Rendered at 06:49:33 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
NaiveEmpiricist 2 days ago [-]
Norvig (and Chomsky) are talking at cross purposes.
Norvig confuses engineering for science. If an engineer wants something that works (and guarantees that it will work) then you use the most effective tool available to achieve that goal. No problem there. But there's absolutely no reason to think that tells you anything scientific about the phenomenon you're modeling. For example, aeronautical engineering says very little about how birds fly. And - currently relevant - LLMs learn to use and understand language, but they use millions of times more training data, and millions of times more computing power, than a child does. So an LLM may be a great technology, but it's probably not a good model for how humans learn and use language.
Chomsky OTOH doesn't see any use for statistics in science, at least not in linguistics. Fortunately for the field of linguistics, Chomsky's influence in linguistics is waning and linguists are beginning to use statistical tools to do science.
theptip 1 days ago [-]
> Norvig confuses engineering for science
I don’t think Norvig is confused. He explicitly addresses this point right at the top of the article.
His argument is that successfully engineering a system using a statistical method provides some bits of evidence for the hypothesis that the system itself is statistical.
Which is obviously true, so long as we take care to weigh evidence properly.
> there's absolutely no reason to think that tells you anything scientific about the phenomenon you're modeling
No, this is not how evidence works. If you can accurately model a system with method X and you cannot come up with a model that works with method Y, you should think it’s more likely (but not certain) that X is the actual physical model.
And indeed modern cognitive science (see Predictive Processing) also has a bunch of evidence for the brain as a hierarchical predictor, rather than a rule engine.
To your points - planes are not model organisms of birds, but they do shed insight on the bounds that birds must obey, eg thrust-to-weight etc. And the sample efficiency point really needs to die. The human brain is born with an extremely complex prior baked in. An LLM begins pretraining with almost nothing (just a very simple architecture and a few simple algorithms like attention), and has to learn the basic architecture of cognition from scratch. A better baseline would be how many samples were processed by every animal in our genetic ancestry that had a brain, since brains existed.
skinfaxi 1 days ago [-]
> No, this is not how evidence works. If you can accurately model a system with method X and you cannot come up with a model that works with method Y, you should think it’s more likely (but not certain) that X is the actual physical model.
Could you expand on this point?
swid 20 hours ago [-]
Try to model radioactive decay via a process other than statistical and you will almost certainly fail, since that is our best understanding of it.
I was also going to suggest coin flips, but it’s not quite the same… I think any chaotic system is generally best modeled that way, and if you cared about the effect the system has on other things, it’s unlikely you can do better since measuring the initial state will not be feasible.
hocusPokeUs 15 hours ago [-]
Try to express it in anything but the language of stats you will fail because we haven't invented another language for it; we could but have no reason to with the ubiquitous nature of stats syntax and semantics
Math is not physics and physics is not math. Writing a Superman story does not mean the language and pictures represent physics
That's the difference being highlighted here; stats are not evidence of gravity and electromagnetism; physical phenomena when materials influenced by those forces are relative to local gravity, electromagnetic fields are our evidence
Stats of 1960s demographics did not capture immutable physical phenomena because entropy-> generational churn -> those people are now dead
The languages do not matter; they're just to communicate some ground truth.
Shackling society to the past through careful statistical methods relative to finance and politics to avoid societal change is not a scientific conclusion; there's no reason to use 30-40 years ago as the center of our social universe
The concern with stats consumption is we walk away from a result as if its an immutable forever conclusion as we ignore that gravity and electromagnetism, etc have not stopped changing ground truth; past truths that no longer matter to our next choices
RussianBot9580 3 hours ago [-]
> Writing a Superman story does not mean the language and pictures represent physics
You can write many imagined stories, but one of them will match observed reality. It won't be about an airborne male model who wears his underpants on the outside. But you can write a story that describes reality. And you can describe observable phenomena with statistics. Of course, you can describe plenty of pseudoscientific drivel with statistics too.
theptip 2 hours ago [-]
The key point is quite simple: You can write any retroactive explanation you want. But only some stories are _predictive_. And you can weigh models by how good their predictions are; which was my original point above.
1 days ago [-]
trgn 2 days ago [-]
i agree. but it's uncanny that stochastic generation of language yields a form of recognizable intelligence.
in college, i was somewhat involved in cogsci wrt semantic mapping of linguistic concepts based on human surveys (this was well before LLM/AI-word2vec), and it surprises me today how the mathematical properties of relatedness/distance seem now to drive machine intelligence, something which we did not consider at the time at all.
maybe i'm getting in my metaphysical sentimental stage, a turn a lot of adults make after middle age, but i wouldn't fully dismiss the fact that there may be an equivalence.
NaiveEmpiricist 2 days ago [-]
Indeed. Who'd have thought that mere statistics could do that? Hat's off to them - they made the darn thing work.
There's pretty interesting (I think) research in how effectively it's possible to scale down LLMs. If you're interested, take a look at the BabyLM conferences:
(I have no connection to them.) The work there demonstrates that gap between what we can do now with human-relevant amounts of training text, and what remains to be discovered. Quite a bit, it looks like.
hocusPokeUs 1 days ago [-]
"Mere statistics" is how physical reality works; I don't mean the language, the syntax
I mean shifting influence of gravity relative to an object with enough mass type ratios are everywhere
The language of math started as visual art to depict such ratios; in modern times this ground truth associations is frequently overlooked or never understood to exist in the first place without engagement
Differential diagnosis in medicine is exactly the same process; such processes and their successes are everywhere.
The issue we have in software is the outcome isn't as obvious as jabbing a blood draw needle in a patients eye; we spent a lot of time wasting time on illusory needs
theptip 1 days ago [-]
“Mere statistics” is heavily underselling what the transformer architecture does. Unless you think “mere ion channels” is a sensible way to think about the human brain.
2 days ago [-]
netdevphoenix 2 days ago [-]
> stochastic generation of language yields a form of recognizable intelligence.
Can you use a human language well enough and it not sound intelligent? And no, I don't mean just syntactically correct because semantics is a part of human languages too. I think the astounding thing is that they managed to model a language this well. It's like trying to learn English without a functional understanding of the word "Please".
My take is that a lot of the intelligence we natively ascribe to humans exist in human languages and just like the nutrients in your food, by harnessing a language that possesses latent intelligence, you too can behave like an intelligent person. And how did the nutrients of intellect got in there in the first place? Just like languages evolve nowadays: via contributions from speakers over time. Not unlike soil fertilisation.
crtified 1 days ago [-]
And language itself is so very 'stochastic' too, due to complex subjectivity. Every word floats on a variable sea of individual choice.
For example, if I compose your post as :
--
"Good prose embodies intellect gradually evolved by myriad humans over time, so it wows me to see LLMs create it, though I soon realised that the intelligence is not being created by the LLM but rather is already present in the human-evolved input datasets - the prose itself! - which they statistically arrange."
--
..the particulars of the language are entirely different (and omit some of your chosen nuance) but most readers would receive the same "core truth" from it, the same measure of beneficial information.
Which highlights the great fluidity of language; that is another of the key underlying mechanisms allowing LLMs to flourish.
cgio 1 days ago [-]
I think there is a distinction to draw there between stochasticity in the use and in the evolution of language (which is not clean as they are not independent). The language evolves stochastically, as it does so by being used in exchanges. But the use could be deterministic with the exchanges being random and the function of update based on exchange. I still think there is stochasticity in the use itself, I choose my words, go back and delete and rewrite, explain again in a different way when the audience looks at me unimpressed etc. I just don’t see fluidity as a proof of stochasticity of evolution or the other way around and in fact think the interaction is a quite interesting domain.
crtified 11 hours ago [-]
I present this as by-no-means a complete argument, but a very salient point imo is the notion of the 'accent'. It is the most prominent etymology mutation in language (driving inter-language variance between adjacent cultures), and exists not only in the human but in the animal kingdom. Its basis? The maxillofacial expression of the most dominant speaker(s) within a group, which shapes their vocal inflection and passes to the wider culture by social effect.
Regardless, the fluidity of language is self-evident to an extent : it is partly embodied in the entire concept we call figures-of-speech (a linguistic phrasing can be literal, or sarcastic, or described by metaphor, ...). Partly in the common reality that our dictionaries list multiple possible definitions for many words. Partly in the fact that (many) synonyms exist. Partly in sheer human individuality.
labcomputer 1 days ago [-]
> i agree. but it's uncanny that stochastic generation of language yields a form of recognizable intelligence.
It's funny, I used to think it was complete and utter horseshit when literature professors would say that the meaning of a text is created by the reader, not the author.
Since LLMs have become commonplace, I have completely reversed my opinion on that.
That you see recognizable intelligence in the output of stochastic generation is because you are imposing meaning on it. It is why "claudish" is so hard to read--you are trying to impose meaning on a sequence of words that was stochastically generated, and it takes mental effort to do that.
eru 1 days ago [-]
I'm not so sure about that.
