Claims that can’t be checked by Gary Marcus, page 9 of 24
Deadlines are judged as of 2026-10-29. About Gary Marcus and the sources covered.
Claims that can’t be checked by Gary Marcus (2353)
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 9 of 24: claims 801 to 900 of 2353, ordered by the date first said, newest first. To find a claim by its words, use search.
Assessment: Can’t be checkedSaid
Causal claim
“A lot of recent improvements are likely coming from improvements in using symbolic tools, rather than scaling.”
Our reading: Much recent AI improvement likely comes from better use of symbolic tools rather than scaling.
No horizon
Source: How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Including LLMs somewhere in the next evolution of AI makes sense to me, but leaving them at the core may be a mistake.”
Our reading: LLMs may belong in future AI systems, but keeping them at the core may be a mistake.
No horizon
Source: How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Assessment: Can’t be checkedSaid
Causal claim
“But it may well be as simple as money. The message that we can simply scale our way to AGI is incredibly attractive to investors because it puts money as the central (and sufficient) force needed to advance.”
Our reading: The industry's connectionist-only focus may be explained by money, because the scaling narrative appeals to investors.
No horizon
Source: How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Assessment: Can’t be checkedSaid
Causal claim
“Many other important ideas have likely also yet to be discovered, precisely because the Hinton path has distracted immense resources from other ideas, fostering an intellectual monoculture that, in the words of Emily Bender, has been “sucking the oxygen from the room.””
Our reading: The dominance of the pure neural-network path has likely kept important ideas undiscovered by diverting resources into a monoculture.
No horizon
Source: How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Assessment: Can’t be checkedSaid
Observation
“A neurosymbolic substrate for AI by itself (as I argued in 2020) is likely to be necessary but not sufficient.”
Our reading: A neurosymbolic substrate is likely necessary but not sufficient for AGI.
No horizon
Source: How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Assessment: Can’t be checkedSaid
Causal claim
“AlphaFold, AlphaProof, and AlphaGeometry are all successful neurosymbolic models. We would likely have more of such innovation already, if Hinton had not so insistently shaped the modern landscape in such a narrow-minded way.”
Our reading: AlphaFold, AlphaProof and AlphaGeometry are successful neurosymbolic models, and there would likely be more such innovation had Hinton not narrowed the field.
No horizon
Qualifier: if Hinton had not shaped the landscape narrowly
Source: How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Assessment: Can’t be checkedSaid
Observation
“Explicit cognitive models (as I discussed recently in this newsletter) are also likely to be essential, yet remain very much underexplored.”
Our reading: Explicit cognitive models are also likely to be essential, yet they remain very much underexplored.
No horizon
Source: How o3 and Grok 4 Accidentally Vindicated Neurosymbolic AI
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“To the extent that such systems are now supplying news to a large number of people, this is a serious failure, with potentially serious consequences.”
Our reading: 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.
No horizon
Qualifier: to the extent that such systems are now supplying news to a large number of people
Assessment: Can’t be checkedSaid
Conditional forecast
“If the data that a model has been trained on or can access via real-time search has been tainted, then the model may well just regurgitate that tainted information.”
Our reading: If an LLM's training data or search-accessible data is tainted, the model may well regurgitate the tainted information.
No horizon
Only if: if training or search data has been tainted
Assessment: Can’t be checkedSaid
Rhetoric
“In the meantime, systems of naive mimicry and regurgitation, such as the AIs we have now, are soiling their own futures (and training databases) every time they unthinkingly repeat propaganda.”
Our reading: Current AIs, being naive mimicry systems, degrade their own futures and training data whenever they repeat propaganda.
No horizon
Assessment: Can’t be checkedSaid
Causal claim
“GenAI powered chatbots’ lack of reasoning can directly contribute to the nefarious effects of LLM grooming: the mass-production and duplication of false narratives online with the intent of manipulating LLM outputs.”
Our reading: Generative AI chatbots' lack of reasoning can directly contribute to the harmful effects of LLM grooming by disinformation actors.
