From what I've heard, Chomsky had a stroke which impacted his language. You will, unfortunately, not hear a recent opinion from him on current developments.
He's currently 96, and I started noticing in his 2022-2023 interviews that he seemed to lose a bit of the spark, so while it is always sad and moving to see such things, I don't know how much Chomsky left we were getting anyway.
Manufactured intelligence to modulate a world of manufactured consent!
I agree with the rest of these comments though, listening to Chomsky wax about the topic-du-jour is a bit like trying to take lecture notes from the Swedish Chef.
It's shocking how people are putting him down for the OP interview with just a couple of questions in 2023. The dude was 94 years old. I also did not predict where we would be in 2025 with LLMs. And neither did you. (When I say you, I mean some of the other commenters.)
Are we seriously saying that his ideas are not taken seriously? his theory of grammar/language construction was a major contributor to modern programming languages, for one.
Chomsky is always saying that LLMs and such can only imitate, not understand language. But I wonder if there is a degree of sophistication at which he would concede these machines exceed "imitation". If his point is that LLMs arrive at language in a way different than humans... great. But I'm not sure how he can argue that some kind of extremely sophisticated understanding of natural language is not embedded in these models in a way that, at this point, exceeds the average human. In all fairness, this was written in 2023, but given his longstanding stubbornness on this topic, I doubt it would make a difference.
I think that misses the point entirely. Even if you constructed some system the output of which could not be distinguished from human-produced language but that either (1) clearly operated according to principles other than those that govern human language or (2) operated according to principles that its creators could not adequately explain, it would not be of that much interest to him.
He wants to understand how human language works. If I get him right — and I'm absolutely sure that I don't in important ways — then LLMs are not that interesting because both (1) and (2) above are true of them.
I think what would "convince" Chomsky is more akin to the explainability research currently in it's infancy, producing something akin to a branch of information theory for language and thought.
Chomsky talks about how the current approach can't tell you about what humans are doing, only approximate it; the example he has given in the past is taking thousands of hours of footage of falling leaves and then training a model to make new leaf falling footage versus producing a model of gravity, gas mechanics for the air currents, and air resistance model of leaves. The later representation is distilled down into something that tells you about what is happening at the end of some scientific inquiry, and the former is a opaque simulation for engineering purposes if all you wanted was more leaf falling footage.
So I interpret Chomsky as meaning "Look, these things can be great for an engineering purpose but I am unsatisfied in them for scientific research because they do not explain language to me" and mostly pushing back against people implying that the field he dedicated much of his life to is obsolete because it isn't being used for engineering new systems anymore, which was never his goal.
That's not quite a valid point considering the article's conclusion: sowing dissent in the sciences allows companies to more easily package and sell carcinogens like asbestos, lead paint, and tobacco products.
I understand his diction is a bit impenetrable but I believe the intention is to promote literacy and specificity, not just to be a smarty-pants.
I guess it's because LLM does not understand the meaning as you understand what you read or thought. LLMs are machines that modulate hierarchical positions, ordering the placement of a-signifying sign without a clue of the meaning of what they ordered (that's why machine can hallucinate :they don't have a sense of what they express)
Quite a nice overview. For almost any specific measure, you can find something that is better than human at that point. And now LLMs architecture have made possible for computers to produce complete and internally consistent paragraphs of text, by rehashing all the digital data that can be found on the internet.
But what we're good as using all of our capabilities to transform the world around us according to an internal model that is partially shared between individuals. And we have complete control over that internal model, diverging from reality and converging towards it on whims.
So we can't produce and manipulate text faster, but rarely the end game is to produce and manipulate text. Mostly it's about sharing ideas and facts (aka internal models) and the control is ultimately what matters. It can help us, just like a calculator can help us solve an equation.
EDIT
After learning to draw, I have that internal model that I switch to whenever I want to sketch something. It's like a special mode of observation, where you no longer simply see, but pickup a lot of extra details according to all the drawing rules you internalized. There's not a lot, they're just intrinsically connected with each other. The difficult part is hand-eye coordination and analyzing the divergences between what you see and the internal model.
I think that's why a lot of artists are disgusted with AI generators. There's no internal models. Trying to extract one from a generated picture is a futile exercice. Same with generated texts. Alterations from the common understanding follows no patterns.
> It can help us, just like a calculator can help us solve an equation.
A calculator is consistent and doesn’t “hallucinate” answers to equations. An LLM puts an untrustworthy filter between the truth and the person. Google was revolutionary because it increased access to information. LLMs only obscure that access, while pretending to be something more.
I'm not a native english speaker, so I've used for an essay where they told us to target a certain word count. I was close, but the verbiage to get to that word count doesn't come naturally to me. So I used Germini and tell it to rewrite the text targeting that word count (my only prompt). Then I reviewed the answer, rewriting where it strayed from the points I was making.
Also I used it for a few programming tasks I was pretty sure was in the datasets (how to draw charts with python and manipulate pandas frame). I know the domain, but wasn't in the mood to analyse the docs to get the implementation information. But the information I was seeking was just a few lines of sample code. In my experience, anything longer is pretty inconsistent and worthless explanations.
> Perhaps it is more important to know the limitations of tools rather than dismiss their utility entirely due to the existence of limitations.
Well, yes. And "reasoning" is only something LLMs do coincidentally, to their function as sequence continuation engines. Like performing accurate math on rationale numbers, it can happen if you put in a lot of work and accept a LOT of expensive computation. Even then there exists computations that just are not reasonable or feasible.
Reminding folks to dismiss the massive propaganda engine pushing this bubble isn't "dismissing their utility entirely".
These are not reasoning machines. Treating them like they are will get you hurt eventually.
My point is that computers, when used properly, can absolutely do math. And LLMs, when used properly, can absolutely explain the reasoning behind why a pound of bricks and a pound of feathers weigh the same.
Can they reason? Maybe, depending on your definition of reasoning.
An example: which weighs more a pound of bricks and 453.59 grams of feathers? Explain your reasoning.
LLM: The pound of bricks weighs slightly more.
*Reasoning:*
* *1 pound* is officially defined as *0.45359237 kilograms*, which is *453.59237 grams*.
* You have *453.59 grams* of feathers.
So, the pound of bricks (453.59237 grams) weighs a tiny fraction more than the 453.59 grams of feathers. For most practical purposes, they'd be considered the same, but technically, the bricks are heavier by 0.00237 grams. /llm
It is both correct and the reasoning is sound. Do I understand that the machine is a pattern following machine, yes! Is there an argument to be made that humans are also that? Probably. Chomsky himself argued in favor of a universal grammar, after all.
I’m steel manning this a bit, but the point is that LLMs are capable of doing some things which are indistinguishable from human reasoning in terms of results. Does the process matter in all cases?
"Does how the reasoning work matter in all cases"
Pretty Obviously no, but it may matter in some of them. We also don't really understand which ones yet.
"Does the reasoning work as intended in all cases?"
Pretty Obviously no, but it doesn't work for at least some of them. We also don't really understand which ones yet.
"We also don't really understand which ones yet"
Is the critical point of caution.
I could retort with another gotcha argument, but instead of doing that perhaps we can do better than that?
An attempt: They are bad reasoning machines that already are useful in a few domains and they're improving faster than evolutionary speeds. So even if they're not useful today in a domain relevant to you there's a significant possibility they might be in a few months.
AlphaEvolve would have been scifi a decade ago.
"It's like if a squirrel started playing chess and instead of "holy shit this squirrel can play chess!" most people responded with "But his elo rating sucks""
No. The marginal cost of an LLM is much, much lower than a high schooler. It is not even close. There is a lot of investment happening but revenue will continue to increase as the product improves and more use it or the money will stop flowing. If training stopped LLMs would be immensely profitable right now
Surely it just reasoned that you made a typo and "autocorrected" your riddle. Isn't this what a human would do? Though to be fair, a human would ask you again to make sure they heard you correctly. But it would be kind of annoying if you had to verify every typo when using an LLM.
I shortened it. I think this is one instance where it's actually relevant. The "opinion" of an LLM about a LLM criticism from one the leading linguists in history.
It got flagged, but I feel the flagging was knee-jerk and failed to understand the irony in the context.
> It’s as if a biologist were to say: “I have a great new theory of organisms. It lists many that exist and many that can’t possibly exist, and I can tell you nothing about the distinction.”
> Again, we’d laugh. Or should.
Should we? This reminds me acutely of imaginary numbers. They are a great theory of numbers that can list many numbers that do 'exist' and many that can't possibly 'exist'. And we did laugh when imaginary numbers were first introduced - the name itself was intended as a derogatory term for the concept. But who's laughing now?
This is the point where i realized he has no clue what he is saying. Theres so many creatures that once existed that can never again exist on earth due to the changes that the planet has gone through over millions, billions of years. The oxygen rich atmosphere that supported the dinosaurs for instance. If we had some kind of system that can put together proper working DNA for all the creatures that ever actually existed on this planet, some half of them would be completely nonviable if introduced to the ecosystem today. He is failing to see that there is an incredible understanding of systems that we are producing with this work, but he is a very old man from a very different time and contrarianism is often the only way to look smart or reasoned when you have no clue whats actually going on, so I am not shocked by his take.
Imaginary numbers are not relevant at all. There’s nothing whatsoever to do with the everyday use of the word imaginary. They could just as easily have been called “vertical numbers” and real numbers called “horizontal numbers” in order to more clearly illustrate their geometric interpretation in the complex plane.
The term “imaginary number” was coined by Rene Descartes as a derogatory and the ill intent behind his term has stuck ever since. I suspect his purpose was theological rather than mathematical and we are all the worse for it.
Insect behaviour. Flight of birds. Turtle navigation. A footballer crossing the field to intercept a football.
This is what Chomsky always wanted ai to be... especially language ai. Clever solutions to complex problems. Simple once you know how they work. Elegant.
I sympathize. I'm a curious human. We like elegant, simple revelations that reveal how out complex world is really simple once you know it's secrets. This aesthetic has also been productive.
And yet... maybe some things are complicated. Maybe LLMs do teach us something about language... that language is complicated.
So sure. You can certainly critique "ai blogosphere" for exuberance and big speculative claims. That part is true. Otoh... linguistics is one of the areas that ai based research may turn up some new insights.
> Maybe LLMs do teach us something about language... that language is complicated.
It certainly teaches us many things. But an LLM trained on as many words (or generally speaking an AI trained on sounds) in similar quantities of a toddler learning to understand, parse and apply language, would not perform well with current architectures. They need orders of magnitude more training material to get even close. Basically, current AI learns slowly, but of course it’s much faster in wall clock time because it’s all computer.
