Claims that can’t be checked, page 39 of 40

Claims by Gary Marcus and Ed Zitron, as of 2026-11-11.

Claims that can’t be checked (3,902)

Opinions, rhetoric, hedged or vague claims, and things no public evidence could show. Listed, not scored.

Claims 3,801 to 3,900 of 3,902, newest first. Page 39 of 40.

Assessment: Can’t be checkedSaid 2022-09-29
Gary Marcus
“The fleets of driverless taxis he promised are likely still years or decades away.”
Assessment: Can’t be checkedSaid 2022-09-29
Gary Marcus
“None of this is to say that Tesla could not solve all of these problems eventually given enough time and money, but Friday may not be the big demo his fans are hoping for.”
Assessment: Can’t be checkedSaid 2022-09-29
Gary Marcus
“In my view the Google demo is pretty problematic (despite being cool), particularly in terms of likely reliability, but no matter.”
Assessment: Can’t be checkedSaid 2022-09-25
Gary Marcus
“A large part of LeCun’s new manifesto is a well-motivated call for incorporating a “configurable predictive world model” into deep learning.”
Assessment: Can’t be checkedSaid 2022-09-25
Gary Marcus
“I won’t accuse LeCun of plagiarism, because I think he probably reached these conclusions honestly, after recognizing the failures of current architectures.”
Assessment: Can’t be checkedSaid 2022-09-23
Gary Marcus
“It may be boring to hear about the same challenges over and over, but it’s hard to see how we will get to trustworthy general intelligence until these problems are resolved, in ways that are robust, rather than merely statistically contingent.”
Assessment: Can’t be checkedSaid 2022-09-23
Gary Marcus
“Those who wish that AI criticism would go away are welcome to their views, but misrepresenting critics and their views (and obscuring the degree to which even mainstream researchers agree) won’t magically solve the problems AI has yet to solve.”
Assessment: Can’t be checkedSaid 2022-09-18
Gary Marcus
“In assessing progress towards general intelligence, the critical question should be, how much do systems like Dall-E, Imagen, Midjourney, and Stable Diffusion really understand the world, such that they can reason on and act on that knowledge?”
Assessment: Can’t be checkedSaid 2022-09-14
Gary Marcus
“(Dare I make obvious guess? They probably tried and failed, and don’t want to admit it.)”
Assessment: Can’t be checkedSaid 2022-09-14
Gary Marcus
“For me, compositionality has always been the touchstone for AI; when I say AI has hit a wall, it’s compositionality, more than anything else, that I have been talking about:”
Assessment: Can’t be checkedSaid 2022-09-14
Gary Marcus
“A will to believe in AI will never replace the need for careful science.”
Assessment: Can’t be checkedSaid 2022-09-14
Gary Marcus
“Having a Silicon Valley celebrity like Alexander weigh in for Google is great PR for Google, but not so great for the public understanding of AI.”
Assessment: Can’t be checkedSaid 2022-09-14
Gary Marcus
“In 2017, a team of Google researchers invented the Transformer, arguably the biggest advance in AI in the last 10 years.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“Without a much greater degree of interpretability and verifiability, its hard to see how we can get to that level of safety.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“Misunderstandings will arise; without really firm checks on these systems, some of those misunderstandings could lead to mayhem.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“It’s not just that large language models can counsel suicide, or sign off on genocide, or that they can be toxic and that they are incredibly (over)sensitive to the details of their training set - it’s that if you put them inside a robot, and they misunderstand you, or fail to fully appreciate the implications of your request, they could cause major damage.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“And the more actions a robot might consider, the worse performance is likely to be.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“Putting aside robots for the moment, we already know that so-called large language models are like bulls in a china shop: awesome, powerful, and reckless.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“First, the language technology that the new system relies on is well-known to be problematic and second it is likely to be even more problematic in the context of robots.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“In a general-purpose, humanoid domestic robot, 75% is nowhere near good enough.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“The PaLM-SayCam experiment makes it pretty clear that the list of 21 problems is due for an update.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“As I have argued throughout the essays in this Substack (eg. Three ideas from linguistics that everyone in AI should know and The New Science of Alt Intelligence) , that’s because large language models are superficial statistical mimics, rather than systems that traffic in rich cognitive models of the world around them.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“The reality is that the field has no compelling idea at the current time how to solve the many “alignment” problems that bedevil large language models.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“But that precisely what Google’s new system is, a stitching-together of superficial and incorrigible language understanders with powerful and potentially dangerous humanoid robots.”