If Claude can write formal mathematical proofs (that a machine can check), or it can write software that works, where's the head of the reader that imposes the meaning?
andrekandre 1 days ago [-]
> you see recognizable intelligence in the output of stochastic generation is because you are imposing meaning on it
i am not an expert in any of this, but intuitively i largely feel the same... the character we see as "claude" is mostly in our heads; its almost like a "super rubber duck" phenomenon or something
broken-kebab 1 days ago [-]
I fail to see what exactly is explained, or clarified by the theory here. Postmodernist philosophers are notorious for writing texts which make readers struggle. Surely, it doesn't mean "Foucault-ese" is somehow more stochastically generated than "Claudese".
jibalt 14 hours ago [-]
Reading what you wrote, I must interpretively attempt to construct a model of what you may have meant and what you may believe, because on the surface it contains multiple claims that are frankly false in my experience (itself a probabilistic world model).
> a sequence of words that was stochastically generated
As was your comment (and mine, and the others on this page and elsewhere).
With Claude I know that there is no source meaning or belief, but it helps to impose Dennett's "intentional stance" and treat it as if there were, up to a point. That there is such a "point" allows be to manipulate it in my attempts to extract factual information implicit in its training data, expressed probabilistically via its stochastic processes.
codeulike 2 days ago [-]
The existence of LLMs is telling us something, surely it means something. I'm not sure we've quite figured it out yet but its probably something along the lines of 'intelligence might not be as complicated as we thought' ... a fourth narcissistic blow to humanity after Copernicus, Darwin and Freud.
peterfirefly 1 days ago [-]
Freud doesn't belong in that group. Freud was a fraud.
20 hours ago [-]
goatlover 1 days ago [-]
We train them on the sum of human knowledge available in digital format, so I'm not sure that's an indictment, and the large models are incredibly complicated to the point it's hard to reverse engineer how they work so well and what they're actually doing to build internal models to predict the next token. Several billion to several trillion parameters is anything but simple.
codeulike 1 days ago [-]
Right, we needed compute on a large and fast scale to make LLMs actually work. But the basic principles are quite simple: train on a huge dataset for long enough, using some fairly simple ML maths but across a very large dimensional space, and then out the other side you get something that can competently use language, and is capable of doing useful work. This is somewhat surprising, no?
goatlover 1 days ago [-]
Yes, it is surprising.
ckcheng 2 days ago [-]
The narcissism might not be in thinking 'intelligence might not be as complicated as we thought’ - the narcissism might be in thinking ‘LLM intelligence is the way we thought’.
trgn 2 days ago [-]
I understand the charge, but it's soemwhat unfair. It's more like we start to see a correspondence which even 20 years ago was barely considered
goatlover 1 days ago [-]
It's also not the first time a breakthrough technology was compared to the human brain/mind, but we also know biologically animal nervous systems are not digital computers, telephone/telegraph networks or hydraulic systems.
codeulike 1 days ago [-]
I'm not saying LLMs work in the same way as us. But a few years ago humans used to be the only competent users of language on the planet. Now its humans + LLMs who are competent users of langauge. This is a big change, and I'm saying that it might be pointing to something deeper about complexity and how approaches to deal with complexity arise in the universe.
trgn 17 hours ago [-]
sure, but its facile to dismiss these breakthroughs as naive metaphors. physicans do understand now many bodily systems can be modeled with the same tools an engineer will model his machines. its not because these abstractions are incomplete, they are wrong.
2 days ago [-]
bigstrat2003 1 days ago [-]
> it's uncanny that stochastic generation of language yields a form of recognizable intelligence.
It most certainly does not. It's obvious to anyone who critically examines LLMs, without letting the hype override sense, that there's zero intelligence there.
PaulHoule 2 days ago [-]
To the cynic, Chomskyism is exactly what Thomas Kuhn warned you about "normal science".
That is, you can apply for grants and write papers in the transformational grammar paradigm endlessly. You can pose and answer questions that make sense within in the system.
Chomskyism was disappointing in two respects: (1) was that it was an inadequate basis for machine understanding of text, (2) was that structuralists were hoping everything in culture could be understood in terms of "... having the structure of language" but syntactic structures weren't all adequate for that task.
(1) and (2) both pointed to Chomskyism being hermetic and not something really useful or inspiring to other fields and maybe just a little less "true".
NaiveEmpiricist 1 days ago [-]
Actually, there have been a number of rounds of paradigm shift (or attempted paradigm shift) in linguistics over the past 70 years. The unique thing about linguistics is that they've been compressed into the lifetime of the same central figure.
Also, I'll mention that there is a lot of empirical knowledge about natural language syntax in generative grammar, even if the details aren't quite right about their theoretical underpinnings. And a lot of them were "low-hanging fruit" waiting for someone to point out how to find them and characterize them. That Chomsky did that is the best argument for his intellectual value - permitting all that "normal science" to happen.
Science doesn't guarantee technological applications, whether or not that science is correct.
Finally, just saying: use of a word like "Chomskyism" makes you sound like one of those creationists who call evolution "Darwinism".
jibalt 13 hours ago [-]
So we can't use the word "creationism" without validating their use of "Darwinism"?
This is wholly fallacious ... you may disagree with what you interpret as the implications of attaching "ism" to Chomsky, but in no way does that equate it to attaching "ism" to Darwin -- each of these attachments and their validity are independent.
BTW, even what you see as pejorative in such an attachment is a cultural interpretation ... e.g., it is far more common among British evolutionary biologists than American to view "Darwinism" neutrally -- Dawkins for one uses the term and describes himself as a Darwinist. See, e.g., https://www.edge.org/conversation/richard_dawkins-chapter-3-...
breppp 1 days ago [-]
The irony is that while (1) is currently dying a terrible death, (2) is on its peak influence of culture and politics
PaulHoule 15 hours ago [-]
Officially structuralism had a crisis around the release of Syntactic Structures and the morphed into post-structuralism which has been through relentless post[n]-modernization and is whatever it is now.
whoanoway 6 hours ago [-]
[dead]
huijzer 2 days ago [-]
> You can pose and answer questions that make sense within in the system.
Yes this is one of my critiques of the academic system (I have finished a PhD and will hopefully never work at a university ever again). Many papers in the introduction start with explaining how the study builds on plenty of previous research, but that alone doesn’t prove that something is a good direction to investigate. If academics were completely honest, many papers would write “We researched topic X because we got funding for it. But since we actually don’t care about X, we just wrote a proposal about our favorite niche topic that only one of the co-authors cares about, namely topic Y, and wrote it such that it sounds like X. Next we show that Y is actually very cool in the hope to convince others that it is cool plus to convince others of the conclusion that our most senior professor already found during the PhD which was 30 years ago. Apart from the professor, we don’t believe in the main argument of the paper but since he got the grant we’ll just go with it in the hope to one day be able to push our own ideas forward”.
jppittma 1 days ago [-]
> Norvig confuses engineering for science. If an engineer wants something that works (and guarantees that it will work) then you use the most effective tool available to achieve that goal. No problem there. But there's absolutely no reason to think that tells you anything scientific about the phenomenon you're modeling.
this distinction is superficial. Scientists use less precise models all the time. Like, you don't solve the schrodinger equation even when you're a scientist doing organic chemistry. You reach for lewis dot models. Science as a discipline is about building better models of the world, not uncovering fundamental truth.
dist-epoch 2 days ago [-]
> LLM may be a great technology, but it's probably not a good model for how humans learn and use language.
but from all models, it's the closest, "all models are wrong, but some are useful"
Calavar 2 days ago [-]
> but from all models, it's the closest
Care to back this up? What evidence do we have to suggest that back propagation, self attention, and byte pair encoding are mechanistically similar to human cognition? Similarity of the outputs says nothing about similarity of the machinery underneath. Turboprop engines tell us nothing about how birds fly - doesn't matter that they successfully transport millions of people around the world.
I would say that when it comes to higher level cognition, logic engines are more human like even if they only have a tiny fraction of the practical utility of LLMs.
adrian_b 2 days ago [-]
> Turboprop engines tell us nothing about how birds fly
That is not true.
Turboprop engines and flying birds are different things, but there is a set of common properties.
A blade of a screw propeller functions exactly in the same way as a flapping wing of a bird, except that the wing cannot rotate without limits around the body of the bird, so it must stop, change angles and revert the motion, while the blade can rotate continuously around the axis, functioning on one side like a falling wing and on the other side like a raising wing. The similarities between propeller blades and insect wings are even greater, because insect wings cannot change their shapes so much as those of birds.
For now, our understanding of the networks composed of the neurons of animals is much less advanced than our understanding of wings and propellers, which is why there is a much greater uncertainty about whether there are any similar mechanisms in biological neural networks and in LLMs.