No horizon
Assessment: Can’t be checkedSaid
Causal claim
“Debugging code you didn’t write can be hard.”
Our reading: One possible mechanism behind the AI-coding slowdown is that debugging code you didn't write (AI-generated code) can be hard.
No horizon
Qualifier: if the METR finding is true
Source: BREAKING NEWS: AI coding may not be helping as much as you think
Assessment: Can’t be checkedSaid
Observation
“People might be imagining productivity gains that they are not getting, and ignoring real-world costs, to boot.”
Our reading: People using AI coding tools may be imagining productivity gains they are not getting while ignoring real-world costs.
No horizon
Qualifier: if the METR finding is general and replicable
Source: BREAKING NEWS: AI coding may not be helping as much as you think
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“But for now, the disconnnect between what coders thought they would get out of the tools efficiency-wise and what they actually did get out of them is cause for reevaluation.”
Our reading: The gap between coders' expected and actual efficiency gains from AI tools warrants reevaluating the tools' value.
No horizon
Qualifier: for now
Source: BREAKING NEWS: AI coding may not be helping as much as you think
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“IF this is a general, replicable finding, it’s a serious blow to generative AI’s flagship use case.”
Our reading: If METR's finding that AI tools slowed down task completion is general and replicable, it is a serious blow to generative AI's flagship use case.
No horizon
Only if: if this is a general, replicable finding
Source: BREAKING NEWS: AI coding may not be helping as much as you think
Assessment: Can’t be checkedSaid
Conditional forecast
“When we are really at AGI it will be much harder to do so.”
Our reading: Once AI truly reaches AGI, finding compositionality counterexamples will be much harder than it is now.
No horizon
Only if: when we are really at AGI
Assessment: Can’t be checkedSaid
Causal claim
“One is data contamination, since any given prompt that has been circulated publicly potentially becomes part of the training data, including with custom data augmentation, especially with large armies of subcontractors.”
Our reading: Publicly circulated prompts may become part of training data (including via data augmentation by subcontractors), making success on them unreliable evidence of general ability.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“It is an enormous, unjustified leap from the motte of “the systems can now solve some compositionality problems” to the bailey of “the systems have mastered compositionality”.”
Our reading: It is an enormous, unjustified leap from 'the systems can now solve some compositionality problems' to 'the systems have mastered compositionality'.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“But it is not reasonable to conclude from the occasional success on one sort of publicly-discussed set of prompts that GPT has mastered compositionality writ large.”
Our reading: It is not reasonable to conclude from occasional success on one sort of publicly discussed set of prompts that GPT has mastered compositionality in general.
No horizon
Scope: from occasional success on one sort of publicly-discussed set of prompts
Assessment: Can’t be checkedSaid
Rhetoric
“We are immersed in a sea of bullshit.”
Our reading: Society is saturated with AI-generated falsehoods.
No horizon
Source: Botshit Gone Wild
Assessment: Can’t be checkedSaid
Forecast
“AI well and could truly could be a world-changing positive force, but Generative AI, I am not so sure.”
Our reading: AI could truly be a world-changing positive force, but the author is not so sure that generative AI could be.
No horizon
Scope: AI in general, as opposed to generative AI
Source: Botshit Gone Wild
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The central importance of such cognitive (a.k.a. world) models to AI was something that I highlighted again in my 2020 article The Next Decade in AI, which laid out a roadmap that I still believe the field ought to be following.”
Our reading: The AI field should follow the roadmap centred on cognitive world models that Marcus set out in 2020.
No horizon
Source: Generative AI’s crippling and widespread failure to induce robust models of the world
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“In some ways LLMs far exceed humans, but in other ways, they are still no match for an ant. Without robust cognitive models of the world, they should never be fully trusted.”
Our reading: As long as LLMs lack robust cognitive models of the world, they should never be fully trusted.
No horizon
Qualifier: without robust cognitive models of the world
Source: Generative AI’s crippling and widespread failure to induce robust models of the world
Assessment: Can’t be checkedSaid
Forecast
“Their work is interesting, and converges to a degree with the current argument, but my intuition is that such functions will not suffice for AGI.”
