Claims that can’t be checked, page 14 of 37
Claims by Gary Marcus and Ed Zitron. Deadlines are judged as of 2026-10-29.
Claims that can’t be checked (3643)
Public evidence cannot settle these claims, so they are recorded and attributed, not scored. Most were screened out: both AI providers read the claim on its own and found it a value judgment, rhetoric, hedged, too vague to test, or about something no public evidence could show. The rest were found to be recommendations or too open to test in a full review.
Page 14 of 37: claims 1301 to 1400 of 3643, ordered by the date first said, newest first. To find a claim by its words, use search.
Assessment: Can’t be checkedSaid
Gary Marcus Forecast
“More likely, to my mind, is that generative AI will become a mildly useful, widely used commodity that never makes much money for anyone other than chip companies.”
Our reading: In Marcus's view, it is more likely than generative AI being truly 'magic' that generative AI will become a mildly useful, widely used commodity that never makes much money for anyone other than chip companies.
No horizon
Qualifier: more likely, to my mind (compared with generative AI really being magic)
Assessment: Can’t be checkedSaid
Gary Marcus Forecast
“The consequences of being all-in on a single, easily-replicated technology could well turn out to be severe — with or without pressure from China.”
Our reading: The consequences of the US being all-in on a single, easily-replicated technology (scaling large language models) could well turn out to be severe, with or without pressure from China.
No horizon
Qualifier: with or without pressure from China
Assessment: Can’t be checkedSaid
Gary Marcus Causal claim
“Earlier today I wrote about how vibe coding seemed to be losing steam, with usage declining, quite likely because people are growing disenchanted with the results.”
Our reading: The decline in vibe coding usage is quite likely because people are growing disenchanted with the results.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Causal claim
“I smell a really big dose of people believing what they want to believe.”
Our reading: Quick acceptance of Bubeck's claim was likely driven by people believing what they wanted to believe.
No horizon
Source: Erdosgate
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“Some things never change.”
Our reading: 'Some things never change': a closing quip linking the Erdős episode to OpenAI's 2019 Rubik's cube 'solved' claim, which Marcus found misleading.
No horizon
Source: Erdosgate
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“The claim would have been extraordinary, and should have been vetted closely.”
Our reading: Bubeck's Erdős claim, which would have been extraordinary, should have been closely vetted.
No horizon
Scope: this particular Erdős claim
Source: Erdosgate
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“Nevertheless, Anthropic’s costs exploded in the space of a month by $135.2 million (35%) - likely due to the fact that users, as I reported in mid-July, were costing it thousands or tens of thousands of dollars in compute, a problem it still faces to this day”
Our reading: Anthropic's costs rose by $135.2 million (35%) in a month, to September 2025, likely because some users were costing it thousands or tens of thousands of dollars in compute.
No horizon
Qualifier: likely
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“And I was god damn right. Or, at least, it sure looks like I am. I’m hedging, forgive me. I cannot say for certain, but I see a pattern.”
Our reading: The Subprime AI Crisis he predicted, in which AI providers raise subsidized prices and customers pass them on, appears to be happening.
No horizon
Qualifier: cannot say for certain
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“When I wrote that Anthropic and OpenAI had begun the Subprime AI Crisis back in July, I assumed that the increase in costs was burdensome, but having the information from its AWS bills, it seems that Anthropic’s actions directly caused Cursor’s costs to explode by over 100%.”
Our reading: Anthropic's actions appear to have directly caused Cursor's costs to rise by over 100%.
No horizon
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“Now, the jump in costs — $65.1 million or so between April and May — likely comes as a result of the final training for Sonnet and Opus 4, as well as, I imagine, some sort of testing to make sure Claude Code was ready to go.”
Our reading: Anthropic's roughly $65.1 million AWS cost jump from April to May 2025 likely came from final training of Sonnet and Opus 4 and Claude Code testing.