NaiveEmpiricist 2 days ago [-]
Also, AFAIK, insects rely on their (flat) flapping wings to create lift by creating vortices.
"Nothing" may be a little too strong -- physics is physics, after all -- but (small) birds and airplanes have Reynolds numbers about 1000x different and different underlying material science.
Airplanes are engineered (with a very small number of exceptions) to be stable during flight, whereas bird flight is inherently unstable, requiring a lot of specialized cognition so birds can fly.
I would think a lot of cognitive ornithologists (who I really hope exist!) would be interested in finding out about this.
At this point, I'm at the complete end of my knowledge of bird flight!
suddenlybananas 2 days ago [-]
We already know that the most basic mechanisms in ANN and actual neurons are very different. Spiking and neurotransmitters are basically everything in brains and don't exist in ANNs.
casualrandomcom 2 days ago [-]
ok, compare a wheel to a human leg. If you look long enough you will find similarities, but you honestly cannot say that you "understand how humans walk" when all you can do is put a man in a wheelbarrow, move it, and say "look! my device accomplishes the same task!"
Calavar 24 hours ago [-]
> A blade of a screw propeller functions exactly in the same way as a flapping wing of a bird, except that the wing cannot rotate without limits around the body of the bird
The "except" in your argument is doing an incredible amount of work. The rotation of a turboprops generates forward force, wings (whether a bird's or a plane's) generate lift. And plane's wings are called "wings" because they, and not the turboprop, mimic the structure and function of a bird's wings.
Regardless, this whole line of argument is missing the forest for the trees.
cma 2 days ago [-]
attention + residual seems kind of like Chomsky's merge operation.
randcraw 2 days ago [-]
LLMs behave like they're closest, but only during execution. During training, obviously, they don't resemble the human brain at all. No human reads a million books before they begin to speak.
And I'd be hard pressed to identify many cognitive characteristics of LLMs during execution that resemble human thinking, aside from those that are readily parrotable.
As a model of human cognition, LLMs deep suck.
swid 1 days ago [-]
Humans have billions of years of evolution going for them, that must count…
gizmo686 2 days ago [-]
What reason to we have to believe that LLMs are the closest model we have for how humans learn language.
We considered the hypothesis that humans are a general intelligence that figures language out by exposure. The evidence has consistently come back that that is not how humans actually work. A model where humans are hardwired with a lot of knowledge about language, then fill in the gaps based on exposure still appears to be far more accurate.
The only thing LLMs show is that it is possible to learn language de-novo.
krackers 1 days ago [-]
>closest model we have for how humans learn language
Is there some other competing model which can give results as good? If not, the statement seems true in the trivial sense that it's the _only_ model which we've observed to give good results.
jibalt 13 hours ago [-]
People here aren't talking about results, they are talking about the form of the model. Thus the statement you quoted is about how, not about results or outcomes. And it was about how humans learn language, which isn't a result or outcome at all.
NaiveEmpiricist 2 days ago [-]
> The only thing LLMs show is that it is possible to learn language de-novo.
I agree, but would add, "But not under conditions that humans learn language."
emmelaich 2 days ago [-]
I think it's far too early to say LLMs are not how humans learn language.
Regarding "hard-wired"; it's not if you consider that "hard-wiring" of an individual is the result of millions of years of evolution.
lostmsu 2 days ago [-]
Humans decidedly take much more data than LLMs. Like orders of magnitude. We happen to feed only distilled data to LLMs though. That's why they appear so efficient.
rcxdude 2 days ago [-]
If you are counting by bytes and including all sensory input, sure. In terms of amount of text, it's not close in the other direction. It's very difficult to make an apples to apples comparison here.
lostmsu 1 days ago [-]
You can obviously count by bytes. They are bytes for the model or for a person. Bytes to bytes is apples to apples.
You can't count by text, because text is not the only thing children learn on. So unlike the bytes comparison the text comparison is totally invalid.
jibalt 13 hours ago [-]
What they said, which is true, is that an LLM's inputs measured in bytes of language is much higher than for humans. Throwing in all the non-language bytes of sensory input received by humans is not valid and thus is not apples to apples. Children learning from watching a dog do zoomies around the room or butterflies in the garden is not learning language.
> We happen to feed only distilled data to LLMs though. That's why they appear so efficient.
That is wild handwaving that really explains nothing.
But then, most of the comments on this page are wild handwaving, pontificating like a blind man describing an elephant. Very little engages with or matches the quality of Norvig's arguments.
lostmsu 12 hours ago [-]
> Throwing in all the non-language bytes of sensory input received by humans is not valid and thus is not apples to apples. Children learning from watching a dog do zoomies around the room or butterflies in the garden is not learning language.
Have you ever had kids? Either you never had, you forgot, or you are intentionally deceiving us (and perhaps yourself) to support your point. If you never had kids, I can tell you you have very poor imagination about how they learn language.
1 days ago [-]
suddenlybananas 2 days ago [-]
The BabyLM challenge has shown that multimodal models don't seem to learn in a more data-efficient way with respect to tokens. Adding an unlimited amount of images doesn't help you learn syntactic islands faster.
lostmsu 1 days ago [-]
Which entry are you referring to?
suddenlybananas 22 hours ago [-]
Not a specific entry, but in four years, there's never been a clear advantage of multimodal models and the results have been extremely lackluckster, even when you allow unbounded images alongside a fixed quantity of text.
lostmsu 20 hours ago [-]
> when you allow unbounded images alongside a fixed quantity of text
"This year, we have merged the Multimodal and Interaction tracks into the Strict and Strict-small tracks. This is for two primary reasons: Firstly, there were insufficient submissions to warrant standalone tracks. Secondly, none of the submissions in the Interaction track outperformed models submitted to the Strict track, and *none of the submissions from Multimodal outperformed our baselines*. We will still allow participants to use multimodal data or interactive training procedures if they choose, but such data or learning procedures must conform to the rules of the Strict or Strict-small tracks.
Emphasis mine.
lostmsu 12 hours ago [-]
> none of the submissions from Multimodal outperformed our baselines
I think that's a pretty weak point to your argument. They are saying that none of the contestants outperformed their baseline multimodal models and that there was very little multimodal submissions in the first place.
12 hours ago [-]
stillpointlab 2 days ago [-]
I find it hard to see how Chomsky has a leg to stand on.
> I understand how Chomsky might arrive at the conclusion that probabilistic models are unnecessary, from his study of the generation of language.
I think the point turns here for me. We have probabilistic models now that generate unbelievably convincing language. Norvig's following argument on the value of interpreting language statistically becomes secondary.
In order to stay moored in reality, our models must follow from evidence. I think it would take a peculiar obstinance to deny that the latest models show evidence of generating language.
suddenlybananas 1 days ago [-]
The models require orders of magnitude more data to learn language than children do, and children's task is harder as they have to segment speech themselves.
theptip 1 days ago [-]
A newborn brain is not a blank slate, it comes with a very rich prior (as Chomsky himself would argue), for example the ability to rapidly acquire language, control a body, etc.
An LLM must construct all of its grammatical structure, world models, etc. from scratch during pretraining.
LLM pre-training is more like recapitulating evolution of the mammalian brain than childhood development.
suddenlybananas 22 hours ago [-]
Of course children are not a blank state. My point is to argue that LLMs are not an argument against UG.
cjauvin 1 days ago [-]
The comparison isn't so straightforward, since children's language-learning environment is multimodal and extremely rich. They're exposed not just to sequences of words, but also to behaviors, emotions, social interactions, and the physical world.
NaiveEmpiricist 1 days ago [-]
I think you just explained the situation, even if you didn't realize it.
LLMs process millions of times more text data and use millions of times more compute power to learn a language, mostly because we know how to do that technologically.
Children use much less text data, but also use a lot of other kinds of data, including interacting with other humans, to learn language. We have no idea how that works. And, on top of that, people can interact with pets of various species in the same way as children, but those pets never learn language.
So, something strange is going on that children can do, but that LLMs and puppies can't. Probably, part of it is that human children are much better at learning by copying what adults are doing, and part of it is that humans evolved to learn and use language. The details of what's going on there is what linguistics is about.
andrewflnr 1 days ago [-]
Yes, I believe arguments like GP vastly underestimate the amount of data involved in children's learning. All the data coming into their sensorium, from heard language to physical movement feedback, contributes to training all their skills, from physical movement to language.
stillpointlab 12 hours ago [-]
I agree that is something we need to learn a lot more about, but a human child's brain does not appear to be necessary to learn language and I argue the evidence supports that.
The question isn't "are neural networks just like human brains?" It is "can statistical models generate and understand language?"
The answer to the actual question under discussion appears to be a resounding yes. It does not appear that learning language is innate in the way that Chomsky has argued.