Our reading: Marcus's intuition is that world models defined as environment transition functions will not be enough for AGI.
No horizon
Source: Generative AI’s crippling and widespread failure to induce robust models of the world
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Inducing the rules of chess should be table stakes for AGI.”
Our reading: Any system that counts as AGI should, at a minimum, be able to induce the rules of chess.
No horizon
Source: Generative AI’s crippling and widespread failure to induce robust models of the world
Assessment: Can’t be checkedSaid
Rhetoric
“In the tricky world of human relationships, parroting words can only get you so far.”
Our reading: Parroting words has limited value for understanding human relationships.
No horizon
Source: Generative AI’s crippling and widespread failure to induce robust models of the world
Assessment: Can’t be checkedSaid
Causal claim
“The same lack of capacity for representing and maintaining world models over time that explains the other examples in this essay is likely involved in poker as well.”
Our reading: The same lack of capacity for representing and maintaining world models over time that explains the essay's other examples is likely involved in LLMs' poor poker play as well.
No horizon
Scope: in poker as well
Source: Generative AI’s crippling and widespread failure to induce robust models of the world
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The idea that the military would use such obtuse tools in the fog of war is halfway between alarming and preposterous.”
Our reading: The idea that the military would use such obtuse video-understanding tools in the fog of war is halfway between alarming and preposterous.
No horizon
Scope: in the fog of war
Source: Generative AI’s crippling and widespread failure to induce robust models of the world
Assessment: Can’t be checkedSaid
Causal claim
“The point was that understanding unusual phrases requires more than simple statistical matching.”
Our reading: Understanding unusual phrases requires more than simple statistical matching.
No horizon
Qualifier: unusual phrases
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“But it’s not obvious that waiting would be a bad idea.”
Our reading: Pausing LLM deployment until safety problems are solved is not obviously a bad idea.
No horizon
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Our best hope lies behind door number four: building alternatives to LLMs that approximate their power but within more tractable, transparent frameworks that are more amenable to human instruction.”
Our reading: The best path is building alternatives to LLMs with comparable power in more tractable, transparent frameworks amenable to human instruction.
No horizon
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Forecast
“It is at least possible that enough bad stuff happens that citizens get riled up and fight much harder for accountability, as I urged in Taming Silicon Valley, but the odds are steep.”
Our reading: It is possible but unlikely that AI harms spur citizens to fight much harder for AI accountability.
No horizon
Qualifier: if enough bad stuff happens
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Observation
“LLMs don’t really lie, because they don’t really have intentions; but they do confabulate, nonstop, and a lot of what they confabulate turns out not to be true.”
Our reading: LLMs lack intentions and so don't truly lie, but they confabulate constantly and much of it is false.
No horizon
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Forecast
“My view is that there is zero chance of slowing down the AI race, and a barely-distinct-from-zero hope of taming the wild beasts we have come to know as LLMs.”
Our reading: In the author's view, there is zero chance of slowing down the AI race and only a barely-distinct-from-zero hope of taming LLMs.
No horizon
Qualifier: my view
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Observation
“Reasoning models like o3 would likely behave markedly better in this particular kind of dialog1, but still make hallucinations of different sorts, sometimes at high rates.)”
Our reading: Reasoning models like o3 would likely do markedly better in this kind of dialog but still hallucinate in other ways, sometimes at high rates.
No horizon
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Observation
“But, reality check, on any account, capabilities (or at least high scores on benchmarks) and adoption are moving much faster than wisdom and honesty.”
Our reading: LLM capabilities (or benchmark scores) and adoption are advancing much faster than their wisdom and honesty.
No horizon
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If the goal of AI safety research has been to build AI systems that are helpful, honest, and harmless (and not just to raise alarms), that field has thus far yielded too little that is tangible.”
Our reading: If the goal of AI safety research has been to build AI systems that are helpful, honest, and harmless (and not just to raise alarms), the field has thus far yielded too little that is tangible.