No horizon
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Recommendation or value judgment
“It’s very difficult to see Service Tiers as anything other than an aggressive rent-seeking maneuver.”
Our reading: It is very difficult to see Anthropic's Service Tiers as anything other than an aggressive rent-seeking maneuver.
No horizon
Qualifier: very difficult to see as anything other than
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Recommendation or value judgment
“I’m tired that so many people have tried to pretend there’s a justification for burning billions of dollars every year, clinging to empty tropes about how this is just like Uber or Amazon Web Services, when Anthropic has built something far more mediocre.”
Our reading: Claims that Anthropic's billions in annual burn are justified by comparisons to Uber or AWS are empty tropes, because Anthropic has built something far more mediocre.
No horizon
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Forecast
“What I believe happened to Cursor will likely happen to every AI-native company, because in a very real sense, Anthropic’s products are a wrapper for its own models, except it only has to pay the (unprofitable) costs of running them on Amazon Web Services and Google Cloud.”
Our reading: Every AI-native company will likely suffer what happened to Cursor: price increases from model providers that also compete with it.
No horizon
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“In my opinion — and this is just my gut instinct — I believe that it is either significantly more expensive to run Anthropic than we know, or Anthropic’s leaked (and stated) revenue numbers are worse than we believe.”
Our reading: In his opinion, as a gut instinct, either Anthropic is significantly more expensive to run than is known, or its leaked and stated revenue numbers are worse than believed.
No horizon
Qualifier: in my opinion, just my gut instinct
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“I am guessing, because I do not know, but with $4.85 billion of other expenses to account for, it’s reasonable to believe Anthropic spent an amount similar to its AWS spend on Google Cloud.”
Our reading: Anthropic plausibly spent about as much on Google Cloud in 2024 as on AWS.
No horizon
Qualifier: a guess
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“It’s likely the June 2025 spike in revenue came from the introduction of service tiers, which specifically target prompt caching, increasing the amount of tokens you’re charged for as an enterprise customer based on the term of the contract, and your forecast usage.”
Our reading: Anthropic's June 2025 revenue spike likely came from introducing service tiers that raised charges on prompt caching.
No horizon
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“I have sat with these numbers for a great deal of time, and I can’t find any evidence that Anthropic has any path to profitability outside of aggressively increasing the prices on their customers to the point that its services will become untenable for consumers and enterprise customers alike.”
Our reading: After long study of the numbers, he cannot find any evidence that Anthropic has a path to profitability other than raising customer prices so aggressively that its services become untenable for consumers and enterprises.
No horizon
Qualifier: he can't find any evidence
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Forecast
“As a result, both OpenAI and Anthropic can (and may very well!) devour the market of any company that builds on top of their models.”
Our reading: OpenAI and Anthropic may take over the markets of companies that build on their models.
No horizon
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“Anthropic and Amodei’s sudden willingness to take large swaths of capital from the Gulf States does not suggest that it’s not at least a little desperate for capital, especially given Anthropic has, according to Bloomberg, “recently held early funding talks with Abu Dhabi-based investment firm MGX” a month after raising $13 billion.”
Our reading: Anthropic's sudden willingness to take large amounts of Gulf State capital, along with its early funding talks with MGX a month after raising $13 billion, does not suggest that it is not at least a little desperate for capital.
No horizon
Qualifier: according to Bloomberg's report of early talks with MGX
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“As you can see, Cursor’s costs continue to balloon in July, and I am guessing it’s because of the Service Tiers situation — which, I believe, indirectly resulted in Cursor pushing more users to models that it runs on Amazon’s infrastructure.”
Our reading: He guesses that Cursor's AWS costs kept ballooning in July 2025 because of the Service Tiers situation, which he believes indirectly led Cursor to push more users to models it runs on Amazon's infrastructure.
No horizon
Qualifier: I am guessing
Source: This Is How Much Anthropic and Cursor Spend On Amazon Web Services
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“First slowly, and then all at once, dreams of LLMs bringing us to the cusp of AGI have fallen apart.”
