly the same place.
Published after both AI reviews and the final check passed.
2026-10-28
CorrectionBy usGary Marcus
By the end of 2024, hallucinations by large language models will not have been reduced to the level of human experts.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
OpenAI will not release GPT-5 in 2024.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
During 2024 and 2025, leading AI figures will further revise and lower their public expectations for AI, such as AGI timelines.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
Marcus seriously doubts the projections of near-term AGI, including Dario Amodei's projection that we will have AGI in 2-3 years from August 2023.
A new assessment with the same outcome replaced the previous one.
2026-10-28
Held backBy usGary Marcus
25 AI Predictions for 2025, from Marcus on AI
The final check failed: a cited source didn’t say what the write-up said.
2026-10-28
CorrectionBy usGary Marcus
Marcus suspects that OpenAI's GPT-6 Astra will show loads of problems with open-ended real-world tasks, despite its success on ARC-AGI.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
If 'great' AI research agents do not arrive by the end of 2025, everything later in the AI 2027 scenario's timeline is pushed back.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usGary Marcus
Throughout 2025, AI agents will be endlessly hyped but far from reliable, except possibly in very narrow use cases.
A new assessment with the same outcome replaced the previous one.
2026-10-28
Held backBy usEd Zitron
By the end of the first quarter of 2025 at the latest, investors will begin punishing big tech companies for their AI spending, and the punishment will hit them far harder than NVIDIA.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-28
Held backBy usEd Zitron
If Oracle loses OpenAI's revenue, Oracle will collapse.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-28
Held backBy usEd Zitron
Once OpenAI has raised its current round of $6.5 billion to $7 billion, it will have to raise another round of at least the same size by July 2025, if not sooner.
A reviewer said our test wasn’t fair to what was actually said.
2026-10-28
CorrectionBy usEd Zitron
If OpenAI does not convert to a for-profit structure by the end of 2025, it will be effectively dead.
A new assessment with the same outcome replaced the previous one.
2026-10-28
Held backBy usEd Zitron
OpenAI and Broadcom will not build their custom AI chip (ASIC) in 2026.
The two reviewers reached different verdicts.
2026-10-28
Held backBy usEd Zitron
What Happens If OpenAI Dies?
The final check of the write-up failed.
2026-10-28
CorrectionBy usEd Zitron
By the end of 2025, Microsoft will have spent over $200 billion in capital expenditures since 2023.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usEd Zitron
By the end of 2026, the hyperscalers (Meta, Google, Amazon, and Microsoft) will have sunk over $1.3 trillion into generative AI since March 2022.
A new assessment with the same outcome replaced the previous one.
2026-10-28
CorrectionBy usEd Zitron
If generative AI becomes an albatross around the necks of Google, Microsoft, or Amazon (which Zitron believes it will), these companies will face serious pressure to curtail their investments and capital expenditures.
A new assessment with the same outcome replaced the previous one.