What I mean is: what makes an ALU (CPU) better than a human at arithmetic? It’s just faster and makes fewer errors. Similarly, what makes Google or Wikipedia better than an educated person? It’s just storing and helping you access stored information, it’s not magic (anymore). You can manually do everything mechanically, if you’re willing to waste the time to prove a point.
An LLM does many things better than humans, but we forget they’ve been trained on all written history and have hundreds of billions of parameters. If you compare what an LLM can do with the same amount of training to a human, the human is much better even at picking up patterns – current AIs strongest skill. The magic comes from the unseen vast amounts of training data. This is obvious when using them – stray just slightly outside of the training zone to unfamiliar domains and ”ability” drops rapidly. The hard part is figuring out these fuzzy boundaries. How far does interpolating training data get you? What are the highest level patterns are encoded in the training data? And most importantly, to what extent do those patterns apply to novel domains?
Alternatively, you can use LLMs as a proxy for understanding the relationship between domains, instead of letting humans label them and decide the taxonomy. One such example is the relationship between detecting patterns and generating text and images – it turns out to be more or less reversible through the same architecture. More such remarkable similarities and anti-similarities are certainly on the horizon. For instance, my gut feeling says that small talk is closer to driving a car but very different from puzzle solving. We don’t really have a (good) taxonomy over human- or animal brain processes.
Chomsky’s notion is: LLMs can only imitate, not understand language. But what exactly is understanding? What if our „understanding“ is just unlocking another level in a model? Unlocking a new form of generation?
Understanding is probably not much more than making abstractions into simpler terms until you are left with something one can relate to by intuition or social consensus.
I have trouble with the notion "understanding". I get the usefulness of the word, but I don't think that we are capable to actually understand. I also think that we are not even able to test for understanding - a good imitation is as good as understanding. Also, understanding has limits. In school, they often say on class that you should forget whatever you have been taught so far, because this new layer of knowledge that they are about to teach you. Was the previous knowledge not "understanding" then? Is the new one "understanding"?
If we define "understanding" like "useful", as in, not an innate attribute, but something in relation to a goal, then again, a good imitation, or a rudimentary model can get very far. ChatGPT "understood" a lot of things I have thrown at it, be that algorithms, nutrition, basic calculations, transformation between text formats, where I'm stuck in my personal development journey, or how to politely address people in the email I'm about to write.
>What if our „understanding“ is just unlocking another level in a model?
I believe that it is - that understanding is basically an illusion. Impressions are made up from perceptions and thinking, and extrapolated over the unknown. And just look how far that got us!
He alludes to quite a bit here - impossible languages, intrinsic rules that don’t actually express in the language, etc - that leads me to believe there’s a pretty specific sense by which he means “understanding,” and I’d expect there’s a decent literature in linguistics covering what he’s referring to. If it’s a topic of interest to you, chasing down some of those leads might be a good start.
(I’ll note as several others have here too that most of his language seems to be using specific linguistics terms of art - “language” for “human language” is a big tell, as is the focus on understanding the mechanisms of language and how humans understand and generate languages - I’m not sure the critique here is specifically around LLMs, but more around their ability to teach us things about how humans understand language.)
Actually no. Chomsky has never really given a stuff about Chinese Room style arguments about whether computers can “really” understand language. His problem with LLMs (if they are presented as a contribution to linguistic science) is primarily that they don’t advance our understanding of the human capacity for language. The main reasons for this are that (i) they are able to learn languages that are very much unlike human languages and (ii) they require vastly more linguistic data than human children have access to.
> The world’s preeminent linguist Noam Chomsky, and one of the most esteemed public intellectuals of all time, whose intellectual stature has been compared to that of Galileo, Newton, and Descartes, tackles these nagging questions in the interview that follows.
In all seriousness tho, not much of anything he says is taken seriously in an academic sense any more. Univeral Grammar, Minimalism, etc. He's a very petty dude. The reason he doesn't engage with GPT is because it suggests that linguistic learning is unlike a theory he spent his whole life [unsuccessfully] promoting, but he's such a haughty know-it-all, that I guess dummies take that for intelligence? It strikes me as not dissimilar to Trump in a way, where arrogance is conflated with strength, intelligence, etc. Fake it til you make it, or like, forever, I guess.
The comparison to Trump seems very unfair. I'm not in the academy and didn't know the current standing of his work, but he was certainly a big name that popped up everywhere (as a theorists in the field, not as a general celebrity) when I took an introduction to linguistics 20+ years ago.
As this is Hacker News, it is worth mentioning that he developed the concept of context-free grammars. That is something many of us encounter on a regular basis.
No matter what personality flaws he might have and how misguided some of his political ideas might be, he is one of the big thinkers of the 20th century. Very much unlike Trump.
"Expert in (now-)ancient arts draws strange conclusion using questionable logic" is the most generous description I can muster.
Quoting Chomsky:
> These considerations bring up a minor problem with the current LLM enthusiasm: its total absurdity, as in the hypothetical cases where we recognize it at once. But there are much more serious problems than absurdity.
> One is that the LLM systems are designed in such a way that they cannot tell us anything about language, learning, or other aspects of cognition, a matter of principle, irremediable... The reason is elementary: The systems work just as well with impossible languages that infants cannot acquire as with those they acquire quickly and virtually reflexively.
Response from o3:
LLMs do surface real linguistic structure:
• Hidden syntax: Attention heads in GPT-style models line up with dependency trees and phrase boundaries—even though no parser labels were ever provided. Researchers have used these heads to recover grammars for dozens of languages.
• Typology signals: In multilingual models, languages that share word-order or morphology cluster together in embedding space, letting linguists spot family relationships and outliers automatically.
• Limits shown by contrast tests: When you feed them “impossible” languages (e.g., mirror-order or random-agreement versions of English), perplexity explodes and structure heads disappear—evidence that the models do encode natural-language constraints.
• Psycholinguistic fit: The probability spikes LLMs assign to next-words predict human reading-time slow-downs (garden-paths, agreement attraction, etc.) almost as well as classic hand-built models.
These empirical hooks are already informing syntax, acquisition, and typology research—hardly “nothing to say about language.”
It's completely irrelevant because the point he's making is that LLMs operate differently from human languages as evidenced by the fact that they can learn language structures that humans cannot learn. Put another way, I'm sure you can point out an infinitude of similarities between human language faculty and LLMs but it's the critical differences that make LLMs not useful models of human language ability.
> When you feed them “impossible” languages (e.g., mirror-order or random-agreement versions of English), perplexity explodes and structure heads disappear—evidence that the models do encode natural-language constraints.
This is confused. You can pre-train an LLM on English or an impossible language and they do equally well. On the other hand humans can't do that, ergo LLMs aren't useful models of human language because they lack this critical distinctive feature.
Yes it's true, you can read my response to one of the authors @canjobear describing the problem with that paper in the comment linked below. But to summarize: in order to show what they want to show they have to take the simple, interesting languages based on linear order that Moro showed a human cannot learn and show that LLMs also can't learn them and they don't do that.
The reason the Moro languages are of interest are that they are computationally simple so it's a puzzle why humans can't learn them (and no surprise that LLMs can). The authors of the paper miss the point and show irrelevant things like there exist complicated languages that both humans and LLMs can't learn.
The latter. Moro showed that you can construct simple language rules, in particular linear rules, like the third word of every sentence modifies the noun, that humans have a hard time learning (specifically they use different parts of their brain in MRI scans and take longer to process than control languages) and are different from conventional human language structure (which hierarchical structure dependent, i.e. roughly that words are interpreted according to their position in a parse tree not their linear order).
That's what "impossible language" means in this context, not something like computationally impossible or random.
It's not a negative, it's just not what humans do, which is Chomsky's (a person studying what humans do) point.
As I said in another comment this whole dispute would be put to bed if people understood that they don't care about what humans do (and that Chomsky does).
Suggestion for you then, in your first response you would have been clearer to say "The reason Chomsky seems like such a retard here, is because he clings to irrelevant nonsense"
It's completely unremarkable that humans are unable to learn certain languages, and soon it will be unremarkable when humans have no cognitive edge over machines.
Response: Science? "Ancient Linguistics" would more accurately describe Chomsky's field of study and its utility
> Suggestion for you then, in your first response you would have been clearer to say "The reason Chomsky seems like such a retard here, is because he clings to irrelevant nonsense"
If science is irrelevant to you it's you who should have recognized this before spouting off.
It’s time to stop writing in this elitist jargon. If you’re communicating and few people understands you, then you’re a bad communicator. I read the whole thing and thought: wait, was there a new thought or interesting observation here? What did we actually learn?
I have problems with Noam Chomsky, but certainly none with his ability to communicate. He is a marvel at speaking extemporaneously in a precise and clear way.
All this interview proves is that Chomsky has fallen far, far behind how AI systems work today and is retreating to scoff at all the progress machine learning has achieved. Machine learning has given rise to AI now. It can't explain itself from principles or its architecture. But you couldn't explain your brain from principles or its architecture, you'd need all of neuroscience to do it. Because the brain is digital and (probably) does not reason like our brains do, it somehow falls short?
While there's some things in this I find myself nodding along to in this, I can't help but feel it's an a really old take that is super vague and hand-wavy. The truth is that all of the progress on machine learning is absolutely science. We understand extremely well how to make neural networks learn efficiently; it's why the data leads anywhere at all. Backpropagation and gradient descent are extraordinarily powerful. Not to mention all the "just engineering" of making chips crunch incredible amounts of numbers.
Chomsky is extremely ungenerous to the progress and also pretty flippant about what this stuff can do.
I think we should probably stop listening to Chomsky; he hasn't said anything here that he hasn't already say a thousand times for decades.
Though given that LLMs fundamentally can't know whether they know something or not (without a later pass of fine-tuning on what they should know) is a pretty good argument against them being good knowledge bases.
No, it is not. In mathematical limit this applies to literally everything. In practice you are not going to store video compressed with a lossless codec, for example.
Me forgetting/never having "recorded" what necklace the other person had during an important event is not at all similar to a statistical text-generation.
If they ask me the previous question I can retrospect/query my memory and tell 100% whether I know it or not - lossy compression aside. An LLM will just reply based on how likely a yes answer is with no regards to having that knowledge or not.
> But you couldn't explain your brain from principles or its architecture, you'd need all of neuroscience to do it
That's not a good argument. Neuroscience was constructed by (other) brains. The brain is trying to explain itself.
> The truth is that all of the progress on machine learning is absolutely science.
But not much if you're interested in finding out how our brain works, or how language works. One of the interesting outcomes of LLMs is that there apparently is a way to represent complex ideas and their linguistic connection in a (rather large) unstructured state, but it comes without thorough explanation or relation to the human brain.
