427
Noam Chomsky, 95, ‘no longer able to talk’ as intellectual’s ‘health deteriorates’
(www.independent.co.uk)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 11 Jun 2024
427 points (98.2% liked)
Philosophy
1230 readers
11 users here now
Discussion of philosophy
founded 1 year ago
MODERATORS
This comment is global-warming-denialism levels of stupid. I’m honestly shocked.
LLM’s have no such implications for the field of linguistics. They’re barely relevant at all.
Do I really need to point out that human beings do not learn language the way LLMs “learn” language? That human beings do not use language the way LLM’s use language? Or that human beings are not mathematical models. Not even approximately. I fucking hate this timeline.
Thank you for saying it. That really was a depressingly incurious comment.
Chomsky, Ian Roberts and Jeffrey Watumull on the topic: https://www.nytimes.com/2023/03/08/opinion/noam-chomsky-chatgpt-ai.html
I'm not saying this from a defending LLMs point, I genuinely think they are a waste of so many things in this world and I can't wait for this hype cycle to be over.
However, there is a lot of research and backing behind statistical learning in language acquisition, this is specifically the research subject of my friends. It's a very big thing in intervention for delays in language.
It is opposed to Chomsky's innate language theory, which at this point I think almost any linguist or language/speech sciences researcher would tell you isn't a well accepted theory (at least as a holistic explanation, certainly it could still be true to an extent and a part of other systems).
tl;Dr LLMs are stupid, but it's not broadly true that the way they "learn language" is entirely different from how humans do. The real difference is that they fail to actually learn anything even when imitating humans.
Be that as it may, we aren’t getting any answers from LLM’s. And given that Universal Grammar was the dominant view for so long, the jury is still out on a viable alternative.
Here’s one relevant discussion.
Here’s another.
This would be true if chomskys claim was that he was simply studying human language acquisition and that machines are different, but his claim was that machines can't learn human languages because they don't have some intuitive innate grammar.
Saying an llm hasn't learned language becomes harder and harder the more you talk to it and the more it starts walking like a duck and quacking like a duck. To make that claim you'll need some evidence to counter the demonstrable understanding the llm displays. Chomsky in his nytimes response just gives his own unprovable theories on innate grammar and some examples of questions llms "can't answer" but if you actually ask any modern llm they answer them fine.
You can define "learning" and "understanding" in a way that excludes llms but you'll end up relying upon unprovable abstract theories until you can come up with an example of a question/prompt that any human would answer correctly and llms won't to demonstrate that difference. I have yet to see any such examples. There's plenty of evidence of them hallucinating when they reach the edge of their understanding, but that is something humans do as well.
Chomsky is still a very important figure and his work on politics with manufacturing consent is just as relevant as when it was written over 20 years ago. His work on language though is on shaky grounds and llms have made it even shakier.
Do you really think Chomsky’s UG hypothesis from half a century ago was formulated to deny that some dumb mathematical model would be able to simulate human speech?
Almost nothing you’ve written has any grounding in empirical reality. You have a sentence that reads something like “the more you talk to LLM’s the harder it is to deny that they can use and understand language.”
You might as well say that the longer you stare at a printed painting, the harder it is to deny that printers make art. LLM’s do not “understand” their outputs or their inputs. If we feed them nonsense, they output nonsense. There’s no underlying semantics whatsoever. An LLM is a mathematical model.
I know it looks like magic, but it’s not actually magic. And even if it were, it would have nothing to do with linguistics, which is concerned with how humans, not computers, understand and manipulate language. This whole ridiculous conversation is a non-sequitur.
Imagine a man in a room
You are like, science denial-ism level of ignorance when it comes to this conversation, or perhaps, it might be that you don't actually understand the underlying philosophy of scientific inquiry to understand why LLMs basically were able to break the back of both UH and innate acquisition.
You seem like the kind of person who cheer-leads "in the spirit of science" but doesn't actually engage in it as a philosophical enterprise. You didn't seem to notice the key point that @Not_mikey@slrpnk.net made, which is that unlike UG, LLM's are actually testable. That's the whole thing right there, and if you don't get the difference, that's fine, but it speaks to your level of understanding to how one actually goes about conducting scientific inquiry.