Assessment: Can’t be checkedSaid 2022-08-18
Gary Marcus
“I am not surprised that this can (kinda sorta) be done, but I am not sure it should be done.”
Assessment: Can’t be checkedSaid 2022-07-20
Gary Marcus
“And it is worth noting that even if they did work, there would be, of course, no guarantee that they would work in a fashion that was similar to humans; AlphaGo plays a mean game of Go but that doesn’t mean it gives insight into how human players learn Go.)”
Assessment: Can’t be checkedSaid 2022-07-20
Gary Marcus
“If we want to make progress, we need to start by recognizing the enormity of the problem.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“Ellipsis, comparative and pronouns are routine, everyday components of language, readily acquired by children. DALLE-mini does not seem to grasp them.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“But we see no reason to think that current large language models encode stable representations of the world at all. All they are doing, literally, is predicting next words in sentences.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“Moving beyond prediction into reference, particularly in a system that is inherently uninterpretable, is an immense leap that (a) cannot be taken for granted and (b) should be a focus of research.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“We would suggest that all the attention that has been placed on predicting sequences of words has come at the expense of developing a theory of how such processes might culminate in cognitive models of the world.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“Rather, it means identify the second of the balls that are blue, regardless of which position; systems like GPT-3 likely have trouble with this sort of thing, as Ernie Davis just reported to us in an email.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“Considerable efforts are being made towards making “cross-modal” models like Gato that take in visual input as well strings of words; we think this is well worth investigating but unlikely in itself to suffice.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“For that, we may yet need another revolution.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“It may produce both a blue cube and a red cube, but as likely as not, the red cube winds up on top of the blue cube.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“Current AI systems don’t really show up for any of that.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“Yet even more wild is the widespread belief that language models lacking the major foundational components of human language (e.g., bounded compositionality mixed with unbounded generativity) will have any hope of meaningfully interfacing with audio/visual models in a way that accurately reflects human conceptual structure.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“A good language understanding (or language production) system almost certainly needs to have some large-scale learning as a vital component.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“But if all a system is doing is predicting the next word in what someone says, its understanding is inevitably superficial.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“In our view, large language models are basically sidestepping this entire set of issues, and to their peril.”
Assessment: Can’t be checkedSaid 2022-06-22
Gary Marcus
“Ultimately, the right approach to large-scale learning may not rest on machinery that is finely geared to predicting sequences of words, but rather on developing a new approach to machine learning that is fundamentally geared towards learning how hierarchically structured sets of words map onto meanings that are used in the service of updating cognitive models.”
Assessment: Can’t be checkedSaid 2022-06-17
Gary Marcus
“A couple weeks ago I got so fed up with so many smart people spending so much time experimenting with GPT-3 and DALL-E and so forth (when I, personally, think that there are deeper problems to be solved), that I wrote a poem.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“For example, we lack data on whether drivers use such systems as directed, and need to enforce vigorous mechanisms to make sure drivers don’t tune out.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“Clearly, the number of incidents would go up, as the number of vehicles went up, and would also go up significantly as the systems were more regularly challenged with more difficult driving conditions (bad weather, more erratic pedestrians, etc), and likely jump radically if we took humans out of the loop altogether.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“Is that good, or bad? I don’t know it, but it’s enough to be concerning. Why? Because .75 incidents a day comes from a very limited trial of computer-assisted driving relative to the ambition of full-self-driving.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“My informal read (still lacking a tremendous amount of information that I as a scientist would really like to see) is that neither Tesla nor Waymo is anywhere close to something we could use every day, all the time, wherever one wanted to go.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“The best argument that Tesla might not survive if its massive bet on self-driving fails comes from Elon Musk himself.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“It’s fantastic that NHTSA is now pushing manufacturers to release incident data, and but we need soooo much more data (eg about miles driven) if we are to look at the money question—how safe is all this, really—in a truly scientific way.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“But our experiments in building them need to be strictly regulated, especially when nonconsenting human beings become unwilling participants. And our requirements on data transparency need to be vastly stricter.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“Does that mean that society should give up on building driverless cars? Absolutely not.”