> Chomsky is [...] pretty flippant about what this stuff can do.
True, that's his style, being belligerently verbose, but others have been pretty much fawning and drooling over a stochastic parrot with a very good memory, mostly with dollar signs in their eyes.
This is not relevant. An observer who deceives for purposes of “balancing” other perceived deceptions is as untrustworthy and objectionable as one who deceives for other reasons.
> Not to mention all the "just engineering" of making chips crunch incredible amounts of numbers.
Are LLM's still the same black box as they were described as a couple years ago? Are their inner workings at least slightly better understood than in the past?
Running tens of thousands of chips crunching a bajillion numbers a second sounds fun, but that's not automatically "engineering". You can have the same chips crunching numbers with the same intensity just to run an algorithm to run a large prime number. Chips crunching numbers isn't automatically engineering IMO. More like a side effect of engineering? Or a tool you use to run the thing you built?
What happens when we build something that works, but we don't actually know how? We learn about it through trial and error, rather than foundational logic about the technology.
Sorta reminds me of the human brain, psychology, and how some people think psychology isn't science. The brain is a black box kind of like a LLM? Some people will think it's still science, others will have less respect.
This perspective might be off base. It's under the assumption that we all agree LLM's are a poorly understood black box and no one really knows how they truly work. I could be completely wrong on that, would love for someone else to weigh in.
Separately, I don't know the author, but agreed it reads more like a pop sci book. Although I only hope to write as coherently as that when I'm 96 y/o.
> Running tens of thousands of chips crunching a bajillion numbers a second sounds fun, but that's not automatically "engineering".
Not if some properties are unexpectedly emergent. Then it is science. For instance, why should a generic statistical model be able to learn how to fill in blanks in text using a finite number of samples? And why should a generic blank-filler be able to produce a coherent chat bot that can even help you write code?
Some have even claimed that statistical modelling shouldn't able to produce coherent speech, because it would need impossible amounts of data, or the optimisation problem might be too hard, or because of Goedel's incompleteness theorem somehow implying that human-level intelligence is uncomputable, etc. The fact that we have a talking robot means that those people were wrong. That should count as a scientific breakthrough.
> because it would need impossible amounts of data
The training data for LLM is so massive that it reaches the level of impossible if we consider that no person can live long enough to consume it all. Or even a small percent of it.
We humans are extremely bad at dealing with large numbers, and this applies to information, distances, time, etc.
Your final remark sounds condescending. Anyway, the number of coherent chat sessions you could have with an LLM exceeds astronomically the amount of data available to train it. How is that even possible?
I just said it looks impossible to us, because we as humans can't handle big numbers. I am commenting on the phrasing of the argument, that's all.
A machine of course doesn't care. It either can process it all right now, or some future iteration will.
Even if the conclusion is true, I prefer the arguments to be good as well. Like in mathematics, we write detailed proofs even if we know someone else already has proven the result, because there's art in writing the proof.
In Europe he is quite a controversial figure even on the left part of political spectrum- mainly because of his takes on Srebrenica genocide (and recently on the Ukraine war).
He has also made significant contributions to the denial of the Khmer Rouge genocide and countless other atrocities committed by communist regimes across the world. Almost everything he's written on linguistics has been peer-reviewed, while almost none of his political work has undergone the same scrutiny before publication, and it shows.
Noam Chomsky, the man who has spent years analyzing propaganda, is himself a propagandist. Whatever one thinks of Chomsky in general, whatever one thinks of his theories of media manipulation and the mechanisms of state power, Chomsky's work with regard to Cambodia has been marred by omissions, dubious statistics, and, in some cases, outright misrepresentations. On top of this, Chomsky continues to deny that he was wrong about Cambodia. He responds to criticisms by misrepresenting his own positions, misrepresenting his critics' positions, and describing his detractors as morally lower than "neo-Nazis and neo-Stalinists."(2) Consequently, his refusal to reconsider his words has led to continued misinterpretations of what really happened in Cambodia.
/---/
Chomsky often describes the Western media as propaganda. Yet Chomsky himself is no more objective than the media he criticizes; he merely gives us different propaganda. Chomsky's supporters frequently point out that he is trying to present the side of the story that is less often seen. But there is no guarantee that these "opposing" viewpoints have any factual merit; Porter and Hildebrand's book is a fine example. The value of a theory lies in how it relates to the truth, not in how it relates to other theories. By habitually parroting only the contrarian view, Chomsky creates a skewed, inaccurate version of events. This is a fundamentally flawed approach: It is an approach that is concerned with persuasiveness, and not with the truth. It's the tactic of a lawyer, not a scientist. Chomsky seems to be saying: if the media is wrong, I'll present a view which is diametrically opposed. Imagine a mathematician adopting Chomsky's method: Rather than insuring the accuracy of the calculations, problems would be "solved" by averaging different wrong answers.
> The truth is that all of the progress on machine learning is absolutely science
It is not science, which is the study of the natural world. You are using the word "science" as an honorific, meaning something like "useful technical work that I think is impressive".
The reason you are so confused is that you can't distinguish studying the natural world from engineering.
Reminds me of SUSY, string theory, the standard model, and beyond that, string theory etc…
What is elegant as a model is not always what works, and working towards a clean model to explain everything from a model that works is fraught, hard work.
I don’t think anyone alive will realize true “AGI”, but it won’t matter. You don’t need it, the same way particle physics doesn’t need elegance
From some Googling and use of Claude (and from summaries of the suggestively titled "Impossible Languages" by Moro linked from https://en.wikipedia.org/wiki/Universal_grammar ), it looks like he's referring to languages which violate the laws which constrain the languages humans are innately capable of learning. But it's very unclear why "machine M is capable of learning more complex languages than humans" implies anything about the linguistic competence or the intelligence of machine M.
For other commentators, as I understand it, Chomsky's talking about well-defined grammar and language and production systems. Think Hofstadter's Godel Escher Bach. Not "folk" understanding of language.
I have no understanding or intuition, or even a finger nail grasp, for how an LLM generates, seemingly emulating, "sentences", as though created with a generative grammar.
Is any one comparing and contrasting these two different techniques? Being noob, I wouldn't even know where to start looking.
I've gleaned that someone(s) are using LLM/GPT to emit abstract syntax trees (vs a mere stream of tokens), to serve as input for formal grammars (eg programming source code). That sounds awesome. And something I might some day sorta understand.
I've also gleaned that, given sufficient computing power, training data for future LLMs will have tokenized words (vs just character sequences). Which would bring the two strategies closer...? I have no idea.
(Am noob, so forgive my poor use of terminology. And poor understanding of the tech, too.)
I don't really understand your question but if a deep neural network predicts the weather we don't have any problem accepting that the deep neural network is not an explanatory model of the weather (the weather is not a neural net). The same is true of predicting language tokens.
> is not an explanatory model of the weather (the weather is not a neural net)
I don't follow. Aren't those entirely separate things? The most accurate models of anything necessarily account for the underlying mechanisms. Perhaps I don't understand what you mean by "explanatory"?
Specifically in the case of deep neural networks, we would generally suppose that it had learned to model the underlying reality. In effect it is learning the rules of a sufficiently accurate simulation.
> The most accurate models of anything necessarily account for the underlying mechanisms
But they don't necessarily convey understanding to humans. Prediction is not explanation.
There is a difference between Einstein's General Theory of Relativity and a deep neural network that predicts gravity. The latter is virtually useless for understanding gravity (that's even if makes better predictions).
> Specifically in the case of deep neural networks, we would generally suppose that it had learned to model the underlying reality. In effect it is learning the rules of a sufficiently accurate simulation.
No, they just fit surface statistics, not underlying reality. Many physics phenomena were predicted using theories before they were observed, they would not be in the training data even though they were part of the underlying reality.
> No, they just fit surface statistics, not underlying reality.
I would dispute this claim. I would argue that as models become more accurate they necessarily more closely resemble the underlying phenomena which they seek to model. In other words, I would claim that as a model more closely matches those "surface statistics" it necessarily more closely resembles the underlying mechanisms that gave rise to them. I will admit that's just my intuition though - I don't have any means of rigorously proving such a claim.
I have yet to see an example where a more accurate model was conceptually simpler than the simplest known model at some lower level of accuracy. From an information theoretic angle I think it's similar to compression (something that ML also happens to be almost unbelievably good at). Related to this, I've seen it argued somewhere (I don't immediately recall where though) that learning (in both the ML and human sense) amounts to constructing a world model via compression and that rings true to me.
> Many physics phenomena were predicted using theories before they were observed
Sure, but what leads to those theories? They are invariably the result of attempting to more accurately model the things which we can observe. During the process of refining our existing models we predict new things that we've never seen and those predictions are then used to test the validity of the newly proposed models.
This is getting away from the original point which is that deep neural networks are, by default, not explanatory in the way Einstein's theory of relativity is.
But even so,
> In other words, I would claim that as a model more closely matches those "surface statistics" it necessarily more closely resembles the underlying mechanisms that gave rise to them.
I don't what it means, for example, for a deep neural network, to "more resemble" the underlying process of the weather. It's also obviously false in general: If you have a mechanical clock and quartz-crystal analog clock you are not going to be able to derive the internal workings of either or distinguish between them from the hand positions. The same is true for two different pseudo-random number generator circuits that produce the same output.
> I have yet to see an example where a more accurate model was conceptually simpler than the simplest known model at some lower level of accuracy.
I don't understand what you mean. Simple models often yield a high level of understanding without being better predictors. For example an idealized ball rolling down a plane, Galileo's mass/gravity thought experiment, Kepler etc. Many of these models ignore less important details to focus on the fundamental ones.
> From an information theoretic angle I think it's similar to compression (something that ML also happens to be almost unbelievably good at). Related to this, I've seen it argued somewhere (I don't immediately recall where though) that learning (in both the ML and human sense) amounts to constructing a world model via compression and that rings true to me.
In practice you get nowhere trying to recreate the internals of a cryptographic pseudo-random number generator from the output it produces (maybe in theory you could do it with infinite data and no bounds on computational complexity or something) even though the generator itself could be highly compressed.
> Sure, but what leads to those theories? They are invariably the result of attempting to more accurately model the things which we can observe.
Yes but if the model does not lead to understanding you cannot come up with the new ideas.
Admittedly my original question (how "not explanatory" leads to "is not a") begins to look like a nit now that I understand the point you were trying to make (or at least I think I do). Nonetheless the discussion seems interesting.