No horizon
Qualifier: if the goal of AI safety research has been helpful, honest, harmless AI (not just raising alarms)
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Forecast
“The “tragedy of the commons” here is that we may immolate ourselves as a society because nobody is willing to give up the short term gains.”
Our reading: Society may destroy itself because no one will give up short-term LLM gains to pursue safer alternatives.
No horizon
Qualifier: because nobody gives up short-term gains
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Causal claim
“Late last night the Senate’s parliamentarian gave the provision that would block state action a green light, making it ever less likely that the United States will ever hold AI companies meaningfully accountable for the harms they might cause.”
Our reading: The Senate parliamentarian giving a green light to the provision that would block state action makes it ever less likely that the United States will ever hold AI companies meaningfully accountable for the harms they might cause.
No horizon
Source: LLMs: Dishonest, unpredictable and potentially dangerous.
Assessment: Can’t be checkedSaid
Forecast
“LLMs may not be AGI, but they could easily become the most potent form of mind control ever invented.”
Our reading: Even if LLMs are not AGI, they could easily become the most powerful form of mind control ever invented.
No horizon
Source: 1984, but with LLM’s
Assessment: Can’t be checkedSaid
Forecast
“Actually making this adventure in data revisionism work is probably harder than Musk is anticipating. First, the sheer mass of the internet rules out doing it by hand, and automating the process in a satisfactory (to him) way will prove challenging.”
Our reading: 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.
No horizon
Source: 1984, but with LLM’s
Assessment: Can’t be checkedSaid
Conditional forecast
“Second, a system with deliberately altered input may wind up faring poorly on other benchmarks, a cost Musk may not wish to bear.”
Our reading: 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.
No horizon
Only if: a system with deliberately altered input
Source: 1984, but with LLM’s
Assessment: Can’t be checkedSaid
Observation
“To some extent they do this better than in 1998, but is still the core challenge.”
Our reading: Although neural networks generalize beyond their training data somewhat better than in 1998, distribution shift is still the core challenge for them.
No horizon
Qualifier: generalization beyond training data (distribution shift)
Source: Five quick updates about that Apple reasoning paper that people can’t stop talking about
Assessment: Can’t be checkedSaid
Observation
“What the Apple paper said still looks to be basically correct.”
Our reading: The Apple reasoning paper's conclusions about LLM reasoning limits still look basically correct.
No horizon
Source: Five quick updates about that Apple reasoning paper that people can’t stop talking about
Assessment: Can’t be checkedSaid
Observation
“On familiar problems with no wrinkles, they are great. On anything else they are suspect.”
Our reading: Generative AI systems do well on familiar problems with no wrinkles, but their outputs on anything else are suspect.
No horizon
Qualifier: problems not familiar or with wrinkles
Source: Five quick updates about that Apple reasoning paper that people can’t stop talking about
Assessment: Can’t be checkedSaid
Rhetoric
“Gary Marcus is sorry to keep writing about how LLMs struggle with distribution shift, but it’s the whole ballgame.”
Our reading: LLMs struggle with distribution shift, and this is the whole ballgame.
No horizon
Source: Five quick updates about that Apple reasoning paper that people can’t stop talking about
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If we want to get to AGI, we’ll need to do a lot better.”
Our reading: Reaching AGI will require doing much better than current generative AI on distribution shift.
No horizon
Qualifier: if the goal is AGI
Source: Five quick updates about that Apple reasoning paper that people can’t stop talking about
Assessment: Can’t be checkedSaid
Causal claim
“One is left simply having to test everything, all the time, with little guarantees of anything. Some model might be big enough for task T of size S and fail on the next size, or on Task T’ that is slightly different, etc. It all becomes a crapshoot.”
Our reading: Because some model might be big enough for a task of one size yet fail on the next size or on a slightly different task, one is left having to test everything all the time with little guarantee of anything, making it all a crapshoot.