Assessment: Can’t be checkedSaid 2022-06-15
Gary Marcus
“This morning’s fresh data dump from NHTSA should be a bit of a wakeup call to those who have gone long on Tesla, if they are thinking that a solution to Full Self Driving is in sight. The data aren’t devastating, but they aren’t so promising either, and they raise of lot of questions.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“In truth, literally everything that the system says is bullshit.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“Fooling people into thinking a program is intelligent is just not the same as building programs that actually are intelligent.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“Neither LaMDA nor any of its cousins (GPT-3) are remotely intelligent.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“All they do is match patterns, draw from massive statistical databases of human language. The patterns might be cool, but language these systems utter doesn’t actually mean anything at all.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“But there is no absolutely no reason whatever for us to waste time wondering whether anything anyone in 2022 knows how to build is sentient. It is not.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“None of the software in either of those systems has survived in modern efforts at “artificial general intelligence”, and I am not sure that LaMDA and its cousins will play any important role in the future of AI, either.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“What these systems do, no more and no less, is to put together sequences of words, but without any coherent understanding of the world behind them, like foreign language Scrabble players who use English words as point-scoring tools, without any clue about what that mean.”
Assessment: Can’t be checkedSaid 2022-06-12
Gary Marcus
“To be sentient is to be aware of yourself in the world; LaMDA simply isn’t.”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“In my view, no way; I think the actual odds are less than 20%.”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“While there’s no denying that the two of us have often clashed, on this point—the point about scale alone not likely to be enough, and about the need for some new discovery (i.e., paradigm shifts)—we are actually in complete agreement.”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“GPT-2 is both a triumph for empiricism, and, in light of the massive resources of data and computation that have been poured into them, a clear sign that it is time to consider investing in different approaches.”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“I say that we have given the scaling hypothesis a really good look (with a larger budget than all but a handful of projects in history), and as such the failures of large scale systems are evidence (not proof) that we ought to seriously consider alternatives, e.g., here:”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“That said, I, of course, could be right about foreseeing the need for a paradigm shift, but wrong about what that paradigm shift turns out to be.”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“That’s why we need new approaches, exactly as I have been arguing.”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“Personally I’d put the probability that we need to shift at 90%, well above the 60% that Alexander suggests, and I would put the probability that we will need to embrace symbol-manipulation as part of the mix at 80%, more than double Alexander’s 34%.”
Assessment: Can’t be checkedSaid 2022-06-11
Gary Marcus
“Personally I’d put the probability that we need to shift at 90%, well above the 60% that Alexander suggests, and I would put the probability that we will need to embrace symbol-manipulation as part of the mix at 80%, more than double Alexander’s 34%.”
Assessment: Can’t be checkedSaid 2022-06-07
Gary Marcus
“Scott’s more marked success on six questions may have been a fluke.”
Assessment: Can’t be checkedSaid 2022-06-07
Gary Marcus
“The core problem that follows from that has always been that these systems aren’t stable or trustworthy; a crumbling house follows from a rotten foundation.”
Assessment: Can’t be checkedSaid 2022-06-07
Gary Marcus
“I think all that’s scandalous.”
Assessment: Can’t be checkedSaid 2022-06-07
Gary Marcus
“But I see no reason whatsoever to think that the underlying problem — a lack of cognitive models of the world —have been remedied.”
Assessment: Can’t be checkedSaid 2022-06-07
Gary Marcus
“Either way, though, there is still no real physical, social, or psychological understanding, just richer parroting of larger corpora (and some reinforcement learning to keep the system more on check).”
Assessment: Can’t be checkedSaid 2022-06-07
Gary Marcus
“Maybe we will find fewer absolute howlers per attempt than we did with GPT-3, but it still won’t be so proficient as to have developed a genuine causal model of the world.”
Assessment: Can’t be checkedSaid 2022-06-07
Gary Marcus
“And yes, there has been improvement eg from GPT-2 to GPT-3—but zero qualitative change.”
Assessment: Can’t be checkedSaid 2022-05-31
Gary Marcus
“Four years later, many people continue to hope that deep learning will be a panacea; that still seems pretty unrealistic to me; I still think we need a whole lot more techniques.”
Assessment: Can’t be checkedSaid 2022-05-31
Gary Marcus
“Seven years is a long time, but the field is going to need to invest in other ideas if we are going to get to AGI before the end of the decade. Or else outliers alone might be enough to keep us from getting there.”
Assessment: Can’t be checkedSaid 2022-05-31
Gary Marcus
“Unexpected circumstances have been and continue to be the bane of contemporary AI techniques, and probably will be until there is a real revolution.”