It may be that learning language as efficiently as human children will involve some innate property that we have not yet discovered. But that is a perfect example of moving the goal posts. We go from "there is something special about the brain that allows us to learn language, something that cannot be reproduced statistically" to "there is something special about the brain that allows us to learn language efficiently, more than statistical models".
yed 1 days ago [-]
You're forgetting the millions of years of evolution that "pre-trained" child brains for this task.
suddenlybananas 1 days ago [-]
I agree that language is innate.
yed 1 days ago [-]
Yes but it can be both innate and built on statistical models, contrary to Chomksy.
roenxi 1 days ago [-]
We could innately have built-in statistical models, then everyone can be happy. Given how well statistics seems to work for language acquisition the brain is probably approximating a statistical formula at some level.
jibalt 13 hours ago [-]
A statistical model is approximating the brain at some level.
theptip 1 days ago [-]
Well, Chomsky wouldn’t be happy, that’s the point of OP; he explicitly argues against this.
suddenlybananas 23 hours ago [-]
But the statistics only work well at vastly unrealistic scales of data, and the learning pattern doesn't match children's behaviour at all.
suddenlybananas 1 days ago [-]
The idea that there is some level of statistical inference involved in language acquisition is not contrary to generative thought, look at Charles Yang's work or Lidz or a fair number of other people. Chomsky just thinks this is uninteresting so he is dismissive of it as a central explanation for language acquisition.
yed 1 days ago [-]
The central question remains whether there are categorical rules built into the brain for grammar (plus "some" statistical inference possibly), or whether the entirety of our grammatical processing is statistical. It is entirely possible that evolution encoded statistical priors into our brains with grammar overall being a fully statistical process. Therefore, the fact that children learn more quickly isn't really hard evidence one way or another.
red75prime 1 days ago [-]
LLMs (at the autoregressive pretraining stage) are trained to predict what an (adult) human who could have said preceding words would have said next. You should add 15-30 years to the human training time.
breppp 1 days ago [-]
The hardest theory to disprove is one an entire field livelihood rests upon
mturmon 1 days ago [-]
Peter makes some excellent points in TFA and it’s a fun read.
But the situation has moved on, moving even further beyond the assertions Peter makes about the utility of the black-box generative approach.
In particular, LLMs are telling us a lot about those low-probability, never before seen sentences — it’s called CALU, the creative aspects of language use.
LLMs exhibit CALU; as we know they will happily create such sentences. And inspecting the LLMs does provide insight:
“It does change the epistemic situation substantially. A working system can be inspected, perturbed, retrained, lesioned, and compared with other systems. Building is not understanding, but it gives understanding a strong foothold.
Explaining creative language use in LLMs would not, by itself, explain creative language use in humans. But the artificial case matters even if the mechanisms turn out to be substantially different. At a minimum, LLMs are an existence proof of one pathway by which open-ended, context-sensitive language can arise.”
For much more, see this substack by Chandra Sripada, a U of Michigan linguist:
(Reposts are fine after a year or so; links to past threads are just to satisfy extra-curious readers)
slibhb 1 days ago [-]
Chomsky is incredibly dogmatic; his views and personality have held the field back.
There's enormous irony in Chomsky's mathematicization of linguistics sweeping away the linguists who came before him compared to his own stubborn refusal to acknowledge the fruits of the statistical approach. As far as I know, he's more or less been swept aside.
0xpgm 1 days ago [-]
A field of study being held back by an individual is more of an indictment of the field of study rather than the individual.
Appeal to Authority is over-emphasized in academia in general though.
jibalt 13 hours ago [-]
One can argue that Chomsky's existence has done far more to bring the field forward than to hold it back, even if one only looks at how people have been prompted to challenge Chomsky's views.
Alexander-Barth 2 days ago [-]
> But the vast majority of people who study interpretation tasks (such as speech recognition) quickly see that interpretation is an inherently probabilistic problem: given a stream of noisy input to my ears, what did the speaker most likely mean? Einstein said to make everything as simple as possible, but no simpler. Many phenomena in science are stochastic, and the simplest model of them is a probabilistic model
Why the "simplest" model? Is a probabilistic model not the only model able to deal with uncertainty? What other kind of models are possible to represent uncertainty and inherent ambiguity that are not probabilistic?
bux93 2 days ago [-]
I don't think it's the simplest, actually. Off the top of my mind:
Ternary logic is a super simple model; where true AND unknown equals unknown, and true OR unknown equals true. This deals with uncertainty and ambiguity, just not the probability; notionally you could say everything is 50-50.
Same thing if you have e.g. text-to-speech which returns "the b??? dog" rather than "the (0.8:black, 0.1:blue, 0.1:boy) dog" - no probabilities. Of course, maybe later in the processing you add probabilities back in.
Then there's fuzzy logic, which uses scores that are typically numbers, but those don't express actual probabilities, but more like vibes. They don't necessarily add up to 1, for instance, and you don't apply Bayes' formula to them, or even multiply them, you can use addition too - yolo.
Edit: and also of course backtracking, beam search, that sort of stuff - needn't use probabilities, just heuristics.
Alexander-Barth 2 days ago [-]
That is true, but to me it seems like ternary logic is like probabilities in disguise (false, true and unknown would be probability of 0, 1, and "between 0 and 1"). For most application, the unknown class is too broad, thus we need actual probabilities. But indeed, ternary logic is simpler than working with probabilities.
tgv 2 days ago [-]
Old school search methods allow you to pursue multiple paths, but not necessarily based on probability. At the end (or while pruning) you can use "word knowledge".
Anyway, the probabilistic model doesn't exist. Simple probabilistic models work quite badly on speech recognition (word classification, etc.). A lot of structure needs to be added before they start performing.
pessimizer 2 days ago [-]
My objection is "simpler." What's so "simple" about a probabilistic model? It takes an enormous amount of calculations, so much that you have to predigest incoming information into high-level chunks in order to have any chance of calculating it over small inputs with reasonable effort.
A probabilistic model is the most complicated model. It doesn't even attempt to model the problem, it just looks at the results of an enormous number of tries with different inputs, goes through them all, finds the most similar cases to the current input, and blends them together in a way that carefully avoids any self-contradictory or absurd result.
We see this plainly with LLMs. You train an LLM on a thousand times the (high-level, pre-digested) input that the average human has gotten, and the average human thinks it's a hallucinating moron. It doesn't start to converge on something that passes the smell test to you average human idiot until it has been trained on a million times the input the average human has gotten.
It's not doing what a human does, it's guessing what a human (or any system) might do when confronted with X input. They're not even similar activities, even if they have identical output in the limit. It's like confusing gravity for equations about gravity. Being able to make perfect predictions about the positions of moving, nearby frictionless spherical things in a vacuum at time t doesn't tell us a thing about what gravity is. It tells us how it looks. It should, we distilled it by watching things for a very long time.
> Is a probabilistic model not the only model able to deal with uncertainty?
No, all models deal with uncertainty. The entire point of a model is that it is what you know about a system when you don't know its inputs. The variables in a model are uncertainty. The model is dealing with it.
edit: I think you're interpreting "uncertainty" as a numerical probability. A numerical probability is an artifact of the set of data you're calculating it from, not a thing with independent existence. "Uncertainty" is just a thing that you don't know yet.
If you assume that "uncertainty" is a knowable numeric probability that contributes to some calculation, you're begging the question when you say that the only way you can model it is through probabilistic calculations.
Alexander-Barth 2 days ago [-]
To me uncertainty on a given quantity, is represented by a probability distribution function giving how probable the value of a given quantity is. Loosely, it can also be represented by a confidence interval. It is not intrinsic to the quantity, but how much we know about this quantity.
For the domain I am working in, a model is considered deterministic if for a set of input it always provide the same output and the input are assumed to be perfectly known. A model become probabilistic, once the input or the model itself are considered as uncertain. As an example: previously weather models always provided singe prediction for tomorrow, for example rain or no rain. But now, such models are probabilistic in the sense that an ensemble of model forecast are made (with slightly different input representing our uncertainties) to provide an estimate of the probabilities. Indeed such probabilistic model is not the simplest, but to most complete and so far the only practical way to handle with uncertainty for system like the weather as far as I am aware of.
2 days ago [-]
jibalt 13 hours ago [-]
The statement was that for many STOCHASTIC phenomena in science, the simplest model of them is a probabilistic model. The statement was not that probabilistic models are the simplest models.
(One could make a meta statement here about how faulty some human's modeling of their linguistic inputs is.)
sambejk 2 days ago [-]
Really great post, though a bit old, still relevant. Peter Norvig is one of the better thinkers of our generation.
Have to say that I confronted myself with the fact that I can't read longer posts anymore due to overuse of AI. Guess a detox is required...
crtified 1 days ago [-]
No empirical fact is restricted to a singular vector through human language.