And if you want to talk about incurious:
Specifically "There’s no underlying semantics whatsoever" is the key lynch pin that UG demands that LLM's demonstrate are not strictly necessary. Its exactly why Chomskys house of cards crumbles with the the counter-factual to UG/ Innate acquisition that LLM's offer. I had a chance to ask him this question directly about 6 months prior to that op-ed being published. And he gave a response that's about as incurious about why LLM, and basically, big complex networks in general, are able to learn as you've offered here. His was response was basically the same regurgitation on UG and innate acquisition that he offers in the op-ed. And the key point is that yes, LLM's are just a big bucket of linear algebra; but they represent an actually testable instrument for learning how a language might be learned. This is the most striking part of Chomskys response and I found it particularly galling.
And it is interesting that yes, if you feed ~~LLM's~~ transformers (I'm going to start using the right term here: transformers) unstructured garbage, you get unstructured garbage out. However, if there is something there to learn, they seem to be at least some what effective at finding it. But that occurs in non-language based systems as well, including image transformers, transformers being used to predict series data like temperature or stock prices, even even DNA and RNA sequences. We're probably going to be having transformers capable of translating animal vocalizations like whale and dolphin songs. If you have structured series data, it seems like transformers are effective at learning patterns and generating coherent responses.
Here's the thing. Chomsky UG represented a monolith in the world of language, language acquisition and learning, and frankly, was an actual barrier to progress in the entire domain, because we now have a counter factual where learning occurs and neither UG or innate acquisition are necessary or at all relevant. Its a complete collapse of the ideas, but its about as close as we'll get because at least in one case of language acquisition, they are completely irrelevant.
And honestly, if you can't handle criticism of ideas in the sciences, you don't belong in the domain. Breaking other peoples ideas is fundamental the process, and its problematic when people assume you need some alternative in place to break someone elses work.
You know what, I’m going to be patient. Let’s syllogize your argument so everyone can get on the same page, shall we.
This argument is not valid, because it’s missing at least one premise. Once you come up with a valid argument, we can debate its premises. Until then, I can’t actually respond, because you haven’t said anything substantive.
The mainstream opinion in linguistics is that LLM’s are mostly irrelevant. If you believe otherwise — for instance, that LLM’s can offer insight into some abstract UG hypothesis about developmental neurobiology — explain why, and maybe publish your theory for peer review.
You don't need to put project a false argument onto what I was saying.
Chomsky's basic arguments:
1: UG requires understanding the semantic roles of words and phrases to map syntactic structures onto semantic structures.
2: UG posits certain principles of grammar are universal, and that syntactic and semantic representation is required as meaning changes with structure. The result is semantic universals - basic meanings that appear across all languages.
3: Semantic bootstrapping is then invoked to explain where children using their understanding of semantic categorizes to learning syntactic structures of language.
LLM's torpedo all of this as totally unnecessary as fundamental to language acquisition, because they offer at least one example where none of the above need to be invoked. LLM's have no innate understanding of language; its just pattern recognition and association. In UG semantics is intrinsically linked to syntactic structure. In this way, semantics are learned indirectly through exposure, rather than through an innate framework. LLM's show that a UG and all of its complexity is totally unnecessary in at least one case of demonstrated language acquisition. That's huge. Its beyond huge. It gives us a testable, falsifiable path forwards that UG didn't.
Largely, because Chomsky. To invoke Planck's principle: Science advances one funeral at a time. Linguistics will finally be able to evolve past the rut its been in, and we now have real technical tools to do the kind of testable, reproducible, quantitative analysis at scale. We're going to see more change in what we understand about language over the next five years than we've learned in the previous fifty. We didn't have anything other than baby humans prior to now to study the properties of language acquisition. Language acquisition in humans is now a subset of the domain because we can actually talk about and study language acquisition outside of the context of humans. In a few more years, linguistics won't look at-all like it did 4 years ago. If departments don't adapt to this new paradigm, they'll become like all those now laughable geography departments that didn't adapt to the satellite revolution of the 1970s. Funny little backwaters of outdated modes of thinking the world has passed by. LLM's for the study of language acquisition is like the invention of the microscope, and Chomsky completely missed the boat because it wasn't his boat.
Your conclusion (which I assume is implied, since you didn’t bother to write it anywhere) might be something like,
My issue is that I just don’t see how to draw this conclusion from your premises. If you were to reformulate your premises into a valid argument structure, we can discuss them and find some common ground.
You haven't demonstrated that you have any real comprehension of the domain, or that you bring anything interesting enough to this conversation to warrant furtherance.
Harsh words for someone who can’t even state a valid argument. I mean do you expect me to guess how your conclusion comes from your unrelated premises?