Our reading: Hopes that LLMs would bring us close to AGI have collapsed over the past few months.
No horizon
Source: Game over. AGI is not imminent, and LLMs are not the royal road to getting there.
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“If the strengths of A.I. are to truly be harnessed, the tech industry should stop focusing so heavily on these one-size-fits-all tools, and instead concentrate on narrow, specialized A.I. tools engineered for particular problems”
Our reading: To harness AI's strengths, the tech industry should shift focus from general-purpose tools to narrow, specialized AI tools for particular problems.
No horizon
Qualifier: if the strengths of AI are to be harnessed
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“In an OpEd today in The New York Times, I argue that at least for now, AGI is maybe not the right near-term goal:”
Our reading: At least for now, AGI is maybe not the right near-term goal for AI.
No horizon
Qualifier: at least for now
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“Right now, it feels as if Big Tech is throwing general-purpose A.I. spaghetti at the wall and hoping that nothing truly terrible sticks.”
Our reading: Right now, it feels as if Big Tech is throwing general-purpose AI at the wall like spaghetti and hoping nothing truly terrible sticks.
No horizon
Qualifier: it feels as if
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Causal claim
“Those obstacles may be one reason generative A.I. hasn’t led to the skyrocketing in profits and productivity that many in the tech industry predicted.”
Our reading: Hallucinations and errors may be one reason generative AI has not produced the predicted surge in profits and productivity.
No horizon
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“The whole thing should be viewed as evolving, with future measures replacing existing ones as researchers devise better benchmarks — and not as a static checklist.”
Our reading: The AGI definition framework should be treated as evolving, with better benchmarks replacing existing ones, not as a static checklist.
No horizon
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“I am someone who believes that AGI (Artificial General Intelligence) could change the world, but also someone who thinks that LLMs are not the right path there, morally or technically.”
Our reading: Marcus thinks LLMs are not the right path to AGI, either morally or technically.
No horizon
Qualifier: morally or technically
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“The main alternative — definitions of AGI that are in terms of economic criteria like profits (AGI as being $100 billion in profits) or percentage of jobs displaced— seem to me to be fundamentally misguided.”
Our reading: To Marcus, definitions of AGI based on economic criteria such as profits or percentage of jobs displaced seem fundamentally misguided.
No horizon
Qualifier: seem to me
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“Humanity would be better served by labs devoting more resources on building specialized tools for science, medicine, technology and education.”
Our reading: AI labs should devote more resources to building specialized tools for science, medicine, technology and education.
No horizon
Source: Is AGI the right goal for AI?
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“I think we are making a huge mistake.”
Our reading: Marcus thinks that we are making a huge mistake by reshaping our world on the premise that we are close to AGI.
No horizon
Qualifier: I think
Source: The Grand AGI Delusion
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“The conservative estimate of capital expenditures related to data centers is around $400 billion, but given the $50 billion a quarter in private credit, I’m going to guess it breaks $500 billion, all to build capacity for an industry yet to prove itself.”
Our reading: Given $50 billion a quarter in private credit, he guesses that data-center-related capital expenditures break $500 billion, above a conservative estimate of around $400 billion, all to build capacity for an industry yet to prove itself.
No horizon
Qualifier: his guess
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Rhetoric
“The only thing “powerful” about generative AI is its mythology.”
Our reading: The only powerful thing about generative AI is its mythology.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Rhetoric
“This is nothing like anything you’ve seen before, because this is the dumbest shit that the tech industry has ever done.”
Our reading: The AI bubble is unlike previous bubbles and is the dumbest thing the tech industry has done.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Forecast
“Anyway, I cannot be clear enough how unlikely it is that (as NVIDIA has said) “the first gigawatt of NVIDIA systems will be deployed in the second half of 2026,” and that’s if it has bought the land and got the permits and ordered the construction, none of which has happened yet.”