Assessment: Can’t be checkedSaid 2022-05-31
Gary Marcus
“So far as I know your cars are still mostly just doing a kind of glorified hash-table-like lookup (viz deep learning), and hence still having a lot of problems with unexpected situations.”
Assessment: Can’t be checkedSaid 2022-05-31
Gary Marcus
“Being realistic, 7 years might well not be enough to (a) invent those tools (if they are don’t already exist) and (b) take them from the lab into production.”
Assessment: Can’t be checkedSaid 2022-05-31
Gary Marcus
“Things like programming verification might eventually help, but again, in deep learning we don’t yet have the tools for writing verifiable code. If you are to win the bet, we are probably going to have to solve that problem too, and soon.”
Assessment: Can’t be checkedSaid 2022-05-28
Gary Marcus
“Against a long history of overclaiming in AI, such strong statements demand an awful lot of evidence. Because Google won’t allow skeptics to try the system, we can only make guesses. But from the few clues that have filtered out, there is enormous reason for skepticism.”
Assessment: Can’t be checkedSaid 2022-05-28
Gary Marcus
“it casts doubt on “an astronaut riding a horse”, and suggests success there is not because the network knows how to extract meanings from individual words”
Assessment: Can’t be checkedSaid 2022-05-28
Gary Marcus
“I am pretty sure this kind of failure would be easy to replicate.”
Assessment: Can’t be checkedSaid 2022-05-28
Gary Marcus
“The answer to the second question—does Imagen have a deep understanding of language—is almost certainly “no”: if a system can draw a given picture with incantation X, but not with perfectly sensible alternative incantation Y, it’s pretty hard to say that the system understands Y.”
Assessment: Can’t be checkedSaid 2022-05-28
Gary Marcus
“DALL-E 2 is, I dare say, not as smart as it seems.”
Assessment: Can’t be checkedSaid 2022-05-28
Gary Marcus
“Judging by all the premature declarations I have seen from the machine learning community of late, I am not sure AI has yet fully learned it.”
Assessment: Can’t be checkedSaid 2022-05-21
Gary Marcus
“A new breed of AI systems, with a firmer innate basis, closer to the Kantian model, but out of step with current Zeitgeist, might eventually tell us more.”
Assessment: Can’t be checkedSaid 2022-05-21
Gary Marcus
“Human minds try to relate their language to what they know about the world; large language models don’t. Which means the latter just aren’t telling us much about the former.”
Assessment: Can’t be checkedSaid 2022-05-21
Gary Marcus
“Large language models can, parasitically, glom onto human data, but they could glom onto any other data set just as well (e.g. programming languages); as such, they tell us far too little about how humans have come to be what they are.”
Assessment: Can’t be checkedSaid 2022-05-21
Gary Marcus
“Systems like GPT-3 don’t tell us why human languages have the special character that they do. As such, there is little explanatory value.”
Assessment: Can’t be checkedSaid 2022-05-21
Gary Marcus
“Humans know these things; if AI’s are to understand language, they should refract language through this sort basic understanding of common sense, rather than needlessly starting as blank slates and acquiring statistics without concepts.”
Assessment: Can’t be checkedSaid 2022-05-21
Gary Marcus
“They are an embarrassment to Locke’s conjecture, and a strong argument for exploring their very antithesis, which would be Kantian systems that start with representations of space and time and causality, and at least some kernels of universal grammar.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“I don’t think any pure deep learning/big data system can match it.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“We are in early days in AI.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“But anyone who is candid will recognize that these kinds of errors reveal that something is, for now, deeply amiss.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“Let us all encourage a field that is open-minded enough to work in multiple directions, without prematurely dismissing ideas that happen to be not yet fully developed.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“It may just be that the best path to artificial (general) intelligence isn’t through Alt Intelligence, after all.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“A large fraction of current AI researchers, or at least those currently in power, don’t (so far as I can tell) give a damn about any of this.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“AI should certainly not be a slavish replica of human intelligence (which after all is flawed in its own ways, saddled with lousy memory and cognitive bias). But it should look to human (and animal cognition) for clues.”
Assessment: Can’t be checkedSaid 2022-05-14
Gary Marcus
“It may well be better than anything else we currently have, but the fact that it still doesn’t really work, even after all the immense investments that have been made in it, should give us pause.”