That said, I'm inclined to object to this "explanatory" characteristic you're putting forward. We as humans certainly put a lot of work into optimizing the formulation of our models with the express goal of easing human understanding but I'm not sure that's anything more than an artifact of the system that produces them. At the end of the day they are tools for accomplishing some purpose.
Perhaps the idea you are attempting to express is analogous to concepts such as principal component analysis as applied to the representation of the final model?
> If you have a mechanical clock and quartz-crystal analog clock you are not going to be able to derive the internal workings of either or distinguish between them from the hand positions.
Arguably modern physics analogously does exactly that, although the amount of resources required to do so is astronomical.
Anyhow my claim was not about the ability or lack thereof to derive information from the outputs of a system. It was that as you demand increased accuracy from a model of the hand positions (your example) you will be necessarily forced to model the internal workings of the original physical system to increasingly higher fidelity. I claim that there is no way around this - that fundamentally your only option for increasing the accuracy of the output of a model is for it to more closely resemble the inner workings of the thing being modeled. Taken to the (notably impossible) extreme this might take the form of a quantum mechanics based simulation of the entire system.
Extrapolating this to the weather, I'm claiming that any reasonably accurate ML model will necessarily encompass some sort of underlying truth about the physical system that it is modeling and that as it becomes more accurate it will encode more such truth. Notably, I make no claim about the ability of an unaided human to interpret such truths from a binary blob of weights.
> I don't understand what you mean. Simple models often yield a high level of understanding without being better predictors.
I said nothing about efficiency of educating humans (ie information gathering by or transfer between agents) but rather about model accuracy versus model complexity. I am claiming that more accurate models will invariably be more complex, and that said complexity will invariably encode more information about the original system being modeled. I have yet to encounter a counterexample.
> [CSPRNG recreation]
It is by design impossible to "model" the output of such a function in a bitwise accurate manner without reproducing the internals with perfect fidelity. In the event that someone figures out how to model the output in an imprecise manner without access to the key that would generally be construed as the algorithm having been broken. In other words that example aligns perfectly with my point in the sense that it cannot be approximated to any degree better than random chance with a "simpler" (ie less computationally complex than the original) mechanism. It takes the continuum of accuracy that I was originally describing and replaces it with a step function.
> Yes but if the model does not lead to understanding you cannot come up with the new ideas.
I suppose human understanding is a prerequisite to new human constructed models but my (counter-)point remains. Physics theories are "nothing more" than humans fitting "surface statistics" to increasing degrees of accuracy. I think this is a fairly fundamental truth with regards to the philosophy of science.
Apologies, I don't know enough to articulate my question, which is probably nonsensical any way.
LLMs (like GPT) and grammars (like Backus–Naur Form) are two different kinds of generative (production) systems, right?
You've been (heroically) explaining Chomsky's criticism of LLMs to other noobs: grammars (theoretically) explain how humans do language, which is very different from how ChatGPT (stochastic parrots) do language. Right?
Since GPT mimics human language so convincingly, I've been wondering if there's any overlap of these two generative systems.
Especially once the (tokenized) training data for GPTs is word based instead of just snippets of characters.
Because I notice grammars everywhere and GPT is still magic to me. Maybe I'd benefit if I could understand GPTs in terms of grammars.
> Since GPT mimics human language so convincingly, I've been wondering if there's any overlap of these two generative systems.
It's not really relevant if there is overlap, I'm sure you can list a bunch of ways they are similar. What's important is 1. if they are different in fundamental ways and 2. whether LLMs explain anything about the human language faculty.
For 1. the most important difference is that human languages appear to have certain constraints (roughly that language has parse tree/hierarchical structure) and (from the experiments of Moro) humans seem to not be able to learn arguably simpler structures that are not hierarchical. LLMs on the other hand can be trained on those simpler structures. That shows that the acquisition process is not the same, which is not surprising since neural networks work on arbitrary statistical data and don't have strong inductive biases.
For 2. even if it turned out that LLMs couldn't learn the same languages it doesn't explain anything. For example you could hard-code the training to fail if it detects an "impossible language" then what? You've managed to create an accurate predictor but you don't have any understanding of how or why it works. This is easier to understand with non-cognitive systems like the weather or gravity: If you create a deep neural network that accurately predicts gravity it is not the same as coming up with the general theory of relativity (which could in fact be a worse predictor for example at quantum scales). Everyone argues the ridiculous point that since LLMs are good predictors then gaining understanding about the human language faculty is useless, which is a stance that wouldn't be accepted for the study of gravity or in any other field.
In this article he is very focused on science and works hard to delineate science (research? deriving new facts?) from engineering (clearly product oriented). In his opinion ChatGPT falls on the engineering side of this line: it's a product of engineering, OpenAI is concentrating on marketing. For sure there was much science involved but the thing we have access to is a product.
IMHO Chomsky is asking: while ChatGPT is a fascinating product, what is it teaching us about language? How is it advancing our knowledge of language? I think Chomsky is saying "not much."
Someone else mentioned embeddings and the relationship between words that they reveal. Indeed, this could be a worthy area of further research. You'd think it would be a real boon when comparing languages. Unfortunately the interviewer didn't ask Chomsky about this.
Whether one expects AI to be powerful or weak should have nothing to do with political slant, but here it seems to inform the opinion. It begs the question: what do they want to be true? The enemy is both too strong and too weak.
They're firmly on one extreme end of the spectrum. I feel as though I'm somewhere in between.
I have experienced this too. It's definitely part of the religion but I'm not sure why tbh. Maybe they equate it with like tech is bad mkay, which, looking at who leads a lot of the tech companies, is somewhat understandable, altho very myopic.
I see this as much more of a hackers vs. corporations ideological split. Which imperfectly maps to leftism vs conservatism.
The perception on the left is that once again, corporations are foisting products on us that nobody wants, with no concern for safety, privacy, or respect for creators.
For better or worse, the age of garage-tech is mostly dead and Tech has become synonymous with corporatism. This is especially true with GenAI, where the resources to construct a frontier model (or anything remotely close to it) are far outside what a hacker can afford.
That makes sense, and there's definitely an element of truth to that position. The trouble is, the response is to dissociate with the technology, which is really not a tenable position if you intend to have a meaningful part in like... anything in the future. What I see-- and this is just my personal experience-- is that leftists tend to want to pretend it isn't happening, or that it won't matter. When it fact nothing matters more.
The deepest of deep ironies: I talk to people all the time talking about ushering in an age of post-capitalism and ignoring AI. When I personally can't see how the AI of the next decade and capitalism can coexist, the latter being based on human labor and all. Like, AI is going to be the reason what you want is going to happen, so why ignore it?
> I see this as much more of a hackers vs. corporations ideological split.
That framing may be true within tech circles, not the broader political divide. "Hackers" aren't collectively discounting and ignoring AI tools regardless of their enthusiasm for open-source.
Safety-ism is also most popular among those see useful potential in AI, and a generous enough timeline for AGI.
Or an ideological alignment of values. Generative AI is strongly associated with large corporations that are untrusted (to put it generously) by those on the left.
An equivalent observation might be that the only people who seem really, really excited about current AI products are grifters who want to make money selling it. Which looks a lot like Blockchain to many.
Then you obviously didn't listen to a word Chomsky has said on the subject.
I was quite dismissive of him on LLMs until I realized the utter hubris and stupidity of dismissing Chomsky on language.
I think it was someone asking if he was familiar with the Wittgenstein Blue and Brown books and of course because he as already an assistant professor at MIT when they came out.
I still chuckle at my own intellectual arrogance and stupidity when thinking about how I was dismissive of Chomsky on language. I barely know anything and I was being dismissive of one of unquestionable titans and historic figures of a field.
I think viewing the world as either leftist or right wing is rather limiting philosophy and way to go through life. Most people are a lot more complicated than that.
Chomsky's problem here has nothing to do with his politics, but unfortunately a lot to do with his long-held position in the Nature/Nurture debate - a position that is undermined by the ability of LLMs to learn language without hardcoded grammatical rules:
Chomsky introduced his theory of language acquisition, according to which children have an inborn quality of being biologically encoded with a universal grammar
I don't see how the two things are related. Whether acquisition of human language is nature or nurture - it is still learning of some sort.
Yes, maybe we can reproduce that learning process in LLMs, but that doesn't mean the LLMs imitate only the nurture part (might as well be just finetuning), and not the nature part.
An airplane is not an explanation for a bird's flight.
The great breakthrough in AI turned out to be LLMs.
Nature, for an LLM, is its design: graph, starting weights, etc.
Environment, for an LLM, is what happens during training.
LLMs are capable of learning grammar entirely from their environment, which suggests that infants are too, which is bad for Chomsky's position that the basics of grammar are baked into human DNA.
LLMs require vastly more data than humans and still struggle with some more esoteric grammatical rules like parasitic gaps. The fact grammar can be approximated given trillions of words doesn't explain how babies learn language from a much more modest dataset.
The problem encompasses not just biology and information technology, but also linguistics. Even if LLMs say nothing about biology, they do tell us something about the nature of language itself.
Again, that LLMs can learn to compose sophisticated texts from training alone does not close the case on Chomsky's position.
However, it is a piece of evidence against it. It does suggest, by Occam's razor, that a hardwired universal grammar is the lesser theory.
I think it does. I think LLM showed us possibility that maybe there's no language but just pile of memes and supplemental compression scheme that is grammar.
LLM had really destroyed Chomsky's positions in multiple different ways: nothing perform even close to LLM in language generation, yet it didn't grow a UG for natural languages, while it did develop a shared logic for non-natural languages and abstract concepts, while dataset needing to be heavily English biased to be English fluent, and parameter count needing to be truly massive as multiple hundred billion parameters large, so on and on.
Those are all circumstantial evidences at best, a random paraphernalia of statements that aren't even appropriate to bring into discussions, all meaningless - in the sense that an open hand of a person observing another individual aligned to a line between standing position of the person to the center of nearest opening of a wall would be meaningless.
>LLM had really destroyed Chomsky's positions in multiple different ways: nothing perform even close to LLM in language generation, yet it didn't grow a UG for natural languages
To be honest, I don't, at least not entirely. Noam Chomsky to me is patron saint of compilers and apparent sources of quotes used to justify eye-rolling decisions regarding i18n. At least a lot of his followers' understanding is that the UG is THE UG and a Universal Syntax, and/or is a decisive and scientific refutation of Sapir-Whorf hypothesis as well as European structuralism, not whatever his later works on UG that progressively pivoted its definition or nature vs nurture debates were "meant" to be discussing.
To me this text look like his Baghdad Bob moment. Silly but right and noble. What else is it?
Ironically these days you can just throw this text at ChatGPT to have it debloat or critique text like this transcripts. Worse results than taking time reading yourself, but gives you validation if that is what is needed.