Our reading: It is very unlikely that the first gigawatt of NVIDIA systems for OpenAI will be deployed in the second half of 2026.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“It’s often emotive — or at the very least, driven or inspired by a given emotion — which is something that an AI simply can’t replicate in a way that’s authentic and believable.”
Our reading: AI cannot replicate emotive good writing in an authentic, believable way.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Rhetoric
“The bubble has entered its most pornographic, aggressive and destructive stage, where the more obvious it becomes that they’re cooked, the more ridiculous the generative AI industry will act — a dark juxtaposition against every new study that says “generative AI does not work” or new story about ChatGPT’s uncanny ability to activate mental illness in people.”
Our reading: The bubble has entered its most pornographic, aggressive and destructive stage, where the more obvious it becomes that they're cooked, the more ridiculous the generative AI industry will act.
No horizon
Qualifier: the more obvious it becomes that they're cooked
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Forecast
“The only reason anyone humours any of this crap is the endless hoarding of GPUs to build capacity for a revolution that will never arrive.”
Our reading: The AI revolution that GPU hoarding is building capacity for will never arrive.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Recommendation or value judgment
“Members of the media, I am begging you, stop doing this.”
Our reading: The media should stop repeating claims that AI is replacing software engineers without checking them.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Rhetoric
“The model companies similarly started going after individual and business customers, offering software and subscriptions that promised the world, though this mostly boiled down to chatbots that could generate stuff, and then doubled down with the promise of “agents” — a marketing term that’s meant to make you think “autonomous digital worker” but really means “broken digital product.””
Our reading: AI "agents" are a marketing term for what are really broken digital products.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“I must be clear that these deals are intentionally made to continue the myth of generative AI, to pump NVIDIA, and to make sure OpenAI insiders can sell $10.3 billion of shares.”
Our reading: The OpenAI-Oracle and NVIDIA-OpenAI deals are intentionally made to continue the myth of generative AI, to pump NVIDIA, and to make sure OpenAI insiders can sell $10.3 billion of shares.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“In a very real sense, OpenAI threatens the future of available capital for the tech industry.”
Our reading: OpenAI threatens the future availability of capital for the tech industry.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“This NVIDIA deal is a farce — an obvious attempt by the largest company on the American stock market to prop up the one significant revenue-generator in the entire industry, knowing that time is running out for it to create new avenues for eternal growth.”
Our reading: The NVIDIA-OpenAI deal is a farce, an obvious attempt by NVIDIA to prop up OpenAI, the industry's one significant revenue-generator, knowing that time is running out for it to create new avenues for eternal growth.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Rhetoric
“Generative AI exists for two reasons: to cost money, and to make executives look busy.”
Our reading: Generative AI exists only to cost money and make executives look busy.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Causal claim
“Generative AI was always set up to fail, because it was meant to be everything, was talked about like it was everything, is still sold like it’s everything, yet for all the fucking hype, it all comes down to two companies: OpenAI, and, of course, NVIDIA.”
Our reading: Generative AI was always set up to fail because it was oversold, and it all comes down to OpenAI and NVIDIA.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Ed Zitron Rhetoric
“OpenAI is the generative AI industry — and it’s a dog of a company.”
Our reading: OpenAI is the generative AI industry, and it is a dog of a company.
No horizon
Source: The Case Against Generative AI
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“Practically the only people still pretending scaling LLMs is “all you need” are grifters.”
Our reading: Practically the only people still pretending that scaling LLMs is "all you need" are grifters.
No horizon
Qualifier: practically
Source: Game over for pure LLMs. Even Turing Award Winner Rich Sutton has gotten off the bus.
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“With only a small fraction of the current and planned investments in LLMs, we can and should find out.”
Our reading: With only a small fraction of current and planned LLM investment, we can find out whether Sutton's or Marcus's preferred approach is right, and we should do so.
No horizon
Qualifier: with only a small fraction of the current and planned investments in LLMs
Source: Game over for pure LLMs. Even Turing Award Winner Rich Sutton has gotten off the bus.
