Language is not essential for the cognitive processes that underlie thought • Hacker News Board

Re: Language is not essential for the cognitive processes that underlie thought

#261

Post by jstanley » Sun, Oct 20, 2024, 8:42 AM UTC

adrianN wrote: post #238

tsimionescu wrote: post #229

Earlier quoted context omitted.

> Even if we accept the most fringe, anthropocentric theories like Penrose & Hammerhoff’s quantum tubules, that’s just a neural network with fancy weights. First, while it is a fringe idea with little backing it, it's far from the most fringe. Secondly, it is not at all known that animal brains are accurately modeled as an ANN, any more so than any other Turing-compatible system can be modeled as an ANN. Biological n…

It is my understanding that Penrose doesn’t claim that brains are needed for cognition, just that brains are needed for a somewhat nebulous „conscious experience“, which need not have any observable effects. I think that it’s fairly uncontroversial that a machine can produce behavior that is indistinguishable from human intelligence over some finite observation time. The Chinese room speaks Chinese, even if it lacks…

But conscious experience does produce observable effects.

For that not to be the case, you'd have to take the position that humans experience consciousness and they talk about consciousness but that there is no causal link between the two! It's just a coincidence that the things you find yourself saying about consciousness line up with your internal experience?

https://www.lesswrong.com/posts/fdEWWr8St59bXLbQr/zombies-zo...

Re: Language is not essential for the cognitive processes that underlie thought

#262

Post by reverius42 » Sun, Oct 20, 2024, 8:45 AM UTC

Animats wrote: post #48

This is an important result. The actual paper [1] says that functional MRI (which is measuring which parts of the brain are active by sensing blood flow) indicates that different brain hardware is used for non-language and language functions. This has been suspected for years, but now there's an experimental result. What this tells us for AI is that we need something else besides LLMs. It's not clear what that someth…

Interestingly though for AI, this doesn’t necessarily mean we need a different model architecture. A single large multimodal transformer might be capable of a lot that an LLM is not (besides the multimodality).

Re: Language is not essential for the cognitive processes that underlie thought

#263

Post by necovek » Sun, Oct 20, 2024, 8:48 AM UTC

jebarker wrote: post #140

awongh wrote: post #120

Earlier quoted context omitted.

One reason might that LLMs are successful because of the architecture, but also, just as importantly because they can be trained over a volume and diversity of human thought that’s encapsulated in language (that is on the internet). Where are we going to find the equivalent data set that will train this other kind of thinking? Open AI O1 seems to be trained on mostly synthetic data, but it makes intuitive sense that…

I think the data is way more important for the success of LLMs than the architecture although I do think there's something important in the GPT architecture in particular. See this talk for why: [1] Warning, watch out for waving hands: The way I see it is that cognition involves forming an abstract representation of the world and then reasoning about that representation. It seems obvious that non-human animals do thi…

I agree we are not limited with the data set size: all humans learn the language with the much smaller language training set (just look at kids and compare them to LLMs).

OTOH, humans (and animals) do get other data feeds (visual, context, touch/pain, smell, internal balance "sensors"...) that we develop as we grow and tie that to learning about language.

Obviously, LLMs won't replicate that since even adults struggle to describe these verbally.

Re: Language is not essential for the cognitive processes that underlie thought

#264

Post by adrianN » Sun, Oct 20, 2024, 8:51 AM UTC

jstanley wrote: post #261

adrianN wrote: post #238

Earlier quoted context omitted.

It is my understanding that Penrose doesn’t claim that brains are needed for cognition, just that brains are needed for a somewhat nebulous „conscious experience“, which need not have any observable effects. I think that it’s fairly uncontroversial that a machine can produce behavior that is indistinguishable from human intelligence over some finite observation time. The Chinese room speaks Chinese, even if it lacks…

But conscious experience does produce observable effects. For that not to be the case, you'd have to take the position that humans experience consciousness and they talk about consciousness but that there is no causal link between the two! It's just a coincidence that the things you find yourself saying about consciousness line up with your internal experience? https://www.lesswrong.com/posts/fdEWWr8St59bXLbQr/zombie…

That philosophers talk about p-zombies seems like evidence to me that at least some of them don't believe that consciousness needs to have observable effects that can't be explained without consciousness. I don't say that I believe that too. I don't believe that there is anything particularly special about brains.