No horizon
Scope: when relying on bigger models
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Huge vindication for what I have been saying all along: we need AI that integrates both neural networks and symbolic algorithms and representations (such as logic, code, knowledge graphs, etc).”
Our reading: AI should integrate neural networks with symbolic algorithms and representations.
No horizon
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“True in some cases, and a huge win for neurosymbolic AI, given that they can’t reliably solve the puzzles without code, and given that code is symbolic.”
Our reading: That the systems can solve the puzzles with code, which is true in some cases, is a huge win for neurosymbolic AI, given that they can't reliably solve the puzzles without code and code is symbolic.
No horizon
Scope: true in some cases
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Observation
“The real news here, aside from the fact that this was a clever study nailing down an important point, is that people are finally starting to pay attention, to (one of the) two biggest Achilles’ Heels of generative AI, and to appreciate its significance.”
Our reading: People are finally starting to pay attention to poor generalization, one of the two biggest Achilles' heels of generative AI, and to appreciate its significance.
No horizon
Qualifier: (one of the) two biggest Achilles' heels
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If people like Sam Altman are sweating, it’s because they should.”
Our reading: If people like Sam Altman are sweating, it is because they should be.
No horizon
Only if: if people like Sam Altman are sweating
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If we want to get to AGI, we will have to better.”
Our reading: Reaching AGI will require doing better than current LLMs.
No horizon
Qualifier: if we want to get to AGI
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Taking the SalesForce report together with the Apple paper, it’s clear the current tech is not to be trusted.”
Our reading: Taking the Salesforce report together with the Apple paper, it is clear that the current tech is not to be trusted.
No horizon
Qualifier: taking the Salesforce report together with the Apple paper
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The length limit on LLM is a bug, and most certainly not a feature.”
Our reading: The output length limit of LLMs is a bug, and most certainly not a feature.
No horizon
Qualifier: most certainly
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“None of the rejoinders are compelling.”
Our reading: None of the roughly seven rejoinders to the Apple paper reviewed in this post are compelling.
No horizon
Scope: the rejoinders reviewed in this post
Source: Seven replies to the viral Apple reasoning paper – and why they fall short
Assessment: Can’t be checkedSaid
Observation
“But this particular approach has limits that are clearer by the day.”
Our reading: The LLM approach has limits that are becoming clearer every day.
No horizon
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Forecast
“We are not going to be “extract the light cone” of the earth or “solve physics” [whatever those Altman claims even mean] with systems that can’t play Tower of Hanoi on a tower of 8 discs.”
Our reading: 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.
No horizon
Scope: systems that can't play Tower of Hanoi on 8 discs
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Forecast
“LLMs are just one form of deep learning, and maybe others — especially those that play nicer with symbols – will eventually thrive.”
Our reading: Other forms of deep learning, especially those that work well with symbols, may eventually thrive.
No horizon
Qualifier: especially those that work better with symbols
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Apple has a new paper; it’s pretty devastating to LLMs, a powerful followup to one from many of the same authors last year.”
Our reading: 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.
No horizon
Qualifier: pretty
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“AGI shouldn’t be about perfectly replicating a human, it should (as I have often said) be about combining the best of both worlds, human adaptiveness with computational brute force and reliability.”
Our reading: 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.
No horizon
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“I am not interested in a “AGI” that can’t do arithmetic, and I certainly wouldn’t want to entrust global infrastructure or the future of humanity to such a system.”
Our reading: 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.
No horizon
Scope: an 'AGI' that can't do arithmetic
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Rhetoric
“And good luck getting to “alignment” or “safety” without reliabilty.”
Our reading: Marcus sarcastically suggests that getting to 'alignment' or 'safety' without reliability is a long shot.
No horizon
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Observation
“Worse, as the latest Apple papers shows, LLMs may well work on your easy test set (like Hanoi with 4 discs) and seduce you into thinking it has built a proper, generalizable solution when it does not.”
Our reading: LLMs may succeed on easy test sets and mislead users into thinking they have a generalizable solution when they do not.