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“When the LLM crowd has lost Sutton – and when Sutton sounds exactly like me – it’s game over.”
Our reading: With Sutton having turned against pure LLMs, it is game over for the pure LLM approach.
No horizon
Qualifier: Sutton has abandoned the LLM camp
Source: Game over for pure LLMs. Even Turing Award Winner Rich Sutton has gotten off the bus.
Assessment: Can’t be checkedSaid
Gary Marcus Causal claim
“Quite likely we need some of both.”
Our reading: Quite likely, AI needs some of both approaches: Sutton's reinforcement learning and Marcus's neurosymbolic approaches with innate constraints.
No horizon
Qualifier: quite likely
Source: Game over for pure LLMs. Even Turing Award Winner Rich Sutton has gotten off the bus.
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“That said I have always thought it was wrong, or more to the point, overstated, and fatally flawed.”
Our reading: Marcus has always thought Sutton's Bitter Lesson essay was wrong, or more precisely overstated, and fatally flawed.
No horizon
Qualifier: or more to the point, overstated
Source: Game over for pure LLMs. Even Turing Award Winner Rich Sutton has gotten off the bus.
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“Just think how high that percentage can go!”
Our reading: Gary Marcus suggests that the share of workers receiving low-quality AI-generated 'workslop', now 40%, could rise much higher.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“I signed, too—because we have let too much slide, and done too little to face the risks.”
Our reading: We have let too much slide and done too little to face the risks of AI, which is Marcus's reason for signing the global call for AI red lines.
No horizon
Qualifier: given as his personal reason for signing the letter
Source: AI red lines
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“I think the truth is somewhere in between.”
Our reading: Marcus thinks the truth about Yudkowsky and Soares' book on potential catastrophic AGI threats lies somewhere between Max Tegmark's view that it is 'the most important book of the decade' and the New York Times likening it to Scientology.
No horizon
Qualifier: I think
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“Don’t panic. Don’t dismiss.”
Our reading: People should neither panic about nor dismiss the potential catastrophic threats from AGI argued in Yudkowsky and Soares’ book.
No horizon
Qualifier: in response to Yudkowsky and Soares’ book on catastrophic AGI threats
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“Oracle’s new market cap, near a trillion dollars, up nearly 50% this week, driven largely by this one apparently non-binding deal with a party that doesn’t have the money to pay for the services, seems more bonkers than most.”
Our reading: Oracle's new market cap, near a trillion dollars and up nearly 50% this week, largely driven by an apparently non-binding deal with a party that lacks the money to pay for the services, seems more bonkers than most AI valuations.
No horizon
Qualifier: seems
Source: Peak bubble
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“We are well past peak bubble, in fact, and into peak musical chairs.”
Our reading: The AI market is well past peak bubble and has reached a peak game of musical chairs.
No horizon
Source: Peak bubble
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“I don’t know when the GenAI bubble will end. But this has got to be peak bubble:”
Our reading: The generative AI market is a bubble and has reached peak bubble, though Marcus does not know when it will end.
No horizon
Source: Peak bubble
Assessment: Can’t be checkedSaid
Gary Marcus Forecast
“The other problem here is that the total value of the tech market as whole, which is supposed to reflect the future of value of the companies within it, far exceeds what is likely ever to be delivered.”
Our reading: The total value of the tech market as a whole far exceeds the value its companies are likely ever to deliver.
No horizon
Qualifier: likely ever to be delivered
Source: Peak bubble
Assessment: Can’t be checkedSaid
Gary Marcus Conditional forecast
“It’s not going to be pretty when the music stops.”
Our reading: When the AI bubble ends, the outcome will be bad.
No horizon
Only if: when the music stops (the bubble ends)
Source: Peak bubble
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“Let us all hope that we can learn something from the mistakes of the last few years, and find a new path — to an AI that better serves humanity.”