"..if we think of truth as something of granite-like solidity and of personality as something of rainbow-like intangibility and reflect that the aim of [human writing] is to weld these two into one seamless whole, we shall admit that the problem is a stiff one and that we need not wonder if [human writers], for the most part failed to solve it." - Virginia Woolf
[human writing, human writers] = [biography, biographers] in the original quote. My edit is because I infer that her statements key point is that of the biological subjectivity inherent to all human prose.
throwaway27448 14 hours ago [-]
What problem does the concept of an empirical fact solve that isn't solved better with consensus?
crtified 11 hours ago [-]
Given that in human terms, empirical == dominant consensus, you are merely using
a different word for 'primacy' in this context. It makes no difference to the structure.
Because : consensus of what? of sheer numbers of human adherents? of other statistic? historical evidence? of widest agreement by those with higher education in that field? of those with the most money to afford advertising? other? a consensus of multiple consensus(es)?
The dynamic of myriad possible vectors through language to express a given 'truth' remains, whether we define truth as empirical or consensus.
talon8635 2 days ago [-]
Would Chomsky have the name recognition and standing he has, or any at all, without his extracurricular political beliefs?
NaiveEmpiricist 1 days ago [-]
For decades, Chomsky was the most-cited author who still had a pulse. And a lot of that was for his political writing - maybe 2/3rds? But he still has a lot of cites for his intellectual work, in several different fields.
Linguistics, obviously.
Philosophers love him, because he always gives them an argument, and that's all they really care about.
There probably wouldn't even be cognitive science without Chomsky. And psychology is still responding to Chomsky's nuking of behaviorism in the 1950's.
So, there's a lot there, but admittedly most of it is for politics.
jancsika 1 days ago [-]
Without question.
Just look at his Wikipedia entry. He finished his dissertation in 1955, starting teaching at MIT the same year, and continued to publish seminal linguistic works over the next 12 years. That's all before he gained any notoriety whatsoever as a critic of the U.S.'s role in the Vietnam War.
Manufacturing Consent was published in 1988!
Additionally, I strongly doubt Chomsky had any "name recognition" all the way through to the late 90s. Outside of a single episode of Firing Line, where would the general public have heard his views?
jibalt 13 hours ago [-]
Chomsky was a prominent critic of the U.S. invasion of Vietnam and of intellectuals who abdicated their responsibility -- nearly everyone attending universities in the late 1960's knew of him.
That said, as a computer science student I learned of his hierarchy of formal languages before I learned of his politics.
talon8635 8 hours ago [-]
Was Manufacturing Consent about his professional linguistic research, or extracurricular politics?
adamddev1 1 days ago [-]
Every introductory book in theoretical computer science will have a discussion of the Chomsky language hierarchy. It is hugely insightful and helpful.
jibalt 13 hours ago [-]
I was a computer science student and an anti-war activist while at UCLA in the late 1960's. I learned of Chomsky's hierarchy of formal languages before I learned of his politics.
No offense but your question is terribly ignorant, naive, and biased. Just spend a few seconds reading his Wikipedia article.
high_5 23 hours ago [-]
Yup, no regular expressions without his theory.
crab2 15 hours ago [-]
False, we owe them to Kleene, 1956.
A formal theory of finite automatons, both deterministic and non-deterministic, is due to Michael Rabin & Dana Scott, for which they received a Turing prize. 1959.
jibalt 13 hours ago [-]
grep was written by people who knew about regular languages and the Kleene star via Chomsky and via BNF used to describe Algol's syntax.
suddenlybananas 1 days ago [-]
Not with laypeople, but among cognitive scientists, absolutely.
kenferry 1 days ago [-]
I mean, his linguistic work was a very, very big deal
emmelaich 2 days ago [-]
As an aside? The "I before E except after C" works well if the sound makes the long-EE sound. Which is believe was implied in it's original formulation. You obviously don't need the rule for "society" and "science".
Of older English words, there are extremely few genuine exceptions. Seize is one.
jibalt 13 hours ago [-]
weird
emmelaich 6 hours ago [-]
Not the long EE sound
muh_gradle 2 days ago [-]
This is an old, old post by Norvig fyi.
lo_zamoyski 2 days ago [-]
It's more or less the classic clash of two modernist attitudes, that of the rationalist (Chomsky) and the empiricist (Norvig). The intellectual wound this division introduces is longstanding. However, I would say Norvig has the overall advantage.
1. Norvig is correct that observation and fact/data collection are necessary for science. However, Chomsky is correct that the ultimate goal of science is to determine the principles governing something. Ergo: you need the first to do the second. The intellect abstracts from source data, but nothing is in the intellect that is not first in the senses.
2. Chomsky here is uninterested in things like speech - he only cares about the underlying syntax. But syntax - expressed and instantiated in matter - is known through speech, and probabilities presuppose an underlying determinate syntax. Ergo: a complete science of speech is end-to-end, from the "reading" or interpretation of noisy data through syntactic processing all the way through to expression.
3. Chomsky focuses heavily on internal generation. Norvig points out that interpretation is a separate, highly practical challenge. Figuring out what someone means in a noisy, ambiguous world is an exercise in prudence which involves weighing contingencies and probabilities to arrive at the most likely truth. Of course, this can only be done against a presupposed acceptance of the function of speech which is to communicate meaning and what is true (hence why lies work and why they are an abuse of speech).
Probabilistic reasoning is still reasoning. It's reasoning in the face of incomplete information (against a backdrop of presuppositions) which is a component of prudential reasoning.
jibalt 12 hours ago [-]
Science has multiple goals. One of the most important and valuable ones is to make accurate predictions.
Note that the base epistemology of science is inference to the best explanation, not "truth".
suddenlybananas 2 days ago [-]
It's so clear that Norvig knows so little about linguistics as a subject, I mean the suggestion that English is "kinda" a pro-drop language like Spanish is just unserious. There's this kind of lazy criticism of generative work where people hear the vague idea of a theory and critique without even seriously reading anything about it. These are facts that generativists are perfectly aware of and have explanations for.
It's sort of like criticising Galileo's putative experiment about dropping two balls from the leaning tower of Pisa by pointing out that a feather indeed falls slower than a iron ball.
segmondy 1 days ago [-]
He never made claim that English is kinda a pro-drop language. His argument that it's easy to make a binary statement that Spanish is a pro-drop language and English is not, but reality is messy. He shows counter example in English, might not be common, nothing is stopping English from evolving like that. No matter the rules, trying to describe natural languages with rules always breaks down, the exception just grows. I think his argument was well presented, we can't brush off probabilistic models because we don't understand it all when they work so well. This essay is 15 years old way before the LLMs of today. It would be interesting to see his additional take on it.
suddenlybananas 22 hours ago [-]
My point is that is not at all a counter example. It is left-edge deletion which is a distinct phenomenon which has been studied since the 70s. It might look like pro-drop to a naive person but it is fundamentally different. For example, you can delete plenty of things on the left edge that _aren't_ pronouns and you can't delete pronouns if they're not on the left-most edge. These are things that don't occur in pro-drop languages where you can always exclude the pronoun.
do you want a cracker -> you want a cracker?
I think he is dumb -> * I think is dumb
It's a lazy dismissal from someone who doesn't know the relevant facts.
alethic 1 days ago [-]
I agree with you that the criticism of government & binding theory is lazy. However, it is not too far off from how people actually criticize it, wrt. giving example after example that G&B simply handles inelegantly (particularly wrt. Principle C). It is not a nice theory, and that is why no one believes in it anymore. Despite it being deep in every linguist's bones...
Shacharp 1 days ago [-]
[flagged]
b0rsuk 19 hours ago [-]
[dead]
2 days ago [-]
effed3 2 days ago [-]
[dead]
gyanchawdhary 2 days ago [-]
[dead]
fnord77 2 days ago [-]
[dead]
ath3nd 2 days ago [-]
[dead]
itkovian_ 1 days ago [-]
Chomsky was completely incorrect about everything. Says nothing about him personally or his political activities, but academically he was completely, almost incredibly, wrong.
dang 1 days ago [-]
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."
It was a terribly misjudged relationship - though Chomsky was in his late 80s and past it by then - but there is not one iota of evidence that Chomsky "raped kids", and I think that highly implausible. Why make such a childish, unfounded slur?
tartoran 2 days ago [-]
I haven't seen any allegation that Chomsky raped anyone, let alone kids. But Chomsky was quite close to Epstein and some of their leaked conversations tarnish his reputation quite a bit, irreversibly IMO.
jibalt 12 hours ago [-]
The allegation was explicitly made above by eaisshutting.
locallost 2 days ago [-]
I agree there is no proof, but merely calling it misjudged is too forgiving. I have had a lot of respect for Chomsky and his moral stance in life, so this relationship to Epstein was gravely disappointing. He tried to justify it a long time ago by saying Epstein was out of prison and had paid his dues, and that it was intellectual curiosity or something like that, and on paper I agree with this, but it was clear Epstein got off easy and in hindsight it should've been clear to the people around him he had never stopped. I don't think Chomsky raped anyone for what it's worth. But it's a big stain on his legacy.
jibalt 12 hours ago [-]
> in hindsight it should've been clear
Read what you wrote.
suddenlybananas 2 days ago [-]
There is absolutely no evidence or even any allegations whatsoever that Chomsky "raped kids".