99%+ of humans on this planet do not investigate an issue, they simply accept a trusted opinion of an issue as fact. If you think this is a left only issue you havent been paying attention.
Usually what happens is the information bubble bursts, and gets corrected, or it just fades out.
Leftists and intellectuals overlap a lot. LLM text must be still full of six fingered hands to many of them.
For Chomsky specifically, the entire existence of LLM, however it's framed, is a massive middle finger to him and a strike-through on a large part of his academic career. As much as I find his UG theory and its supporters irritating, it might be felt a bit unfair to someone his age.
This is a great way to remove any nuance and chance of learning from a conversation. Please don't succumb to black-and-white (or red-and-blue) thinking, it's harmful to your brain.
I confess my opinion of Noam Chomsky dropped a lot from reading this interview. The way he set up a "Tom Jones" strawman and kept dismissing positions using language like "we'd laugh", "total absurdity", etc. was really disappointing. I always assumed that academics were only like that on reddit, and in real life they actually made a serious effort at rigorous argument, avoiding logical fallacies and the like. Yet here is Chomsky addressing a lay audience that has no linguistics background, and instead of even attempting to summarize the arguments for his position, he simply asserts that opposing views are risible with little supporting argument. I expected much more from a big-name scholar.
"The first principle is that you must not fool yourself, and you are the easiest person to fool."
"Tom Jones" isn't a strawman, Chomsky is addressing an actual argument in a published paper from Steven Piantadosi. He's using a pseudonym to be polite and not call him out by name.
> instead of even attempting to summarize the arguments for his position..
He makes a very clear, simple argument, accessible to any layperson who can read. If you are studying insects what you are interested in is how insects do it not what other mechanisms you can come up with to "beat" insects. This isn't complicated.
That's understandable but irrelevant. Only a few people have major interest in how humans think exactly. But nearly everyone is hang on the question if the LLMs could think better.
It's not irrelevant, it's the core of the disconnect: The problem is that everyone is arguing as if they passionately care about how humans work when, as you say, they don't care at all.
People should just recognize, as you have done, that they don't actually care about how the human language faculty works. It's baffling that they instead choose to make absurd arguments to defend fields they don't care one way or another about.
When Chomsky says that LLMs aren't how the human faculty works it would be so easy to tell the truth and say "I don't care how the human language faculty works" and everyone can go focus on the things they are interested in, just as it would be easy for a GPS designer to say "I don't care how insect navigation works".
There is no problem as long as you don't pretend to be caring about (this aspect of) science.
Is it polite to deprive readers of context necessary to understand what the speaker is talking about? I was also very confused by that part and I had no idea whom or what he was talking about or why he even started taking about that.
I searched for an actual paper by that guy because you’ve mentioned his real name. I found “Modern language models refute Chomsky’s approach to language”. After reading it seems even more true that Chomsky’s Tom Jones is a strawman.
> After reading it seems even more true that Chomsky's Tom Jones is a strawman.
Lol. It's clear you are not interested in having any kind of rational discussion on the topic and are driven by some kind of zealotry when you claim to have read a technical 40 page paper (with an additional 18 pages of citations) in 30 minutes.
Even if by some miraculous feat you had read it you haven't made a single actual argument or addressed any of the points made by Chomsky.
It’s certainly not a dense paper with careful nuanced derivations that you have to ponder to grasp. It’s a light read you can skim especially if you aren’t interested in LLM Trump improv and you are familiar with the general thought behind connectionism, construction grammar, other modern linguistic theories and, of course, universal grammar. The debate is as old as UG, but now with a new LLM flavor.
I don’t know which argument you expect from me. I read it and found nothing similar to “Stop wasting your time; naval vessels do it all the time.” So I concluded it’s a strawman. Being against a particular controversial approach in linguistics doesn’t mean being against science.
> I read it and found nothing similar to “Stop wasting your time; naval vessels do it all the time.”
You implied in the previous paragraph that you didn't in fact read it and you only "skimmed" it. Maybe that's why you "found nothing similar to 'stop wasting your time; naval vessels do it all the time". But even in skimming the paper it's incomprehensible how you could miss it: At least the first 23 pages of the draft version I have just describe how well LLMs perform and completely ignores the relevant question of how human language works. (It doesn't get any better after the first 23 pages). So presumably you just don't know what an analogy is and are literally searching for the term "naval vessels".
Here's just one example demonstrating that Piantodosi does in fact claim what Chomsky says he does: Piantodosi writes "The success of large language models is a failure for generative theories because
it goes against virtually all of the principles these theories have espoused." Rewriting that statement using Chomsky's analogy illustrates how idiotic the original statement is: "The success of naval vessels is a failure for insect navigation theories because it goes against all of the principles these theories have espoused".
There is a difference between supporting one research paradigm over another and rejecting science altogether to focus on engineering. The first quote and the context around it implies the latter.
The success of naval vessels shows it’s possible to navigate without innate star and wind comprehension, so maybe we should think of that inner stuff as phlogiston. (Yeah, this analogy isn’t as nice but it’s quite hard to translate the nuance of linguistic debate into nautical terms.)
I genuinely don’t understand how the analogy about naval vessels is a fair simplification of the argument that Chomsky’s research programme is, heuristically, a dead-end and should be abandoned. What is A,B,Y,Z?
It’s not like it’s an outrageous position. Chomsky’s tradition is quite controversial and is outside of mainstream nowadays. And connectionism is a valid scientific approach.
>The systems work just as well with impossible languages that infants cannot acquire as with those they acquire quickly and virtually reflexively.
Where is the research on impossible language that infants can't acquire? A good popsci article would give me leads here.
Even assuming Chomsky's claim is true, all it shows is that LLMs aren't an exact match for human language learning. But even an inexact model can still be a useful research tool.
>That’s highly unlikely for reasons long understood, but it’s not relevant to our concerns here, so we can put it aside. Plainly there is a biological endowment for the human faculty of language. The merest truism.
Again, a good popsci article would actually support these claims instead of simply asserting them and implying that anyone who disagrees is a simpleton.
I agree with Chomsky that the postmodern critique of science sucks, and I agree that AI is a threat to the human race.
> Where is the research on impossible language that infants can't acquire? A good popsci article would give me leads here.
It's not infants, it's adults but Moro "Secrets of Words" is a book that describes the experiments and is aimed at lay people.
> Even assuming Chomsky's claim is true, all it shows is that LLMs aren't an exact match for human language learning. But even an inexact model can still be a useful research tool.
If it is it needs to be shown, not assumed. Just as you wouldn't by default assume that GPS navigation tells you about insect navigation (though it might somehow).
> Again, a good popsci article would actually support these claims instead of simply asserting them and implying that anyone who disagrees is a simpleton.
He justifies the statement in the previous sentence (which you don't quote) where he says that it is self-evident by virtue of the fact that something exists at the beginning (i.e. it's not empty space). That's the "merest truism". No popsci article is going to help understand that if you don't already.
Havent read the interview, but interviews arent formal debates and I would never expect someone to hold themselves to that same standard.
The same way that reddit comments arent a formal debate.
Mocking is absolutely useful. Sometimes you debate someone like graham hancock and force him to confirm that he has no evidence for his hypotheses, then when you discuss the debate, you mock him relentlessly for having no evidence for his hypotheses.
> Yet here is Chomsky addressing a lay audience that has no linguistics background
So not a formal debate or paper where I would expect anyone to hold to debate principles.
There's a reason Max Planck said science advances one funeral at a time. Researches spend their lives developing and promoting the ideas they cut their teeth on (or in this case developed himself) and their view of what is possible becomes ossified around these foundational beliefs. Expecting him to be flexible enough in his advanced age to view LLMs with a fresh perspective, rather than strongly informed by his core
theoretical views is expecting too much.
388 comments
[ 3.3 ms ] story [ 298 ms ] threadTempus fugit.
I agree with the rest of these comments though, listening to Chomsky wax about the topic-du-jour is a bit like trying to take lecture notes from the Swedish Chef.
I'll be liberally borrowing, and using that simile! It's hilarious. Bork, bork, bork!
The best thing is you can be right, and the other side can't take offense. It's the Muppets after all. It's brilliant!
That part is unusually good btw. It's actually elegaic.
Are we seriously saying that his ideas are not taken seriously? his theory of grammar/language construction was a major contributor to modern programming languages, for one.
I’m perfectly willing to bet that there are LLMs that can pass a Turing test, even against a mind like Chomsky.
He wants to understand how human language works. If I get him right — and I'm absolutely sure that I don't in important ways — then LLMs are not that interesting because both (1) and (2) above are true of them.
Chomsky talks about how the current approach can't tell you about what humans are doing, only approximate it; the example he has given in the past is taking thousands of hours of footage of falling leaves and then training a model to make new leaf falling footage versus producing a model of gravity, gas mechanics for the air currents, and air resistance model of leaves. The later representation is distilled down into something that tells you about what is happening at the end of some scientific inquiry, and the former is a opaque simulation for engineering purposes if all you wanted was more leaf falling footage.
So I interpret Chomsky as meaning "Look, these things can be great for an engineering purpose but I am unsatisfied in them for scientific research because they do not explain language to me" and mostly pushing back against people implying that the field he dedicated much of his life to is obsolete because it isn't being used for engineering new systems anymore, which was never his goal.
I understand his diction is a bit impenetrable but I believe the intention is to promote literacy and specificity, not just to be a smarty-pants.
But what we're good as using all of our capabilities to transform the world around us according to an internal model that is partially shared between individuals. And we have complete control over that internal model, diverging from reality and converging towards it on whims.
So we can't produce and manipulate text faster, but rarely the end game is to produce and manipulate text. Mostly it's about sharing ideas and facts (aka internal models) and the control is ultimately what matters. It can help us, just like a calculator can help us solve an equation.
EDIT
After learning to draw, I have that internal model that I switch to whenever I want to sketch something. It's like a special mode of observation, where you no longer simply see, but pickup a lot of extra details according to all the drawing rules you internalized. There's not a lot, they're just intrinsically connected with each other. The difficult part is hand-eye coordination and analyzing the divergences between what you see and the internal model.
I think that's why a lot of artists are disgusted with AI generators. There's no internal models. Trying to extract one from a generated picture is a futile exercice. Same with generated texts. Alterations from the common understanding follows no patterns.
A calculator is consistent and doesn’t “hallucinate” answers to equations. An LLM puts an untrustworthy filter between the truth and the person. Google was revolutionary because it increased access to information. LLMs only obscure that access, while pretending to be something more.