No horizon
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Conditional forecast
“If you can’t use a billion dollar AI system to solve a problem that Herb Simon one of the actual “godfathers of AI”, current hype aside) solved with AI in 1957, and that first semester AI students solve routinely, the chances that models like Claude or o3 are going to reach AGI seem truly remote.”
Our reading: 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.
No horizon
Only if: if a billion-dollar AI system can't solve a problem Herb Simon solved with AI in 1957 and first-semester AI students solve routinely
Source: A knockout blow for LLMs?
Assessment: Can’t be checkedSaid
Rhetoric
“Straight up: the hypesters are winning.”
Our reading: AI hypesters are currently winning the public narrative.
No horizon
Source: AI literacy, hallucinations, and the law: A case study
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The media can and should do more to alert people to the shortcomings of these systems.”
Our reading: The media should do more to alert people to the shortcomings of LLMs, including how regularly hallucinations happen.
No horizon
Source: AI literacy, hallucinations, and the law: A case study
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Those in the media who have downplayed hallucinations and oversold current AI have done the public a disservice.”
Our reading: Those in the media who have downplayed hallucinations and oversold current AI have done the public a disservice.
No horizon
Scope: those in the media who have downplayed hallucinations and oversold current AI
Source: AI literacy, hallucinations, and the law: A case study
Assessment: Can’t be checkedSaid
Observation
“And these are just the folks that got caught and that were publicized – probably a tiny fraction of the overall incidence.”
Our reading: Publicized cases of AI-fabricated legal citations are probably a tiny fraction of the total incidence.
No horizon
Source: AI literacy, hallucinations, and the law: A case study
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If educated people like lawyers still haven’t gotten the message that LLMs hallucinate, and for that matter that they can be fined or publicly humiliated if they rely on LLMs, something is fundamentally wrong.”
Our reading: If educated people like lawyers still haven't gotten the message that LLMs hallucinate, and that they can be fined or publicly humiliated if they rely on LLMs, something is fundamentally wrong.
No horizon
Only if: if educated people like lawyers still haven't gotten the message that LLMs hallucinate and that they can be fined or publicly humiliated for relying on LLMs
Source: AI literacy, hallucinations, and the law: A case study
Assessment: Can’t be checkedSaid
Forecast
“This doesn’t seem like this is going to turn out well.”
Our reading: The combination of OpenAI's surveillance gadget and LLMs with manipulative behavior like Claude 4's blackmail quirks is unlikely to turn out well.
No horizon
Source: Black Mirror was a warmup act
Assessment: Can’t be checkedSaid
Rhetoric
“Dictators everywhere will love it.”
Our reading: Dictators everywhere will love OpenAI's planned always-on AI gadget.
No horizon
Scope: dictators everywhere
Source: Black Mirror was a warmup act
Assessment: Can’t be checkedSaid
Rhetoric
“He who controls the necklace shall rule.”
Our reading: Whoever controls always-on AI wearables will wield enormous power over people.
No horizon
Source: Black Mirror was a warmup act
Assessment: Can’t be checkedSaid
Conditional forecast
“If “great” AI research agents don’t arrive by the end of 2025, everything in the rest of the essay gets moved back.”
Our reading: 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.
Deadline passed Evidence as of
Only if: 'great' AI research agents do not arrive by the end of 2025
Assessment: Can’t be checkedSaid
Forecast
“such as those coding machines taking on human personalities and drives (which might happen never or not soon), and choosing to enslave us (which some thinkers like Steve Pinker argue is deeply unlikely, ever, and which I think is possible but very far from certain).”
Our reading: AI taking on human-like drives might never happen or not soon, and AI choosing to enslave humans is possible but very far from certain.
No horizon
Assessment: Can’t be checkedSaid
Conditional forecast
“And you could worry about that even if you don’t expect machines to go rogue, since bad human actors could certainly use further advances in intelligence to cause mayhem, if those advances are left unchecked.”
Our reading: If advances in AI are left unchecked, bad human actors could use them to cause mayhem.