Re: Language is not essential for the cognitive processes that underlie thought

#265

Post by jll29 » Sun, Oct 20, 2024, 8:57 AM UTC

Animats wrote: post #48

This is an important result. The actual paper [1] says that functional MRI (which is measuring which parts of the brain are active by sensing blood flow) indicates that different brain hardware is used for non-language and language functions. This has been suspected for years, but now there's an experimental result. What this tells us for AI is that we need something else besides LLMs. It's not clear what that someth…

Most pre-deep learning architectures had separate modules like "language model", "knowledge base" and "inference component".

Then LLMs came along, and ML folks got rather too excited that they contain implicit knowledge (which, of course, is required to deal with ambiguity). Then the new aspiration as "all in one" and "bigger is better", not analyzing what components are needed and how to orchestrate their interplay.

From an engineering (rather than science) point of view, the "end-to-end black box" approach is perhaps misguided, because the result will be a non-transparent system by definition. Individual sub-models should be connected in a way that retains control (e.g. in dialog agents, SRI's Open Agent Architecture was a random example of such "glue" to tie components together, to name but one).

Regarding the science, I do believe language adds to the power of thinking; while (other) animals can of course solve simple problems without language, language permits us to define layers of abstractions (by defining and sharing new concepts) that goes beyond simple, non-linguistic thoughts. Programming languages (created by us humans somewhat in the image of human language) and the language of mathematics are two examples where we push this even further (beyond the definition of new named concepts, to also define new "DSL" syntax) - but all of these could not come into beying without human language: all formal specs and all axioms are ultimately and can only be formulated in human language. So without language, we would likely be stuck at a very simple point of development, individually and collectively.

EDIT: 2 typos fixed

Re: Language is not essential for the cognitive processes that underlie thought

#266

Post by fallingknife » Sun, Oct 20, 2024, 9:00 AM UTC

NemoNobody wrote: post #233

tines wrote: post #146

Earlier quoted context omitted.

> As some who has a dis-harmonic intelligence profile, this has been obvious for a very long time. In the family of my mother there are several individuals struggling with language while excelling in the field of exact sciences. I very strongly suspect that my non-verbal (performal) IQ is much higher (around 130) than my verbal IQ (around 100) I used to rationalize to myself along similar lines for a long time, then…

I'm a brilliant genius according to IQ tests. Think me arrogant or conceited or whatever - that is literally the truth, fact - proven many times in the educational system (I was homeschooled and didn't follow any sort of curriculum and was allowed to do whatever I wanted bc I kept testing higher than almost everyone) and just for kicks also - the last time I took an IQ test I was in my late 20s and a friend and I had…

> predicting the future with high accuracy

You can't do this. It's not a matter of IQ, it's a matter of math. Higher order effects are essentially impossible to predict because the level of detail you need to know the initial conditions in is not possible. Even in simple systems where all the rules are known like a billiards table. Furthermore, if you could do this, you would be a billionaire by now just from trading the stock market. This claim alone makes me doubt the rest of your comment.

Re: Language is not essential for the cognitive processes that underlie thought

#267

Post by tsimionescu » Sun, Oct 20, 2024, 9:02 AM UTC

necovek wrote: post #260

tsimionescu wrote: post #232

Earlier quoted context omitted.