IncreasePosts 2 days ago [-]
There's also communications of his where he talks with Epstein about the allegations being ridiculous. Obviously, they weren't, and Epstein knew that, but I think the odds are good that Chomsky actually did think they were fabricated. Epstein knew so many people, it's highly unlikely that he would start chatting about child sexual abuse with everyone. Chomsky probably didn't get that version of Epstein - he just got the Epstein who liked to splash money around, help out famous friends with introductions, etc.
NaiveEmpiricist 2 days ago [-]
I haven't seen the evidence for "raped kids", but it's certainly true that Chomsky advised and benefitted from Epstein long after any sensible person would have realized that Epstein JUST MIGHT be a bad person.
Unfortunately, Chomsky's always had a problem pulling the ripcord in similar situations. As one of his colleagues once explained, "When most people step in dogshit on the sidewalk, they scrape it off their shoe and keep walking. Chomsky gets stubborn and just keeps stepping in it again and again."
Not that this is relevant to his intellectual contributions, unless it crosses your personal line where involvement becomes complicity.
jibalt 12 hours ago [-]
> long after any sensible person
Chomsky is a sensible person, and many other sensible people also continued to engage with Epstein. Hindsight and not being personally involved make judgments easy.
Also, Chomsky was willing to talk to anybody, and I have the personal emails to prove it.
rapnie 2 days ago [-]
His Wikipedia article [0] gives a good summary of the connection they had.
Chomsky is also somewhat unique in that he would communicate with literally anyone who wanted to talk to him. I remember seeing Manufacturing Consent for the first time in 9th grade in 1999, and I sent him an email, and he responded within the hour.
jibalt 12 hours ago [-]
I sent Chomsky an email on 9/11 asking about blowback, mentioning that I had once met him when a group I belonged to took him to breakfast before a speech he gave about Reagan and Nicaragua. He responded on 9/12, apologizing for the delay caused by so many people contacting him. The response was filled with salient information about blowback. I also have some very interesting emails from him criticizing Steven Pinker's depravity, among other rather personal statements.
Norvig confuses engineering for science. If an engineer wants something that works (and guarantees that it will work) then you use the most effective tool available to achieve that goal. No problem there. But there's absolutely no reason to think that tells you anything scientific about the phenomenon you're modeling. For example, aeronautical engineering says very little about how birds fly. And - currently relevant - LLMs learn to use and understand language, but they use millions of times more training data, and millions of times more computing power, than a child does. So an LLM may be a great technology, but it's probably not a good model for how humans learn and use language.
Chomsky OTOH doesn't see any use for statistics in science, at least not in linguistics. Fortunately for the field of linguistics, Chomsky's influence in linguistics is waning and linguists are beginning to use statistical tools to do science.
I don’t think Norvig is confused. He explicitly addresses this point right at the top of the article.
His argument is that successfully engineering a system using a statistical method provides some bits of evidence for the hypothesis that the system itself is statistical.
Which is obviously true, so long as we take care to weigh evidence properly.
> there's absolutely no reason to think that tells you anything scientific about the phenomenon you're modeling
No, this is not how evidence works. If you can accurately model a system with method X and you cannot come up with a model that works with method Y, you should think it’s more likely (but not certain) that X is the actual physical model.
And indeed modern cognitive science (see Predictive Processing) also has a bunch of evidence for the brain as a hierarchical predictor, rather than a rule engine.
To your points - planes are not model organisms of birds, but they do shed insight on the bounds that birds must obey, eg thrust-to-weight etc. And the sample efficiency point really needs to die. The human brain is born with an extremely complex prior baked in. An LLM begins pretraining with almost nothing (just a very simple architecture and a few simple algorithms like attention), and has to learn the basic architecture of cognition from scratch. A better baseline would be how many samples were processed by every animal in our genetic ancestry that had a brain, since brains existed.
Could you expand on this point?
I was also going to suggest coin flips, but it’s not quite the same… I think any chaotic system is generally best modeled that way, and if you cared about the effect the system has on other things, it’s unlikely you can do better since measuring the initial state will not be feasible.
Math is not physics and physics is not math. Writing a Superman story does not mean the language and pictures represent physics
That's the difference being highlighted here; stats are not evidence of gravity and electromagnetism; physical phenomena when materials influenced by those forces are relative to local gravity, electromagnetic fields are our evidence
Stats of 1960s demographics did not capture immutable physical phenomena because entropy-> generational churn -> those people are now dead
The languages do not matter; they're just to communicate some ground truth.
Shackling society to the past through careful statistical methods relative to finance and politics to avoid societal change is not a scientific conclusion; there's no reason to use 30-40 years ago as the center of our social universe
The concern with stats consumption is we walk away from a result as if its an immutable forever conclusion as we ignore that gravity and electromagnetism, etc have not stopped changing ground truth; past truths that no longer matter to our next choices
You can write many imagined stories, but one of them will match observed reality. It won't be about an airborne male model who wears his underpants on the outside. But you can write a story that describes reality. And you can describe observable phenomena with statistics. Of course, you can describe plenty of pseudoscientific drivel with statistics too.
in college, i was somewhat involved in cogsci wrt semantic mapping of linguistic concepts based on human surveys (this was well before LLM/AI-word2vec), and it surprises me today how the mathematical properties of relatedness/distance seem now to drive machine intelligence, something which we did not consider at the time at all.
maybe i'm getting in my metaphysical sentimental stage, a turn a lot of adults make after middle age, but i wouldn't fully dismiss the fact that there may be an equivalence.
There's pretty interesting (I think) research in how effectively it's possible to scale down LLMs. If you're interested, take a look at the BabyLM conferences:
https://babylm.github.io/
(I have no connection to them.) The work there demonstrates that gap between what we can do now with human-relevant amounts of training text, and what remains to be discovered. Quite a bit, it looks like.
I mean shifting influence of gravity relative to an object with enough mass type ratios are everywhere
The language of math started as visual art to depict such ratios; in modern times this ground truth associations is frequently overlooked or never understood to exist in the first place without engagement
Differential diagnosis in medicine is exactly the same process; such processes and their successes are everywhere.
The issue we have in software is the outcome isn't as obvious as jabbing a blood draw needle in a patients eye; we spent a lot of time wasting time on illusory needs
Can you use a human language well enough and it not sound intelligent? And no, I don't mean just syntactically correct because semantics is a part of human languages too. I think the astounding thing is that they managed to model a language this well. It's like trying to learn English without a functional understanding of the word "Please".
My take is that a lot of the intelligence we natively ascribe to humans exist in human languages and just like the nutrients in your food, by harnessing a language that possesses latent intelligence, you too can behave like an intelligent person. And how did the nutrients of intellect got in there in the first place? Just like languages evolve nowadays: via contributions from speakers over time. Not unlike soil fertilisation.
For example, if I compose your post as :
--
"Good prose embodies intellect gradually evolved by myriad humans over time, so it wows me to see LLMs create it, though I soon realised that the intelligence is not being created by the LLM but rather is already present in the human-evolved input datasets - the prose itself! - which they statistically arrange."
--
..the particulars of the language are entirely different (and omit some of your chosen nuance) but most readers would receive the same "core truth" from it, the same measure of beneficial information.
Which highlights the great fluidity of language; that is another of the key underlying mechanisms allowing LLMs to flourish.
Regardless, the fluidity of language is self-evident to an extent : it is partly embodied in the entire concept we call figures-of-speech (a linguistic phrasing can be literal, or sarcastic, or described by metaphor, ...). Partly in the common reality that our dictionaries list multiple possible definitions for many words. Partly in the fact that (many) synonyms exist. Partly in sheer human individuality.
It's funny, I used to think it was complete and utter horseshit when literature professors would say that the meaning of a text is created by the reader, not the author.
Since LLMs have become commonplace, I have completely reversed my opinion on that.
That you see recognizable intelligence in the output of stochastic generation is because you are imposing meaning on it. It is why "claudish" is so hard to read--you are trying to impose meaning on a sequence of words that was stochastically generated, and it takes mental effort to do that.
If Claude can write formal mathematical proofs (that a machine can check), or it can write software that works, where's the head of the reader that imposes the meaning?
> a sequence of words that was stochastically generated
As was your comment (and mine, and the others on this page and elsewhere).
With Claude I know that there is no source meaning or belief, but it helps to impose Dennett's "intentional stance" and treat it as if there were, up to a point. That there is such a "point" allows be to manipulate it in my attempts to extract factual information implicit in its training data, expressed probabilistically via its stochastic processes.