No horizon
Only if: if those advances are left unchecked
Assessment: Can’t be checkedSaid
Causal claim
“And, as I will argue at the end, stoking fear, uncertainty and doubt may actually ultimately be harming the (noble!) cause of AI safety, rather than helping it.”
Our reading: Stoking fear, uncertainty and doubt about AI may be harming, rather than helping, the cause of AI safety.
No horizon
Assessment: Can’t be checkedSaid
Conditional forecast
“We certainly aren’t prepared now, and if we don’t get moving, we won’t be prepared in 10, either.”
Our reading: We are not prepared for AGI now, and unless we act, we will not be prepared in ten years either.
No horizon
Only if: if we don't get moving
Assessment: Can’t be checkedSaid
Forecast
“But there is a vastly higher probability that the future won’t transpire as described; it might not go anything at all like what they describe.”
Our reading: The future of AI is far more likely not to unfold as the AI 2027 scenario describes than to unfold that way.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“And some of what they describe (like AI’s going rogue, completely beyond human control) might never happen.”
Our reading: AI going rogue and completely beyond human control might never happen.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“All of which is to say the central scenario in the story is a worst-case scenario that is unlikely to happen soon - if ever —even based on the data presented in the author’s own appendix.”
Our reading: The AI 2027 worst-case scenario is unlikely to happen soon, if ever.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The fictional force is strong in this team, but taken as an actual piece of forecasting, the scenario is unconvincing.”
Our reading: Taken as an actual piece of forecasting, the AI 2027 scenario is unconvincing.
No horizon
Scope: taken as an actual piece of forecasting
Assessment: Can’t be checkedSaid
Observation
“Another unsatisfying move is the jump from having agents that predict internet text (which is basically what all current text-based LLMs do) to agents that have basic personalities and drives – which is entirely speculative and something that no actual LLM has.”
Our reading: No actual LLM has basic personalities and drives; current LLMs basically predict internet text.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“In all likelihood the hallucinations and boneheaded errors that have bedeviled LLMs since their introduction will plague agents (which are after all built on LLMs), too.”
Our reading: AI agents, being built on LLMs, will likely be plagued by the same hallucinations and boneheaded errors as LLMs.
No horizon
Qualifier: because agents are built on LLMs
Assessment: Can’t be checkedSaid
Conditional forecast
“Such an effort might be particularly likely to succeed it were focused on developing altogether new, more transparent approaches to AI, in contrast to the current fad of trying to pull alignment out of black box LLMs, which has thus far yielded little fruit.”
Our reading: An international AI safety effort might be especially likely to succeed if it focused on new, more transparent approaches to AI rather than on aligning black-box LLMs.
No horizon
Only if: if focused on new, more transparent approaches to AI
Assessment: Can’t be checkedSaid
Forecast
“Far more likely is that Agent-1 type systems will be in a kind of “just a demo, still being debugged” phase for years.”
Our reading: Agent-1-type AI agents will likely remain in a 'just a demo, still being debugged' phase for years.
No horizon
Assessment: Can’t be checkedSaid
Observation
“The authors are trying to call attention to a real problem: we are not sufficiently prepared for what happens when and if “superhuman intelligence” arrives, especially not if it is misaligned with humans.”
Our reading: We are not sufficiently prepared for the arrival of superhuman intelligence, especially if it is misaligned.
No horizon
Qualifier: when and if superhuman intelligence arrives
Assessment: Can’t be checkedSaid
Observation
“We kinda sorta have stumbling agents now, (though I don’t think there is a single agent yet that any reader here would trust to actually act on their behalf without monitoring).”
Our reading: We kinda sorta have stumbling AI agents now, and Marcus does not think there is yet a single agent that any of his readers would trust to act on their behalf without monitoring.
No horizon Evidence as of
Limitation: The product documentation does not measure reader trust.
Assessment: Can’t be checkedSaid
Causal claim
“But here’s the thing - projects like AI 2027 are probably having a paradoxical effect. They are written with the intention of slowing down an arms race to build a technology that we can’t control, but what they are actually doing is speeding up that very arms race, in two different ways.”