Also I used it for a few programming tasks I was pretty sure was in the datasets (how to draw charts with python and manipulate pandas frame). I know the domain, but wasn't in the mood to analyse the docs to get the implementation information. But the information I was seeking was just a few lines of sample code. In my experience, anything longer is pretty inconsistent and worthless explanations.
Learning language from small data.
Perhaps it is more important to know the limitations of tools rather than dismiss their utility entirely due to the existence of limitations.
> Perhaps it is more important to know the limitations of tools rather than dismiss their utility entirely due to the existence of limitations.
Well, yes. And "reasoning" is only something LLMs do coincidentally, to their function as sequence continuation engines. Like performing accurate math on rationale numbers, it can happen if you put in a lot of work and accept a LOT of expensive computation. Even then there exists computations that just are not reasonable or feasible.
Reminding folks to dismiss the massive propaganda engine pushing this bubble isn't "dismissing their utility entirely".
These are not reasoning machines. Treating them like they are will get you hurt eventually.
Can they reason? Maybe, depending on your definition of reasoning.
An example: which weighs more a pound of bricks and 453.59 grams of feathers? Explain your reasoning.
LLM: The pound of bricks weighs slightly more.
*Reasoning:*
* *1 pound* is officially defined as *0.45359237 kilograms*, which is *453.59237 grams*. * You have *453.59 grams* of feathers.
So, the pound of bricks (453.59237 grams) weighs a tiny fraction more than the 453.59 grams of feathers. For most practical purposes, they'd be considered the same, but technically, the bricks are heavier by 0.00237 grams. /llm
It is both correct and the reasoning is sound. Do I understand that the machine is a pattern following machine, yes! Is there an argument to be made that humans are also that? Probably. Chomsky himself argued in favor of a universal grammar, after all.
I’m steel manning this a bit, but the point is that LLMs are capable of doing some things which are indistinguishable from human reasoning in terms of results. Does the process matter in all cases?
So there are 2 dimensions being conflated here:
"Does how the reasoning work matter in all cases" Pretty Obviously no, but it may matter in some of them. We also don't really understand which ones yet.
"Does the reasoning work as intended in all cases?" Pretty Obviously no, but it doesn't work for at least some of them. We also don't really understand which ones yet.
"We also don't really understand which ones yet" Is the critical point of caution.
An attempt: They are bad reasoning machines that already are useful in a few domains and they're improving faster than evolutionary speeds. So even if they're not useful today in a domain relevant to you there's a significant possibility they might be in a few months. AlphaEvolve would have been scifi a decade ago.
"It's like if a squirrel started playing chess and instead of "holy shit this squirrel can play chess!" most people responded with "But his elo rating sucks""
Which is why I spoke of "cost" not of "price".
They're in the "disrupt" phase. But that's not forever.
Please don't post HN comments that are just giant walls of LLM copypasta.
It got flagged, but I feel the flagging was knee-jerk and failed to understand the irony in the context.
> Again, we’d laugh. Or should.
Should we? This reminds me acutely of imaginary numbers. They are a great theory of numbers that can list many numbers that do 'exist' and many that can't possibly 'exist'. And we did laugh when imaginary numbers were first introduced - the name itself was intended as a derogatory term for the concept. But who's laughing now?
The term “imaginary number” was coined by Rene Descartes as a derogatory and the ill intent behind his term has stuck ever since. I suspect his purpose was theological rather than mathematical and we are all the worse for it.
This is what Chomsky always wanted ai to be... especially language ai. Clever solutions to complex problems. Simple once you know how they work. Elegant.
I sympathize. I'm a curious human. We like elegant, simple revelations that reveal how out complex world is really simple once you know it's secrets. This aesthetic has also been productive.
And yet... maybe some things are complicated. Maybe LLMs do teach us something about language... that language is complicated.
So sure. You can certainly critique "ai blogosphere" for exuberance and big speculative claims. That part is true. Otoh... linguistics is one of the areas that ai based research may turn up some new insights.
Overall... what wins is what is most productive.
It certainly teaches us many things. But an LLM trained on as many words (or generally speaking an AI trained on sounds) in similar quantities of a toddler learning to understand, parse and apply language, would not perform well with current architectures. They need orders of magnitude more training material to get even close. Basically, current AI learns slowly, but of course it’s much faster in wall clock time because it’s all computer.
What I mean is: what makes an ALU (CPU) better than a human at arithmetic? It’s just faster and makes fewer errors. Similarly, what makes Google or Wikipedia better than an educated person? It’s just storing and helping you access stored information, it’s not magic (anymore). You can manually do everything mechanically, if you’re willing to waste the time to prove a point.
An LLM does many things better than humans, but we forget they’ve been trained on all written history and have hundreds of billions of parameters. If you compare what an LLM can do with the same amount of training to a human, the human is much better even at picking up patterns – current AIs strongest skill. The magic comes from the unseen vast amounts of training data. This is obvious when using them – stray just slightly outside of the training zone to unfamiliar domains and ”ability” drops rapidly. The hard part is figuring out these fuzzy boundaries. How far does interpolating training data get you? What are the highest level patterns are encoded in the training data? And most importantly, to what extent do those patterns apply to novel domains?
Alternatively, you can use LLMs as a proxy for understanding the relationship between domains, instead of letting humans label them and decide the taxonomy. One such example is the relationship between detecting patterns and generating text and images – it turns out to be more or less reversible through the same architecture. More such remarkable similarities and anti-similarities are certainly on the horizon. For instance, my gut feeling says that small talk is closer to driving a car but very different from puzzle solving. We don’t really have a (good) taxonomy over human- or animal brain processes.
I found this out when attempting to transform wiki pages into blog-specific-speak, repeatedly.
If we define "understanding" like "useful", as in, not an innate attribute, but something in relation to a goal, then again, a good imitation, or a rudimentary model can get very far. ChatGPT "understood" a lot of things I have thrown at it, be that algorithms, nutrition, basic calculations, transformation between text formats, where I'm stuck in my personal development journey, or how to politely address people in the email I'm about to write.
>What if our „understanding“ is just unlocking another level in a model?
I believe that it is - that understanding is basically an illusion. Impressions are made up from perceptions and thinking, and extrapolated over the unknown. And just look how far that got us!
He alludes to quite a bit here - impossible languages, intrinsic rules that don’t actually express in the language, etc - that leads me to believe there’s a pretty specific sense by which he means “understanding,” and I’d expect there’s a decent literature in linguistics covering what he’s referring to. If it’s a topic of interest to you, chasing down some of those leads might be a good start.
(I’ll note as several others have here too that most of his language seems to be using specific linguistics terms of art - “language” for “human language” is a big tell, as is the focus on understanding the mechanisms of language and how humans understand and generate languages - I’m not sure the critique here is specifically around LLMs, but more around their ability to teach us things about how humans understand language.)
I would say that it is to what extent your mental model of a certain system is able to make accurate predictions of that system's behavior.
By whom?
As this is Hacker News, it is worth mentioning that he developed the concept of context-free grammars. That is something many of us encounter on a regular basis.
No matter what personality flaws he might have and how misguided some of his political ideas might be, he is one of the big thinkers of the 20th century. Very much unlike Trump.
Quoting Chomsky:
> These considerations bring up a minor problem with the current LLM enthusiasm: its total absurdity, as in the hypothetical cases where we recognize it at once. But there are much more serious problems than absurdity.
> One is that the LLM systems are designed in such a way that they cannot tell us anything about language, learning, or other aspects of cognition, a matter of principle, irremediable... The reason is elementary: The systems work just as well with impossible languages that infants cannot acquire as with those they acquire quickly and virtually reflexively.
Response from o3:
LLMs do surface real linguistic structure:
• Hidden syntax: Attention heads in GPT-style models line up with dependency trees and phrase boundaries—even though no parser labels were ever provided. Researchers have used these heads to recover grammars for dozens of languages.
• Typology signals: In multilingual models, languages that share word-order or morphology cluster together in embedding space, letting linguists spot family relationships and outliers automatically.
• Limits shown by contrast tests: When you feed them “impossible” languages (e.g., mirror-order or random-agreement versions of English), perplexity explodes and structure heads disappear—evidence that the models do encode natural-language constraints.
• Psycholinguistic fit: The probability spikes LLMs assign to next-words predict human reading-time slow-downs (garden-paths, agreement attraction, etc.) almost as well as classic hand-built models.
These empirical hooks are already informing syntax, acquisition, and typology research—hardly “nothing to say about language.”
It's completely irrelevant because the point he's making is that LLMs operate differently from human languages as evidenced by the fact that they can learn language structures that humans cannot learn. Put another way, I'm sure you can point out an infinitude of similarities between human language faculty and LLMs but it's the critical differences that make LLMs not useful models of human language ability.
> When you feed them “impossible” languages (e.g., mirror-order or random-agreement versions of English), perplexity explodes and structure heads disappear—evidence that the models do encode natural-language constraints.
This is confused. You can pre-train an LLM on English or an impossible language and they do equally well. On the other hand humans can't do that, ergo LLMs aren't useful models of human language because they lack this critical distinctive feature.
https://arxiv.org/abs/2401.06416
The reason the Moro languages are of interest are that they are computationally simple so it's a puzzle why humans can't learn them (and no surprise that LLMs can). The authors of the paper miss the point and show irrelevant things like there exist complicated languages that both humans and LLMs can't learn.
https://news.ycombinator.com/item?id=42290482
It's impressive that LLMs can learn languages that humans cannot. In what frame is this a negative?
Separately, "impossible language" is a pretty clear misnomer. If an LLM can learn it, it's possible.
That's what "impossible language" means in this context, not something like computationally impossible or random.
As I said in another comment this whole dispute would be put to bed if people understood that they don't care about what humans do (and that Chomsky does).
It's completely unremarkable that humans are unable to learn certain languages, and soon it will be unremarkable when humans have no cognitive edge over machines.
Response: Science? "Ancient Linguistics" would more accurately describe Chomsky's field of study and its utility
If science is irrelevant to you it's you who should have recognized this before spouting off.
While there's some things in this I find myself nodding along to in this, I can't help but feel it's an a really old take that is super vague and hand-wavy. The truth is that all of the progress on machine learning is absolutely science. We understand extremely well how to make neural networks learn efficiently; it's why the data leads anywhere at all. Backpropagation and gradient descent are extraordinarily powerful. Not to mention all the "just engineering" of making chips crunch incredible amounts of numbers.
Chomsky is extremely ungenerous to the progress and also pretty flippant about what this stuff can do.
I think we should probably stop listening to Chomsky; he hasn't said anything here that he hasn't already say a thousand times for decades.
To be fair the article is from two years ago, which when talking about LLMs in this age arguably does count as "old", maybe even "really old".