It most certainly does not. It's obvious to anyone who critically examines LLMs, without letting the hype override sense, that there's zero intelligence there.
That is, you can apply for grants and write papers in the transformational grammar paradigm endlessly. You can pose and answer questions that make sense within in the system.
Chomskyism was disappointing in two respects: (1) was that it was an inadequate basis for machine understanding of text, (2) was that structuralists were hoping everything in culture could be understood in terms of "... having the structure of language" but syntactic structures weren't all adequate for that task.
(1) and (2) both pointed to Chomskyism being hermetic and not something really useful or inspiring to other fields and maybe just a little less "true".
Also, I'll mention that there is a lot of empirical knowledge about natural language syntax in generative grammar, even if the details aren't quite right about their theoretical underpinnings. And a lot of them were "low-hanging fruit" waiting for someone to point out how to find them and characterize them. That Chomsky did that is the best argument for his intellectual value - permitting all that "normal science" to happen.
Science doesn't guarantee technological applications, whether or not that science is correct.
Finally, just saying: use of a word like "Chomskyism" makes you sound like one of those creationists who call evolution "Darwinism".
This is wholly fallacious ... you may disagree with what you interpret as the implications of attaching "ism" to Chomsky, but in no way does that equate it to attaching "ism" to Darwin -- each of these attachments and their validity are independent.
BTW, even what you see as pejorative in such an attachment is a cultural interpretation ... e.g., it is far more common among British evolutionary biologists than American to view "Darwinism" neutrally -- Dawkins for one uses the term and describes himself as a Darwinist. See, e.g., https://www.edge.org/conversation/richard_dawkins-chapter-3-...
Yes this is one of my critiques of the academic system (I have finished a PhD and will hopefully never work at a university ever again). Many papers in the introduction start with explaining how the study builds on plenty of previous research, but that alone doesn’t prove that something is a good direction to investigate. If academics were completely honest, many papers would write “We researched topic X because we got funding for it. But since we actually don’t care about X, we just wrote a proposal about our favorite niche topic that only one of the co-authors cares about, namely topic Y, and wrote it such that it sounds like X. Next we show that Y is actually very cool in the hope to convince others that it is cool plus to convince others of the conclusion that our most senior professor already found during the PhD which was 30 years ago. Apart from the professor, we don’t believe in the main argument of the paper but since he got the grant we’ll just go with it in the hope to one day be able to push our own ideas forward”.
this distinction is superficial. Scientists use less precise models all the time. Like, you don't solve the schrodinger equation even when you're a scientist doing organic chemistry. You reach for lewis dot models. Science as a discipline is about building better models of the world, not uncovering fundamental truth.
but from all models, it's the closest, "all models are wrong, but some are useful"
Care to back this up? What evidence do we have to suggest that back propagation, self attention, and byte pair encoding are mechanistically similar to human cognition? Similarity of the outputs says nothing about similarity of the machinery underneath. Turboprop engines tell us nothing about how birds fly - doesn't matter that they successfully transport millions of people around the world.
I would say that when it comes to higher level cognition, logic engines are more human like even if they only have a tiny fraction of the practical utility of LLMs.
That is not true.
Turboprop engines and flying birds are different things, but there is a set of common properties.
A blade of a screw propeller functions exactly in the same way as a flapping wing of a bird, except that the wing cannot rotate without limits around the body of the bird, so it must stop, change angles and revert the motion, while the blade can rotate continuously around the axis, functioning on one side like a falling wing and on the other side like a raising wing. The similarities between propeller blades and insect wings are even greater, because insect wings cannot change their shapes so much as those of birds.
For now, our understanding of the networks composed of the neurons of animals is much less advanced than our understanding of wings and propellers, which is why there is a much greater uncertainty about whether there are any similar mechanisms in biological neural networks and in LLMs.
"Nothing" may be a little too strong -- physics is physics, after all -- but (small) birds and airplanes have Reynolds numbers about 1000x different and different underlying material science.
Airplanes are engineered (with a very small number of exceptions) to be stable during flight, whereas bird flight is inherently unstable, requiring a lot of specialized cognition so birds can fly.
I would think a lot of cognitive ornithologists (who I really hope exist!) would be interested in finding out about this.
At this point, I'm at the complete end of my knowledge of bird flight!
The "except" in your argument is doing an incredible amount of work. The rotation of a turboprops generates forward force, wings (whether a bird's or a plane's) generate lift. And plane's wings are called "wings" because they, and not the turboprop, mimic the structure and function of a bird's wings.
Regardless, this whole line of argument is missing the forest for the trees.
And I'd be hard pressed to identify many cognitive characteristics of LLMs during execution that resemble human thinking, aside from those that are readily parrotable.
As a model of human cognition, LLMs deep suck.
We considered the hypothesis that humans are a general intelligence that figures language out by exposure. The evidence has consistently come back that that is not how humans actually work. A model where humans are hardwired with a lot of knowledge about language, then fill in the gaps based on exposure still appears to be far more accurate.
The only thing LLMs show is that it is possible to learn language de-novo.
Is there some other competing model which can give results as good? If not, the statement seems true in the trivial sense that it's the _only_ model which we've observed to give good results.
I agree, but would add, "But not under conditions that humans learn language."
Regarding "hard-wired"; it's not if you consider that "hard-wiring" of an individual is the result of millions of years of evolution.
You can't count by text, because text is not the only thing children learn on. So unlike the bytes comparison the text comparison is totally invalid.
> We happen to feed only distilled data to LLMs though. That's why they appear so efficient.
That is wild handwaving that really explains nothing.
But then, most of the comments on this page are wild handwaving, pontificating like a blind man describing an elephant. Very little engages with or matches the quality of Norvig's arguments.
Have you ever had kids? Either you never had, you forgot, or you are intentionally deceiving us (and perhaps yourself) to support your point. If you never had kids, I can tell you you have very poor imagination about how they learn language.
Any specific entries?
"This year, we have merged the Multimodal and Interaction tracks into the Strict and Strict-small tracks. This is for two primary reasons: Firstly, there were insufficient submissions to warrant standalone tracks. Secondly, none of the submissions in the Interaction track outperformed models submitted to the Strict track, and *none of the submissions from Multimodal outperformed our baselines*. We will still allow participants to use multimodal data or interactive training procedures if they choose, but such data or learning procedures must conform to the rules of the Strict or Strict-small tracks.
Emphasis mine.
I think that's a pretty weak point to your argument. They are saying that none of the contestants outperformed their baseline multimodal models and that there was very little multimodal submissions in the first place.
> I understand how Chomsky might arrive at the conclusion that probabilistic models are unnecessary, from his study of the generation of language.
I think the point turns here for me. We have probabilistic models now that generate unbelievably convincing language. Norvig's following argument on the value of interpreting language statistically becomes secondary.
In order to stay moored in reality, our models must follow from evidence. I think it would take a peculiar obstinance to deny that the latest models show evidence of generating language.
An LLM must construct all of its grammatical structure, world models, etc. from scratch during pretraining.
LLM pre-training is more like recapitulating evolution of the mammalian brain than childhood development.
LLMs process millions of times more text data and use millions of times more compute power to learn a language, mostly because we know how to do that technologically.
Children use much less text data, but also use a lot of other kinds of data, including interacting with other humans, to learn language. We have no idea how that works. And, on top of that, people can interact with pets of various species in the same way as children, but those pets never learn language.
So, something strange is going on that children can do, but that LLMs and puppies can't. Probably, part of it is that human children are much better at learning by copying what adults are doing, and part of it is that humans evolved to learn and use language. The details of what's going on there is what linguistics is about.
The question isn't "are neural networks just like human brains?" It is "can statistical models generate and understand language?"
The answer to the actual question under discussion appears to be a resounding yes. It does not appear that learning language is innate in the way that Chomsky has argued.
It may be that learning language as efficiently as human children will involve some innate property that we have not yet discovered. But that is a perfect example of moving the goal posts. We go from "there is something special about the brain that allows us to learn language, something that cannot be reproduced statistically" to "there is something special about the brain that allows us to learn language efficiently, more than statistical models".
But the situation has moved on, moving even further beyond the assertions Peter makes about the utility of the black-box generative approach.
In particular, LLMs are telling us a lot about those low-probability, never before seen sentences — it’s called CALU, the creative aspects of language use.
LLMs exhibit CALU; as we know they will happily create such sentences. And inspecting the LLMs does provide insight:
“It does change the epistemic situation substantially. A working system can be inspected, perturbed, retrained, lesioned, and compared with other systems. Building is not understanding, but it gives understanding a strong foothold.
Explaining creative language use in LLMs would not, by itself, explain creative language use in humans. But the artificial case matters even if the mechanisms turn out to be substantially different. At a minimum, LLMs are an existence proof of one pathway by which open-ended, context-sensitive language can arise.”