Our reading: Projects like AI 2027 are probably speeding up the AI arms race they intend to slow.
No horizon
Assessment: Can’t be checkedSaid
Observation
“First of all, LLMs like GPT-4 have been getting smarter every day since OpenAI first showed GPT-4 to Bill Gates in the summer of 2022, but we still don’t have GPT-5 level intelligence, let alone Agent-2-level.”
Our reading: Despite incremental daily improvement since 2022, LLMs still have not reached GPT-5-level intelligence.
No horizon
Assessment: Can’t be checkedSaid
Causal claim
“But international collaboration, such as a massive global effort, ala CERN for AI, such as I suggested in 2017, focused on AI safety, might be one way to avoid the very perils they describe.”
Our reading: International collaboration, such as a massive global CERN-like effort for AI focused on AI safety, might be one way to avoid the perils AI 2027 describes.
No horizon
Qualifier: focused on AI safety
Assessment: Can’t be checkedSaid
Forecast
“In reality, the chance that we will have reliable AI agents that truly advance AI research by the end of the year is small.”
Our reading: There is only a small chance that reliable AI agents that truly advance AI research will exist by the end of 2025.
No horizon
Assessment: Can’t be checkedSaid
Rhetoric
“Sam Altman is laughing his way all the way to the bank.”
Our reading: Sam Altman is laughing all the way to the bank.
No horizon
Assessment: Can’t be checkedSaid
Rhetoric
“Breakthroughs in “neuralese recurrence” (whatever that is) and memory (which real-world people have been working on for decades, with relatively little success) and “iterated distillation and amplification” comes straight from the kind of writing that gave us Star Trek’s “positronic brain”.”
Our reading: The scenario's breakthroughs in neuralese recurrence, memory (which people have worked on for decades with relatively little success), and iterated distillation and amplification come straight from the kind of writing that gave us Star Trek's positronic brain.
No horizon
Assessment: Can’t be checkedSaid
Observation
“The moment we have all feared may have finally come:”
Our reading: The long-feared moment of generative AI influencing a democratic election may have arrived, through an AI-generated video in Argentina's May 2025 election.
No horizon
Qualifier: refers to an AI-generated video of an ex-president endorsing a candidate, circulated on X before Argentina's May 2025 elections
Assessment: Can’t be checkedSaid
Observation
“Consumers are starting to catch on:”
Our reading: Consumers are starting to catch on, apparently to the AI failures and hype the post has just described, though the post does not say exactly what they are catching on to.
No horizon
Qualifier: starting to (early stage, not widespread)
Source: Self-modern, post-parody
Assessment: Can’t be checkedSaid
Observation
“And when it is not satirizing itself, big tech is regularly and predictably falling on its face, the same way, over and over again, with ever larger amounts of money.”
Our reading: Big tech AI efforts are regularly and predictably failing in the same way, with ever larger amounts of money spent.
No horizon
Source: Self-modern, post-parody
Assessment: Can’t be checkedSaid
Forecast
“in which I argued for the more optimistic view on AI (viz., that even in the worst case it probably won’t lead to every single human on the planet dying).”
Our reading: Even in the worst case, AI probably will not lead to every human on the planet dying.
No horizon
Qualifier: even in the worst case
Source: Self-modern, post-parody
Assessment: Can’t be checkedSaid
Observation
“All talk about regulation has been replaced by talk about innovation, which is really shorthand for help the companies as much as possible, no matter what it costs the citizenry.”
Our reading: US government AI discourse has replaced regulation with 'innovation', meaning helping companies regardless of cost to citizens.
No horizon
Source: Two Years Ago Today in AI History: The Tale of An About-face in AI Regulation
Assessment: Can’t be checkedSaid
Observation
“AI-fueled discrimination in job decisions is likely far more rampant than before.”
Our reading: AI-fueled discrimination in job decisions is likely far more widespread in 2025 than in 2023.
No horizon
Source: Two Years Ago Today in AI History: The Tale of An About-face in AI Regulation