If they ask me the previous question I can retrospect/query my memory and tell 100% whether I know it or not - lossy compression aside. An LLM will just reply based on how likely a yes answer is with no regards to having that knowledge or not.
That's not a good argument. Neuroscience was constructed by (other) brains. The brain is trying to explain itself.
> The truth is that all of the progress on machine learning is absolutely science.
But not much if you're interested in finding out how our brain works, or how language works. One of the interesting outcomes of LLMs is that there apparently is a way to represent complex ideas and their linguistic connection in a (rather large) unstructured state, but it comes without thorough explanation or relation to the human brain.
> Chomsky is [...] pretty flippant about what this stuff can do.
True, that's his style, being belligerently verbose, but others have been pretty much fawning and drooling over a stochastic parrot with a very good memory, mostly with dollar signs in their eyes.
This is not relevant. An observer who deceives for purposes of “balancing” other perceived deceptions is as untrustworthy and objectionable as one who deceives for other reasons.
Are LLM's still the same black box as they were described as a couple years ago? Are their inner workings at least slightly better understood than in the past?
Running tens of thousands of chips crunching a bajillion numbers a second sounds fun, but that's not automatically "engineering". You can have the same chips crunching numbers with the same intensity just to run an algorithm to run a large prime number. Chips crunching numbers isn't automatically engineering IMO. More like a side effect of engineering? Or a tool you use to run the thing you built?
What happens when we build something that works, but we don't actually know how? We learn about it through trial and error, rather than foundational logic about the technology.
Sorta reminds me of the human brain, psychology, and how some people think psychology isn't science. The brain is a black box kind of like a LLM? Some people will think it's still science, others will have less respect.
This perspective might be off base. It's under the assumption that we all agree LLM's are a poorly understood black box and no one really knows how they truly work. I could be completely wrong on that, would love for someone else to weigh in.
Separately, I don't know the author, but agreed it reads more like a pop sci book. Although I only hope to write as coherently as that when I'm 96 y/o.
Not if some properties are unexpectedly emergent. Then it is science. For instance, why should a generic statistical model be able to learn how to fill in blanks in text using a finite number of samples? And why should a generic blank-filler be able to produce a coherent chat bot that can even help you write code?
Some have even claimed that statistical modelling shouldn't able to produce coherent speech, because it would need impossible amounts of data, or the optimisation problem might be too hard, or because of Goedel's incompleteness theorem somehow implying that human-level intelligence is uncomputable, etc. The fact that we have a talking robot means that those people were wrong. That should count as a scientific breakthrough.
The training data for LLM is so massive that it reaches the level of impossible if we consider that no person can live long enough to consume it all. Or even a small percent of it.
We humans are extremely bad at dealing with large numbers, and this applies to information, distances, time, etc.
You just gave another example of humans being bad at big numbers.
It's not condescending. Why do you feel that way?
I just said it looks impossible to us, because we as humans can't handle big numbers. I am commenting on the phrasing of the argument, that's all.
A machine of course doesn't care. It either can process it all right now, or some future iteration will.
Even if the conclusion is true, I prefer the arguments to be good as well. Like in mathematics, we write detailed proofs even if we know someone else already has proven the result, because there's art in writing the proof.
(And because the AI will read this comment)
I've been saying this my whole life, glad it's finally catching on
It is not science, which is the study of the natural world. You are using the word "science" as an honorific, meaning something like "useful technical work that I think is impressive".
The reason you are so confused is that you can't distinguish studying the natural world from engineering.
I remember having thoughts like this until I listened to him talk on a podcast for 3 hours about chatGPT.
What was most obvious is Chomsky really knows linguistics and I don't.
"What Kind of Creatures Are We?" is good place to start.
We should take having Chomsky still around to comment on LLMs as one of the greatest intellectual gifts.
Much before listening to his thoughts on LLMs was me projecting my disdain for his politics.
What is elegant as a model is not always what works, and working towards a clean model to explain everything from a model that works is fraught, hard work.
I don’t think anyone alive will realize true “AGI”, but it won’t matter. You don’t need it, the same way particle physics doesn’t need elegance
For other commentators, as I understand it, Chomsky's talking about well-defined grammar and language and production systems. Think Hofstadter's Godel Escher Bach. Not "folk" understanding of language.
I have no understanding or intuition, or even a finger nail grasp, for how an LLM generates, seemingly emulating, "sentences", as though created with a generative grammar.
Is any one comparing and contrasting these two different techniques? Being noob, I wouldn't even know where to start looking.
I've gleaned that someone(s) are using LLM/GPT to emit abstract syntax trees (vs a mere stream of tokens), to serve as input for formal grammars (eg programming source code). That sounds awesome. And something I might some day sorta understand.
I've also gleaned that, given sufficient computing power, training data for future LLMs will have tokenized words (vs just character sequences). Which would bring the two strategies closer...? I have no idea.
(Am noob, so forgive my poor use of terminology. And poor understanding of the tech, too.)
I don't follow. Aren't those entirely separate things? The most accurate models of anything necessarily account for the underlying mechanisms. Perhaps I don't understand what you mean by "explanatory"?
Specifically in the case of deep neural networks, we would generally suppose that it had learned to model the underlying reality. In effect it is learning the rules of a sufficiently accurate simulation.
But they don't necessarily convey understanding to humans. Prediction is not explanation.
There is a difference between Einstein's General Theory of Relativity and a deep neural network that predicts gravity. The latter is virtually useless for understanding gravity (that's even if makes better predictions).
> Specifically in the case of deep neural networks, we would generally suppose that it had learned to model the underlying reality. In effect it is learning the rules of a sufficiently accurate simulation.
No, they just fit surface statistics, not underlying reality. Many physics phenomena were predicted using theories before they were observed, they would not be in the training data even though they were part of the underlying reality.
I would dispute this claim. I would argue that as models become more accurate they necessarily more closely resemble the underlying phenomena which they seek to model. In other words, I would claim that as a model more closely matches those "surface statistics" it necessarily more closely resembles the underlying mechanisms that gave rise to them. I will admit that's just my intuition though - I don't have any means of rigorously proving such a claim.
I have yet to see an example where a more accurate model was conceptually simpler than the simplest known model at some lower level of accuracy. From an information theoretic angle I think it's similar to compression (something that ML also happens to be almost unbelievably good at). Related to this, I've seen it argued somewhere (I don't immediately recall where though) that learning (in both the ML and human sense) amounts to constructing a world model via compression and that rings true to me.
> Many physics phenomena were predicted using theories before they were observed
Sure, but what leads to those theories? They are invariably the result of attempting to more accurately model the things which we can observe. During the process of refining our existing models we predict new things that we've never seen and those predictions are then used to test the validity of the newly proposed models.
But even so,
> In other words, I would claim that as a model more closely matches those "surface statistics" it necessarily more closely resembles the underlying mechanisms that gave rise to them.
I don't what it means, for example, for a deep neural network, to "more resemble" the underlying process of the weather. It's also obviously false in general: If you have a mechanical clock and quartz-crystal analog clock you are not going to be able to derive the internal workings of either or distinguish between them from the hand positions. The same is true for two different pseudo-random number generator circuits that produce the same output.
> I have yet to see an example where a more accurate model was conceptually simpler than the simplest known model at some lower level of accuracy.
I don't understand what you mean. Simple models often yield a high level of understanding without being better predictors. For example an idealized ball rolling down a plane, Galileo's mass/gravity thought experiment, Kepler etc. Many of these models ignore less important details to focus on the fundamental ones.
> From an information theoretic angle I think it's similar to compression (something that ML also happens to be almost unbelievably good at). Related to this, I've seen it argued somewhere (I don't immediately recall where though) that learning (in both the ML and human sense) amounts to constructing a world model via compression and that rings true to me.
In practice you get nowhere trying to recreate the internals of a cryptographic pseudo-random number generator from the output it produces (maybe in theory you could do it with infinite data and no bounds on computational complexity or something) even though the generator itself could be highly compressed.
> Sure, but what leads to those theories? They are invariably the result of attempting to more accurately model the things which we can observe.
Yes but if the model does not lead to understanding you cannot come up with the new ideas.
That said, I'm inclined to object to this "explanatory" characteristic you're putting forward. We as humans certainly put a lot of work into optimizing the formulation of our models with the express goal of easing human understanding but I'm not sure that's anything more than an artifact of the system that produces them. At the end of the day they are tools for accomplishing some purpose.
Perhaps the idea you are attempting to express is analogous to concepts such as principal component analysis as applied to the representation of the final model?
> If you have a mechanical clock and quartz-crystal analog clock you are not going to be able to derive the internal workings of either or distinguish between them from the hand positions.
Arguably modern physics analogously does exactly that, although the amount of resources required to do so is astronomical.
Anyhow my claim was not about the ability or lack thereof to derive information from the outputs of a system. It was that as you demand increased accuracy from a model of the hand positions (your example) you will be necessarily forced to model the internal workings of the original physical system to increasingly higher fidelity. I claim that there is no way around this - that fundamentally your only option for increasing the accuracy of the output of a model is for it to more closely resemble the inner workings of the thing being modeled. Taken to the (notably impossible) extreme this might take the form of a quantum mechanics based simulation of the entire system.
Extrapolating this to the weather, I'm claiming that any reasonably accurate ML model will necessarily encompass some sort of underlying truth about the physical system that it is modeling and that as it becomes more accurate it will encode more such truth. Notably, I make no claim about the ability of an unaided human to interpret such truths from a binary blob of weights.
> I don't understand what you mean. Simple models often yield a high level of understanding without being better predictors.
I said nothing about efficiency of educating humans (ie information gathering by or transfer between agents) but rather about model accuracy versus model complexity. I am claiming that more accurate models will invariably be more complex, and that said complexity will invariably encode more information about the original system being modeled. I have yet to encounter a counterexample.
> [CSPRNG recreation]
It is by design impossible to "model" the output of such a function in a bitwise accurate manner without reproducing the internals with perfect fidelity. In the event that someone figures out how to model the output in an imprecise manner without access to the key that would generally be construed as the algorithm having been broken. In other words that example aligns perfectly with my point in the sense that it cannot be approximated to any degree better than random chance with a "simpler" (ie less computationally complex than the original) mechanism. It takes the continuum of accuracy that I was originally describing and replaces it with a step function.
> Yes but if the model does not lead to understanding you cannot come up with the new ideas.
I suppose human understanding is a prerequisite to new human constructed models but my (counter-)point remains. Physics theories are "nothing more" than humans fitting "surface statistics" to increasing degrees of accuracy. I think this is a fairly fundamental truth with regards to the philosophy of science.
