Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Language is not essential for the cognitive processes that underlie thought

scientificamerican.com

Post reply on HN
Print view View original

111–120 of 424 posts

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

#111

Post by voxl »

danielmarkbruce wrote: post #89
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…

Is it important? To who? Anyone with half a brain is aware that language isn't the only way to think. I can think my way through all kinds of things in 3-d space without a single word uttered in any internal monologue and I'm not remotely unique - this kind of thing is put in all kinds of math and iq'ish like tests one takes as a child.

Before you say things this patiently dumb you should probably wonder what question the researchers are actually interested in and why your average experience isn't sufficient proof.

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

#114

Post by dragonwriter »

aniijbod wrote: post #104

Thought and language are intertwined in ways we don’t fully grasp. The fact that certain cognitive tasks, like comprehension, can proceed without engaging traditional language-related brain regions doesn't mean thought doesn't use language—it just means we might not yet understand how it does. Thought could employ other forms of linguistic-like processes that Fedorenko's experiments, or even current brain-imaging tec…

> The fact that certain cognitive tasks, like comprehension, can proceed without engaging traditional language-related brain regions doesn't mean thought doesn't use language

All other things being equal, its is a reason to provisionally reject the hypothesis that those kinds of thought use language as introducing entities (the ties between those kinds of thought and language) into the model of reality being generated that are not needed to explain any observed phenomenon.

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

#115

Post by gotoeleven »

voxl wrote: post #111
danielmarkbruce wrote: post #89

Earlier quoted context omitted.

Is it important? To who? Anyone with half a brain is aware that language isn't the only way to think. I can think my way through all kinds of things in 3-d space without a single word uttered in any internal monologue and I'm not remotely unique - this kind of thing is put in all kinds of math and iq'ish like tests one takes as a child.

Before you say things this patiently dumb you should probably wonder what question the researchers are actually interested in and why your average experience isn't sufficient proof.

I am 3-d rotating this comment in my head right now

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

#116

Post by fhdsgbbcaA »

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…

My first thought as well - “AGI via LLM” implies that our grey matter is merely a substrate for executing language tasks: just swap out bio-neurons for a few H100s and viola, super intelligence.

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

#117

Post by pineaux »

datameta wrote: post #36
Razengan wrote: post #21

Earlier quoted context omitted.

Can you count without using a "language"? Try it now: Tap your hand on the desk randomly. Can you recall how many times you did it without "saying" a sequence in your head like "1, 2, 3" or "A, B, C" etc? If yes, how far can you count? With a language it's effectively infinite. You could theoretically go up to "1 million 5 hundred 43 thousand, 2 hundred and 10" and effortlessly know what comes next.

I can remember the sequence of sounds and like a delay line repeat that sequence in my head. This becomes easier the more distinguishable the taps are or the more of a cadence variability there is. But if it is a longer sequence I compress it by remembering an analogue like so: doo doo da doo da doo da da doo (reminiscent of morse code, or a kind of auditory binary). Would we consider this language? I think in the co…

I think this is what language is. It's a sequence rememberance system.

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

#118

Post by fhdsgbbcaA »

theptip wrote: post #98
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…

LLM as a term is becoming quite broad; a multi-modal transformer-based model with function calling / ReAct finetuning still gets called an LLM, but this scaffolding might be all that’s needed. I’d be extremely surprised if AI recapitulates the same developmental path as humans did; evolution vs. next-token prediction on an existing corpus are completely different objective functions and loss landscapes.

I asked both OpenAI and Claude the same difficult programming question. Each gave a nearly identical response down to the variable names and example values.

I then looked it up and they had each copy/pasted the same Stack overflow answer.

Furthermore, the answer was extremely wrong, the language I used was superficially similar to the source material, but the programming concepts were entirely different.

What this tells me is there is clearly no “reasoning” happening whatsoever with either model, despite marketing claiming as such.

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

#119

Post by zbyforgotp »

jebarker wrote: post #93
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…

> What this tells us for AI is that we need something else besides LLMs Not to over-hype LLMs, but I don't see why this results says this. AI doesn't need to do things the same way as evolved intelligence has.

Ok, but at least it suggests that this other thing might be more efficient in some ways.

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

#120

Post by awongh »

jebarker wrote: post #93
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…

> What this tells us for AI is that we need something else besides LLMs Not to over-hype LLMs, but I don't see why this results says this. AI doesn't need to do things the same way as evolved intelligence has.

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 LLMs work so well because we had the data lying around already.

Post reply on HN