Our reading: Marcus hopes that we all can learn something from the mistakes of the last few years and find a new path to an AI that better serves humanity.
No horizon
Qualifier: let us all hope
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“If there is any true exponential nowadays, I fear it is this: the rapid doubling of AI-generated misinformation.”
Our reading: Marcus fears that, if there is any true exponential nowadays, it is the rapid doubling of AI-generated misinformation.
No horizon
Qualifier: if there is any true exponential nowadays
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“The field is overdue for a rethink.”
Our reading: The AI field should rethink its approach, and that rethink is overdue.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Causal claim
“Because they live in the land of correlations between images and text, not in the land of ideas.”
Our reading: The image-generating AI systems under discussion are unpredictably right or wrong because they operate in the land of correlations between images and text, not in the land of ideas.
No horizon
Scope: these systems
Source: Spinal Tap meets John Malkovich
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“These systems can draw things that kinda look like the world, but they obviously don’t know what’s what or how things work outside of their training data.”
Our reading: Generative image systems can produce world-like images but do not understand what things are or how they work outside their training data.
No horizon
Qualifier: outside of their training data
Source: Spinal Tap meets John Malkovich
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“Beauty is only skin deep.”
Our reading: The visual polish of generative image systems is superficial and hides a lack of understanding.
No horizon
Source: What scaling does and doesn’t buy you: peeling back the hype surrounding Google‘s trendy Nano Banana
Assessment: Can’t be checkedSaid
Gary Marcus Causal claim
“Thanks to scaling (and some other tricks), the graphics are terrific, and the ability to have it edit photos that you upload is genuinely cool.”
Our reading: Thanks to scaling (and some other tricks), Google's Nano Banana has terrific graphics, and its ability to edit uploaded photos is genuinely cool.
No horizon
Qualifier: and some other tricks
Source: What scaling does and doesn’t buy you: peeling back the hype surrounding Google‘s trendy Nano Banana
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“But it’s all still just an extended form of mimicry, not something deeper, as the persistent failures with parts and wholes keep showing us.”
Our reading: Image and video generation systems like these are still just an extended form of mimicry, not something deeper, as their persistent failures with parts and wholes keep showing.
No horizon
Scope: these systems (image and video generators discussed)
Source: What scaling does and doesn’t buy you: peeling back the hype surrounding Google‘s trendy Nano Banana
Assessment: Can’t be checkedSaid
Ed Zitron Forecast
“Big tech has spent over $500 billion in capital expenditures in the last 18 months, and all told — between investments of cloud credits and infrastructure — will likely sink over $600 billion by year's-end.”
Our reading: Big tech, having spent over $500 billion in capital expenditures in the last 18 months, will likely sink over $600 billion all told, counting investments of cloud credits and infrastructure, by year's end.
No horizon
Scope: all told, between investments of cloud credits and infrastructure
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“AI is being sold as a transformational technology, and I am yet to see it transform anything.”
Our reading: Although marketed as transformational, generative AI has not transformed anything he has seen.
No horizon
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“We Are Not In The Early Days Of Generative AI, And Anyone Using This Argument Is Either Ignorant Or Intentionally Deceptive”
Our reading: Generative AI is not in its early days.
No horizon
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Forecast
“In fact, I believe that thanks to effectively infinite resources, we've speed-run the entire Large Language Model era, and we're nearing the end.”
Our reading: Because of effectively unlimited resources, the Large Language Model era has been sped through and is nearing its end.
No horizon
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“The answer, of course, is that it has no intellect, is not intelligent, and Large Language Models are being pushed up a mountain by a cadre of people who are either easily impressed or invested — either emotionally or financially — in its success due to the company they keep or their intentions for the world.”
Our reading: LLMs have no intellect and are not intelligent; they are promoted by people who are easily impressed or have an emotional or financial stake in them.
No horizon
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Rhetoric
“Anyway, AI 2027 is fan fiction, nothing more, and just because it’s full of fancy words and has five different grifters on its byline doesn’t mean anything.”