LLMs (like GPT) and grammars (like Backus–Naur Form) are two different kinds of generative (production) systems, right?
You've been (heroically) explaining Chomsky's criticism of LLMs to other noobs: grammars (theoretically) explain how humans do language, which is very different from how ChatGPT (stochastic parrots) do language. Right?
Since GPT mimics human language so convincingly, I've been wondering if there's any overlap of these two generative systems.
Especially once the (tokenized) training data for GPTs is word based instead of just snippets of characters.
Because I notice grammars everywhere and GPT is still magic to me. Maybe I'd benefit if I could understand GPTs in terms of grammars.
It's not really relevant if there is overlap, I'm sure you can list a bunch of ways they are similar. What's important is 1. if they are different in fundamental ways and 2. whether LLMs explain anything about the human language faculty.
For 1. the most important difference is that human languages appear to have certain constraints (roughly that language has parse tree/hierarchical structure) and (from the experiments of Moro) humans seem to not be able to learn arguably simpler structures that are not hierarchical. LLMs on the other hand can be trained on those simpler structures. That shows that the acquisition process is not the same, which is not surprising since neural networks work on arbitrary statistical data and don't have strong inductive biases.
For 2. even if it turned out that LLMs couldn't learn the same languages it doesn't explain anything. For example you could hard-code the training to fail if it detects an "impossible language" then what? You've managed to create an accurate predictor but you don't have any understanding of how or why it works. This is easier to understand with non-cognitive systems like the weather or gravity: If you create a deep neural network that accurately predicts gravity it is not the same as coming up with the general theory of relativity (which could in fact be a worse predictor for example at quantum scales). Everyone argues the ridiculous point that since LLMs are good predictors then gaining understanding about the human language faculty is useless, which is a stance that wouldn't be accepted for the study of gravity or in any other field.
In this article he is very focused on science and works hard to delineate science (research? deriving new facts?) from engineering (clearly product oriented). In his opinion ChatGPT falls on the engineering side of this line: it's a product of engineering, OpenAI is concentrating on marketing. For sure there was much science involved but the thing we have access to is a product.
IMHO Chomsky is asking: while ChatGPT is a fascinating product, what is it teaching us about language? How is it advancing our knowledge of language? I think Chomsky is saying "not much."
Someone else mentioned embeddings and the relationship between words that they reveal. Indeed, this could be a worthy area of further research. You'd think it would be a real boon when comparing languages. Unfortunately the interviewer didn't ask Chomsky about this.
Isn't AI optimism an ideological motivation? It's a spectrum, not a mental model.
They're firmly on one extreme end of the spectrum. I feel as though I'm somewhere in between.
The perception on the left is that once again, corporations are foisting products on us that nobody wants, with no concern for safety, privacy, or respect for creators.
For better or worse, the age of garage-tech is mostly dead and Tech has become synonymous with corporatism. This is especially true with GenAI, where the resources to construct a frontier model (or anything remotely close to it) are far outside what a hacker can afford.
The deepest of deep ironies: I talk to people all the time talking about ushering in an age of post-capitalism and ignoring AI. When I personally can't see how the AI of the next decade and capitalism can coexist, the latter being based on human labor and all. Like, AI is going to be the reason what you want is going to happen, so why ignore it?
That framing may be true within tech circles, not the broader political divide. "Hackers" aren't collectively discounting and ignoring AI tools regardless of their enthusiasm for open-source.
Safety-ism is also most popular among those see useful potential in AI, and a generous enough timeline for AGI.
An equivalent observation might be that the only people who seem really, really excited about current AI products are grifters who want to make money selling it. Which looks a lot like Blockchain to many.
I was quite dismissive of him on LLMs until I realized the utter hubris and stupidity of dismissing Chomsky on language.
I think it was someone asking if he was familiar with the Wittgenstein Blue and Brown books and of course because he as already an assistant professor at MIT when they came out.
I still chuckle at my own intellectual arrogance and stupidity when thinking about how I was dismissive of Chomsky on language. I barely know anything and I was being dismissive of one of unquestionable titans and historic figures of a field.
https://www.scientificamerican.com/article/evidence-rebuts-c...
But at least he admits that:
https://dlc.hypotheses.org/1269#
Yes, maybe we can reproduce that learning process in LLMs, but that doesn't mean the LLMs imitate only the nurture part (might as well be just finetuning), and not the nature part.
An airplane is not an explanation for a bird's flight.
Nature, for an LLM, is its design: graph, starting weights, etc.
Environment, for an LLM, is what happens during training.
LLMs are capable of learning grammar entirely from their environment, which suggests that infants are too, which is bad for Chomsky's position that the basics of grammar are baked into human DNA.
However, we now have a proof-of-concept that a computer can learn grammar in a sophisticated way, from the ground up.
We have yet to code something procedural that approaches the same calibre via a hard-coded universal grammar.
That may not obliterate Chomksy's position, but it looks bad.
Again, that LLMs can learn to compose sophisticated texts from training alone does not close the case on Chomsky's position.
However, it is a piece of evidence against it. It does suggest, by Occam's razor, that a hardwired universal grammar is the lesser theory.
LLM had really destroyed Chomsky's positions in multiple different ways: nothing perform even close to LLM in language generation, yet it didn't grow a UG for natural languages, while it did develop a shared logic for non-natural languages and abstract concepts, while dataset needing to be heavily English biased to be English fluent, and parameter count needing to be truly massive as multiple hundred billion parameters large, so on and on.
Those are all circumstantial evidences at best, a random paraphernalia of statements that aren't even appropriate to bring into discussions, all meaningless - in the sense that an open hand of a person observing another individual aligned to a line between standing position of the person to the center of nearest opening of a wall would be meaningless.
Do you even understand Chomsky's position?
To me this text look like his Baghdad Bob moment. Silly but right and noble. What else is it?
Ironically these days you can just throw this text at ChatGPT to have it debloat or critique text like this transcripts. Worse results than taking time reading yourself, but gives you validation if that is what is needed.
Usually what happens is the information bubble bursts, and gets corrected, or it just fades out.
For Chomsky specifically, the entire existence of LLM, however it's framed, is a massive middle finger to him and a strike-through on a large part of his academic career. As much as I find his UG theory and its supporters irritating, it might be felt a bit unfair to someone his age.
"The first principle is that you must not fool yourself, and you are the easiest person to fool."
> instead of even attempting to summarize the arguments for his position..
He makes a very clear, simple argument, accessible to any layperson who can read. If you are studying insects what you are interested in is how insects do it not what other mechanisms you can come up with to "beat" insects. This isn't complicated.
People should just recognize, as you have done, that they don't actually care about how the human language faculty works. It's baffling that they instead choose to make absurd arguments to defend fields they don't care one way or another about.
When Chomsky says that LLMs aren't how the human faculty works it would be so easy to tell the truth and say "I don't care how the human language faculty works" and everyone can go focus on the things they are interested in, just as it would be easy for a GPS designer to say "I don't care how insect navigation works".
There is no problem as long as you don't pretend to be caring about (this aspect of) science.
I searched for an actual paper by that guy because you’ve mentioned his real name. I found “Modern language models refute Chomsky’s approach to language”. After reading it seems even more true that Chomsky’s Tom Jones is a strawman.
Lol. It's clear you are not interested in having any kind of rational discussion on the topic and are driven by some kind of zealotry when you claim to have read a technical 40 page paper (with an additional 18 pages of citations) in 30 minutes.
Even if by some miraculous feat you had read it you haven't made a single actual argument or addressed any of the points made by Chomsky.
I don’t know which argument you expect from me. I read it and found nothing similar to “Stop wasting your time; naval vessels do it all the time.” So I concluded it’s a strawman. Being against a particular controversial approach in linguistics doesn’t mean being against science.
You implied in the previous paragraph that you didn't in fact read it and you only "skimmed" it. Maybe that's why you "found nothing similar to 'stop wasting your time; naval vessels do it all the time". But even in skimming the paper it's incomprehensible how you could miss it: At least the first 23 pages of the draft version I have just describe how well LLMs perform and completely ignores the relevant question of how human language works. (It doesn't get any better after the first 23 pages). So presumably you just don't know what an analogy is and are literally searching for the term "naval vessels".
Here's just one example demonstrating that Piantodosi does in fact claim what Chomsky says he does: Piantodosi writes "The success of large language models is a failure for generative theories because it goes against virtually all of the principles these theories have espoused." Rewriting that statement using Chomsky's analogy illustrates how idiotic the original statement is: "The success of naval vessels is a failure for insect navigation theories because it goes against all of the principles these theories have espoused".
The success of naval vessels shows it’s possible to navigate without innate star and wind comprehension, so maybe we should think of that inner stuff as phlogiston. (Yeah, this analogy isn’t as nice but it’s quite hard to translate the nuance of linguistic debate into nautical terms.)
It’s not like it’s an outrageous position. Chomsky’s tradition is quite controversial and is outside of mainstream nowadays. And connectionism is a valid scientific approach.
Where is the research on impossible language that infants can't acquire? A good popsci article would give me leads here.
Even assuming Chomsky's claim is true, all it shows is that LLMs aren't an exact match for human language learning. But even an inexact model can still be a useful research tool.
>That’s highly unlikely for reasons long understood, but it’s not relevant to our concerns here, so we can put it aside. Plainly there is a biological endowment for the human faculty of language. The merest truism.
Again, a good popsci article would actually support these claims instead of simply asserting them and implying that anyone who disagrees is a simpleton.
I agree with Chomsky that the postmodern critique of science sucks, and I agree that AI is a threat to the human race.
It's not infants, it's adults but Moro "Secrets of Words" is a book that describes the experiments and is aimed at lay people.
> Even assuming Chomsky's claim is true, all it shows is that LLMs aren't an exact match for human language learning. But even an inexact model can still be a useful research tool.
If it is it needs to be shown, not assumed. Just as you wouldn't by default assume that GPS navigation tells you about insect navigation (though it might somehow).
> Again, a good popsci article would actually support these claims instead of simply asserting them and implying that anyone who disagrees is a simpleton.
He justifies the statement in the previous sentence (which you don't quote) where he says that it is self-evident by virtue of the fact that something exists at the beginning (i.e. it's not empty space). That's the "merest truism". No popsci article is going to help understand that if you don't already.
The same way that reddit comments arent a formal debate.
Mocking is absolutely useful. Sometimes you debate someone like graham hancock and force him to confirm that he has no evidence for his hypotheses, then when you discuss the debate, you mock him relentlessly for having no evidence for his hypotheses.
> Yet here is Chomsky addressing a lay audience that has no linguistics background
So not a formal debate or paper where I would expect anyone to hold to debate principles.