I've been both cogent and clear as to what my points are, and you've made none. You are a joke if you think yourself an intellectual.
You cogently failed to produce a valid argument. I can’t even engage with your claims because they are unrelated to your conclusion.
The saddest thing about your responses, in spite of their multiple edits, is that you think you are actually serious in whatever it is you think you are doing.
Its disappointing because you can't actually do this thing which you wish you were capable of. You can only imitate it, and in doing so, you mock both yourself and the thing you appear to revere so much.
You could just actually engage with the points being made, but I think we both know you aren't capable. So you resort to self-fellatio. And its sad, because its not just you, but an entire generation of pseudo-intellectuals who almost know how to have a complex discussion on difficult topics. But when your favorite comic book hero gets called out for pushing a unfalsifiable theory, that basically held the field captive for 50 years, you get all tied up in knots. Its because you aren't actually engaging with the material intellectually, but emotionally.
I’m not sure how I confused you so much, or why you would find the request for a syllogized version of your sweeping theoretical claim surprising.
There is nothing confusing about someone as simple as you. You don't understand Chomsky, you don't understand LLM's, and you don't even really understand the conversation we're having now. You don't engage with the points people are actually making, just the ones you wish they made.
You're just a sad little jack off.
"this thread is a wild fucking ride"
-chompsky deepfake
"I smoke two joints after I smoke two joints, and then I smoke two more"
-gnome chimpskie
Premise 1 (UG): Human languages share an underlying structure.
Premise 2: LLM's can be trained to use human languages without the need for any underlying structure. Such an underlying structure is unnecessary for language acquisition.
Therefore, (UG) is false.
Not bad for a first attempt. Unfortunately, the argument above is assuming the consequent. Just because it is not necessary for something to be true, doesn't mean it isn't.
Let's try again.
Premise 1 (UG): Human languages share an underlying structure.
Premise 2: LLM's can be trained to use human languages without the need for any underlying structure. Such an underlying structure is unnecessary for language acquisition.
Premise 3: Human minds and brains operate in a manner relevantly similar to LLM's, at least when it comes to language acquisition.
Therefore, (UG) is false.
This argument is (almost) valid.
Of course, I would push back on both premise 2 and premise 3, which are difficult to believe and would require a lot of empirical evidence, but at least we’re clear about our claims.
The only thing that's clear about this conversation is that you engaged on a topic you know almost nothing about beyond some cursory googling. Go actually read Chomsky on this. Then read the follow ups. Then read Elman, and Fisher, and Vernes. I'd say to also read Sampson but he's a racist fuck nugget, so fuck that guy.
You should actually read Chomsky before commenting any further, because its very very clear you haven't. UG as a theory is relegated to pseudoscience, and Chomsky did a disservice to the entire field pushing theory that had been demonstrated to be false, repeatedly. Even that NYT op-ed was the same pushing of his wrong theories, which were baseless. His preeminence in the field prevented forced research into an unfounded and wrong direction for five decades.
That doesn’t sound like “thank you.” Regardless, feel free to adopt some version of the syllogism above, so that you can have more honest conversations with people in the future. Clarity is important if you’re interested in getting to the truth. It’s a shame we wasted so much time and now can no longer debate the actual premises of your dumb argument. That’s the fun part, after all.
Like I said, the only thing clear in this conversation is that you know actually nothing about anything in it. You don't know what Chomsky is saying because you haven't read him. You are stuck in this weird little self sucking loop where you think you are making a point, but you aren't. You don't know Chomsky, you don't know language, language theory, or computer science. You should just be apologizing for your ignorance and saying that you'll do the work to be taken seriously next time.
The only thing here that's a shame is you.
Your responses are so unhinged. Is your ego that fragile? What’s the point? Why not just learn from your mistake and do better next time?
Such a long post that could be summarized by "I don't like you".
Right or wrong, we had to wait 10 paragraphs for a concise version of your argument, which your opponent made for you. And, without the fog of obscure language, your point isn't the slam dunk you think it is, and that's driving you mad.
And now, you're just insulting them. Let me tell you, if there's someone that sounds like a pseudo-intellectual buffoon here, it's you.
Forget about defining learning. Just look at what "language" means in the dictionary. There are definitions that put an emphasis on using language to communicate thoughts or feelings, which LLMs don't do.
Which definition of language is Chomsky referring to?
I'm not trying to be picky, but I think the answer might help determine whether or not LLMs are proving Chomsky wrong (which he might be, I don't know!)