> However, it also seems plausible that you could build an abstract representation of the world through studying a vast amount of human language and that'll be a good approximation of the real-world too and furthermore it seems possible that reasoning about that abstract representation can take place in the depths of the layers of a large transformer. While I agree this is possible, I don't see why you'd think it's l…

GP argument seems to be about "thinking" when restricted to knowledge through language, and "possible" is not the same as "likely" or "unlikely" — you are not really disagreeing, since either means "possible".

GP said plausible, which does mean likely. It's possible that there's a teapot in orbit around Jupiter, but it's not plausible. And GP is specifically saying that by studying human language output, you could plausibly learn about the world that have birth to the internal models that language is used to exteriorize.

Re: Language is not essential for the cognitive processes that underlie thought

#268

Post by midtake » Sun, Oct 20, 2024, 9:19 AM UTC

NemoNobody wrote: post #233

tines wrote: post #146

Earlier quoted context omitted.

> As some who has a dis-harmonic intelligence profile, this has been obvious for a very long time. In the family of my mother there are several individuals struggling with language while excelling in the field of exact sciences. I very strongly suspect that my non-verbal (performal) IQ is much higher (around 130) than my verbal IQ (around 100) I used to rationalize to myself along similar lines for a long time, then…

I'm a brilliant genius according to IQ tests. Think me arrogant or conceited or whatever - that is literally the truth, fact - proven many times in the educational system (I was homeschooled and didn't follow any sort of curriculum and was allowed to do whatever I wanted bc I kept testing higher than almost everyone) and just for kicks also - the last time I took an IQ test I was in my late 20s and a friend and I had…

Meta: I have noticed many comments written in this style lately. Long-winded with out-of-place internet shorthand. It is as if someone is deliberately trying to sound youthful while not really knowing how. It is the "how do you do, fellow kids" of writing styles.

I am not sure what sort of LLM-powered bot is behind them, or whether it's one person with some sort of schizophrenia, but once you notice it you will see at least one of these per popular post.

Re: Language is not essential for the cognitive processes that underlie thought

#269

Post by necovek » Sun, Oct 20, 2024, 9:29 AM UTC

tsimionescu wrote: post #267

necovek wrote: post #260

Earlier quoted context omitted.

GP argument seems to be about "thinking" when restricted to knowledge through language, and "possible" is not the same as "likely" or "unlikely" — you are not really disagreeing, since either means "possible".

GP said plausible, which does mean likely. It's possible that there's a teapot in orbit around Jupiter, but it's not plausible. And GP is specifically saying that by studying human language output, you could plausibly learn about the world that have birth to the internal models that language is used to exteriorize.

If we are really nitpicking, they said it's plausible you could build an abstract representation of the world by studying language-based data, but that it's possible it could be made to effectively reason too.

Anyway, it seems to me we are generally all in agreement (in this thread, at least), but are now being really picky about... language :)

Re: Language is not essential for the cognitive processes that underlie thought

#270

Post by psychoslave » Sun, Oct 20, 2024, 9:34 AM UTC

throwaway19972 wrote: post #156

psychoslave wrote: post #74

Earlier quoted context omitted.

Not all human languages exhibits recursion though: https://en.wikipedia.org/wiki/Pirah%C3%A3_language And recursion as the unique trait for human language differentiation is not necessarily completely consensual https://omseeth.github.io/blog/2024/recursive_language/ Also, let's recall that in its broader meaning, the scientific consensus is that humans are animals and they evolved through the same basic mechanism as…

Does any other living entity have recursive grammars? It seems uniquely human. It seems that the second link may indicate otherwise but I'm still pretty skeptical. This requires extraordinary evidence. Furthermore there may be a more practical limit of "stack size" or "context size" that effectively exceptionalizes humans (especially considering the size and proportional energy consumption of our brains).

Does it matter in the frame of investigating relations between cognitive processes and languages?

Other animals have cognitive processes, and languages, or at least it seems to be something scientifically consensual. Thus the surprise reading the kind of statement given in introduction.

Whether humans have exceptional language habilites or even "just" a biggest board to play on with the same basic facilities seems to be a completely different matter.