Our reading: The AI 2027 scenario is fiction with no forecasting value.
No horizon
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“We don’t know if AGI is possible, anyone claiming they do is lying.”
Our reading: Nobody knows whether AGI is possible, and anyone who claims to know is lying.
No horizon
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Recommendation or value judgment
“Anyone falling for this is a mark, and OpenAI should have gone out of its way to correct people.”
Our reading: OpenAI should have actively corrected misreported claims that GPT-4 hired a TaskRabbit worker.
No horizon
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“As I've said in the past, OpenAI is deliberately using Weekly Active Users so that it doesn't have to publish its monthly active users, which I believe would be higher.”
Our reading: OpenAI is deliberately using weekly active users so that it doesn't have to publish its monthly active users, which he believes would be higher.
No horizon
Qualifier: he believes monthly active users would be higher
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“As things begin to collapse — and they sure look like they're collapsing, but I am not making any wild claims about "the bubble bursting" quite yet — it will look increasingly more deranged to bluntly publish everything that these companies say.”
Our reading: Things in the AI industry sure look like they're beginning to collapse, though he is not making claims about 'the bubble bursting' quite yet.
No horizon
Qualifier: not making any wild claims about 'the bubble bursting' quite yet
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Ed Zitron Observation
“There's a good chance that Anthropic and OpenAI dominate the capex of Amazon and Microsoft, in part because what the fuck else are they buying all these GPUs for, as their own AI services don't seem to be making that much money at all.”
Our reading: There is a good chance that Anthropic and OpenAI dominate the capital expenditures of Amazon and Microsoft.
No horizon
Qualifier: a good chance
Source: How To Argue With An AI Booster
Assessment: Can’t be checkedSaid
Gary Marcus Forecast
“The Klarna Effect may come back to haunt them.”
Our reading: Employers replacing employees with AI may end up reversing course as Klarna did.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Conditional forecast
“If we stop hiring junior employees, we risk destroying the pipeline to expertis.”
Our reading: Stopping hiring of junior employees in favor of AI risks destroying the pipeline to expertise.
No horizon
Only if: if we stop hiring junior employees
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“Without saying for sure that I know what will happen next — after all things like stock prices and the valuations AI startups are as much a matter of unpredictable crowd psychology as they are of (in this case, shaky) technical and economic fundamentals — I will leave this here, a cartoon I made (with ChatGPT’s help!) in mid-July.”
Our reading: The technical and economic fundamentals behind AI startup valuations are shaky.
No horizon
Source: Is this the moment when the Generative AI bubble finally deflates?
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“Not that Sam is deterred; in his mind, a fuckup is just an excuse to ask for money:”
Our reading: Sam Altman is not deterred by a failure; in his mind, a failure is just an excuse to ask for money.
No horizon
Source: Is this the moment when the Generative AI bubble finally deflates?
Assessment: Can’t be checkedSaid
Gary Marcus Conditional forecast
“Once the markets really understand this, enthusiasm may indeed collapse fairly quickly.”
Our reading: Once the markets really understand the point made in Marcus's cartoon, enthusiasm for generative AI may collapse fairly quickly.
No horizon
Only if: once the markets really understand 'this' (the point of the cartoon he shared, whose contents are not given in the text)
Source: Is this the moment when the Generative AI bubble finally deflates?
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“The technical foundation was, as I argued all the way back to the early days of GPT-2, never strong enough to support the hype.”
Our reading: The technical foundation of LLMs has never been strong enough to support the hype.
No horizon
Source: Is this the moment when the Generative AI bubble finally deflates?
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“the user has the option to accept a code change or not, but with so many changes to make at such a rapid pace, a developer in hurrry (which is almost every developer) could easily miss an attack, and ok the change or run a command”
Our reading: Even with manual approval, hurried developers could easily approve malicious changes suggested by coding agents.