For much more, see this substack by Chandra Sripada, a U of Michigan linguist:
https://chandrasripada.substack.com/p/ai-and-the-retreat-of-...
He is interviewed by Sean Carroll on his excellent Mindscape podcast:
https://preposterousuniverse.com/podcast/2026/08/10/363-chan...
Chomsky and the Two Cultures of Statistical Learning (2011) - https://news.ycombinator.com/item?id=46286533 - Dec 2025 (119 comments)
On Chomsky and the Two Cultures of Statistical Learning - https://news.ycombinator.com/item?id=35081102 - March 2023 (135 comments)
On Chomsky and the Two Cultures of Statistical Learning (2011) - https://news.ycombinator.com/item?id=16489828 - March 2018 (12 comments)
On Chomsky and the Two Cultures of Statistical Learning (2011) - https://news.ycombinator.com/item?id=11951444 - June 2016 (102 comments)
On Chomsky and the Two Cultures of Statistical Learning - https://news.ycombinator.com/item?id=2591154 - May 2011 (107 comments)
(Reposts are fine after a year or so; links to past threads are just to satisfy extra-curious readers)
There's enormous irony in Chomsky's mathematicization of linguistics sweeping away the linguists who came before him compared to his own stubborn refusal to acknowledge the fruits of the statistical approach. As far as I know, he's more or less been swept aside.
Appeal to Authority is over-emphasized in academia in general though.
Why the "simplest" model? Is a probabilistic model not the only model able to deal with uncertainty? What other kind of models are possible to represent uncertainty and inherent ambiguity that are not probabilistic?
Ternary logic is a super simple model; where true AND unknown equals unknown, and true OR unknown equals true. This deals with uncertainty and ambiguity, just not the probability; notionally you could say everything is 50-50.
Same thing if you have e.g. text-to-speech which returns "the b??? dog" rather than "the (0.8:black, 0.1:blue, 0.1:boy) dog" - no probabilities. Of course, maybe later in the processing you add probabilities back in.
Then there's fuzzy logic, which uses scores that are typically numbers, but those don't express actual probabilities, but more like vibes. They don't necessarily add up to 1, for instance, and you don't apply Bayes' formula to them, or even multiply them, you can use addition too - yolo.
Edit: and also of course backtracking, beam search, that sort of stuff - needn't use probabilities, just heuristics.
Anyway, the probabilistic model doesn't exist. Simple probabilistic models work quite badly on speech recognition (word classification, etc.). A lot of structure needs to be added before they start performing.
A probabilistic model is the most complicated model. It doesn't even attempt to model the problem, it just looks at the results of an enormous number of tries with different inputs, goes through them all, finds the most similar cases to the current input, and blends them together in a way that carefully avoids any self-contradictory or absurd result.
We see this plainly with LLMs. You train an LLM on a thousand times the (high-level, pre-digested) input that the average human has gotten, and the average human thinks it's a hallucinating moron. It doesn't start to converge on something that passes the smell test to you average human idiot until it has been trained on a million times the input the average human has gotten.
It's not doing what a human does, it's guessing what a human (or any system) might do when confronted with X input. They're not even similar activities, even if they have identical output in the limit. It's like confusing gravity for equations about gravity. Being able to make perfect predictions about the positions of moving, nearby frictionless spherical things in a vacuum at time t doesn't tell us a thing about what gravity is. It tells us how it looks. It should, we distilled it by watching things for a very long time.
> Is a probabilistic model not the only model able to deal with uncertainty?
No, all models deal with uncertainty. The entire point of a model is that it is what you know about a system when you don't know its inputs. The variables in a model are uncertainty. The model is dealing with it.
edit: I think you're interpreting "uncertainty" as a numerical probability. A numerical probability is an artifact of the set of data you're calculating it from, not a thing with independent existence. "Uncertainty" is just a thing that you don't know yet.
If you assume that "uncertainty" is a knowable numeric probability that contributes to some calculation, you're begging the question when you say that the only way you can model it is through probabilistic calculations.
For the domain I am working in, a model is considered deterministic if for a set of input it always provide the same output and the input are assumed to be perfectly known. A model become probabilistic, once the input or the model itself are considered as uncertain. As an example: previously weather models always provided singe prediction for tomorrow, for example rain or no rain. But now, such models are probabilistic in the sense that an ensemble of model forecast are made (with slightly different input representing our uncertainties) to provide an estimate of the probabilities. Indeed such probabilistic model is not the simplest, but to most complete and so far the only practical way to handle with uncertainty for system like the weather as far as I am aware of.
(One could make a meta statement here about how faulty some human's modeling of their linguistic inputs is.)
Have to say that I confronted myself with the fact that I can't read longer posts anymore due to overuse of AI. Guess a detox is required...
"..if we think of truth as something of granite-like solidity and of personality as something of rainbow-like intangibility and reflect that the aim of [human writing] is to weld these two into one seamless whole, we shall admit that the problem is a stiff one and that we need not wonder if [human writers], for the most part failed to solve it." - Virginia Woolf
[human writing, human writers] = [biography, biographers] in the original quote. My edit is because I infer that her statements key point is that of the biological subjectivity inherent to all human prose.
Because : consensus of what? of sheer numbers of human adherents? of other statistic? historical evidence? of widest agreement by those with higher education in that field? of those with the most money to afford advertising? other? a consensus of multiple consensus(es)?
The dynamic of myriad possible vectors through language to express a given 'truth' remains, whether we define truth as empirical or consensus.
Linguistics, obviously.
Philosophers love him, because he always gives them an argument, and that's all they really care about.
There probably wouldn't even be cognitive science without Chomsky. And psychology is still responding to Chomsky's nuking of behaviorism in the 1950's.
So, there's a lot there, but admittedly most of it is for politics.
Just look at his Wikipedia entry. He finished his dissertation in 1955, starting teaching at MIT the same year, and continued to publish seminal linguistic works over the next 12 years. That's all before he gained any notoriety whatsoever as a critic of the U.S.'s role in the Vietnam War.
Manufacturing Consent was published in 1988!
Additionally, I strongly doubt Chomsky had any "name recognition" all the way through to the late 90s. Outside of a single episode of Firing Line, where would the general public have heard his views?
That said, as a computer science student I learned of his hierarchy of formal languages before I learned of his politics.
No offense but your question is terribly ignorant, naive, and biased. Just spend a few seconds reading his Wikipedia article.
A formal theory of finite automatons, both deterministic and non-deterministic, is due to Michael Rabin & Dana Scott, for which they received a Turing prize. 1959.
Of older English words, there are extremely few genuine exceptions. Seize is one.
1. Norvig is correct that observation and fact/data collection are necessary for science. However, Chomsky is correct that the ultimate goal of science is to determine the principles governing something. Ergo: you need the first to do the second. The intellect abstracts from source data, but nothing is in the intellect that is not first in the senses.
2. Chomsky here is uninterested in things like speech - he only cares about the underlying syntax. But syntax - expressed and instantiated in matter - is known through speech, and probabilities presuppose an underlying determinate syntax. Ergo: a complete science of speech is end-to-end, from the "reading" or interpretation of noisy data through syntactic processing all the way through to expression.
3. Chomsky focuses heavily on internal generation. Norvig points out that interpretation is a separate, highly practical challenge. Figuring out what someone means in a noisy, ambiguous world is an exercise in prudence which involves weighing contingencies and probabilities to arrive at the most likely truth. Of course, this can only be done against a presupposed acceptance of the function of speech which is to communicate meaning and what is true (hence why lies work and why they are an abuse of speech). Probabilistic reasoning is still reasoning. It's reasoning in the face of incomplete information (against a backdrop of presuppositions) which is a component of prudential reasoning.
Note that the base epistemology of science is inference to the best explanation, not "truth".
It's sort of like criticising Galileo's putative experiment about dropping two balls from the leaning tower of Pisa by pointing out that a feather indeed falls slower than a iron ball.
do you want a cracker -> you want a cracker?
I think he is dumb -> * I think is dumb
It's a lazy dismissal from someone who doesn't know the relevant facts.
https://news.ycombinator.com/newsguidelines.html
Read what you wrote.
Unfortunately, Chomsky's always had a problem pulling the ripcord in similar situations. As one of his colleagues once explained, "When most people step in dogshit on the sidewalk, they scrape it off their shoe and keep walking. Chomsky gets stubborn and just keeps stepping in it again and again."
Not that this is relevant to his intellectual contributions, unless it crosses your personal line where involvement becomes complicity.
Chomsky is a sensible person, and many other sensible people also continued to engage with Epstein. Hindsight and not being personally involved make judgments easy.
Also, Chomsky was willing to talk to anybody, and I have the personal emails to prove it.
[0] https://en.wikipedia.org/wiki/Noam_Chomsky#Friendship_with_J...