No horizon
Qualifier: even with Auto-Run off
Assessment: Can’t be checkedSaid
Gary Marcus Forecast
“(Nvidia advised, rightly, that everyone should disable Auto-Run mode if they’ve activated it, but we fear that many users may use it anyway, because it is so fast.)”
Our reading: Many Cursor users may keep using Auto-Run mode despite the security risk because it is fast.
No horizon
Qualifier: because it is so fast
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“As long as we have agents roaming the internet and otherwise incorporating data that they don’t fully understand – and LLMs don’t ever fully understand the data they are leveraging – there is enormous risk.”
Our reading: LLMs never fully understand the data they leverage.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Forecast
“Given current implementations of the technology, it's hard to imagine enough patches on the planet to thwart them all.”
Our reading: Given current implementations, it is unlikely that patches can thwart all data-injection attacks on generative AI systems.
No horizon
Qualifier: given current implementations
Assessment: Can’t be checkedSaid
Gary Marcus Conditional forecast
“Sure the advice is good, as far is it goes, but if the goods inside are valuable enough, motivated thieves may well still find a way.”
Our reading: Even with mitigations, motivated attackers may still compromise valuable systems using coding agents.
No horizon
Only if: even with mitigations
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“The best defense would be not using agentic coding altogether.”
Our reading: The best security defense is to not use agentic coding at all.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Forecast
“But the tools are so seductive that we doubt many developers will resist.”
Our reading: Few developers are likely to refrain from using agentic coding tools.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“There is a real sense in which GPT-5 is starting to look Altman’s Waterloo.”
Our reading: GPT-5 is starting to look like Altman's decisive defeat.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“Two years ago, many treated Altman like an oracle; in the eyes of many, Altman now looks more like a snake oil salesman.”
Our reading: Two years ago many people treated Altman like an oracle, and now, in the eyes of many, he looks more like a snake oil salesman.
No horizon Evidence as of
Limitation: No poll or systematic measure of views of Altman in 2023 or in August 2025 was found.
Assessment: Can’t be checkedSaid
Gary Marcus Conditional forecast
“As people begin to recognize that truth (the bitter lesson about limits of Sutton’s The Bitter Lesson), scientists will start to charts new paths. One of the new paths may even give rise to what we so desperately need: AI that we can trust.”
Our reading: As the limits of scaling become recognized, scientists will pursue new approaches, one of which may produce trustworthy AI.
No horizon
Only if: as people recognize the limits of scaling LLMs
Assessment: Can’t be checkedSaid
Gary Marcus Rhetoric
“But media coverage is not science, and braggadacio alone cannot yield AGI.”
Our reading: Hype and bragging alone cannot produce AGI.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Causal claim
“And you can’t make AI better if you drown out the critics and keep pouring good money after bad.”
Our reading: AI cannot be made better if you drown out the critics and keep pouring good money after bad.
No horizon
Only if: if you drown out the critics and keep pouring good money after bad
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“The LLM itself still isn’t inducing the rules. And this is a very special-purpose hack, not a remotely general solution. (And, e.g., it would likely fall apart if you changed the rules to play an alternative version of chess.)”
Our reading: A big company's added python code for chess legality is a special-purpose hack, the LLM still doesn't induce the rules, and it would likely fail on chess variants.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“But no matter how much LLMs mimic the patterns of human language, they are not like us. They sound like us, but they don’t think like us.”
Our reading: LLMs do not think like humans, however well they mimic human language.
No horizon
Qualifier: no matter how much they mimic human language
Assessment: Can’t be checkedSaid
Gary Marcus Recommendation or value judgment
“Always regard them like the weird function approximators they are; never trust them.”
Our reading: People should never trust LLMs and should treat them as weird function approximators.
No horizon
Assessment: Can’t be checkedSaid
Gary Marcus Observation
“We still feel that way. And the field still hasn’t put the time in.”
Our reading: The AI field still has not put the time in on giving time, space, and causality a solid computational footing.
No horizon