zlacker

>>mef+(OP)
The fact that we have figured out how to translate language into something a computer can "understand" should thrill linguists. Taking a word (token) and abstracting it's "meaning" as a 1,000-dimension vector seems like something that should revolutionize the field of linguistics. A whole new tool for analyzing and understanding the underlying patterns of all language!

And there's a fact here that's very hard to dispute, this method works. I can give a computer instructions and it "understands" them in a way that wasn't possible before LLMs. The main debate now is over the semantics of words like "understanding" and whether or not an LLM is conscious in the same way as a human being (it isn't).

>>caliba+cd
Restricted to linguistics, LLM's supposed lack of understanding should be a non-sequitur. If the question is whether LLMs have formed a coherent ability to parse human languages, the answer is obviously yes. In fact not just human languages, as seen with multimodality the same transformer architecture seems to work well to model and generate anything with inherent structure.

I'm surprised that he doesn't mention "universal grammar" once in that essay. Maybe it so happens that humans do have some innate "universal grammar" wired in by instinct but it's clearly not _necessary_ to be able to parse things. You don't need to set up some explicit language rules or generative structure, enough data and the model learns to produce it. I wonder if anyone has gone back and tried to see if you can extract out some explicit generative rules from the learned representation though.

Since the "universal grammar" hypothesis isn't really falsifiable, at best you can hope for some generalized equivalent that's isomorphic to the platonic representation hypothesis and claim that all human language is aligned in some given latent representation, and that our brains have been optimized to be able to work in this subspace. That's at least a testable assumption, by trying to reverse engineer the geometry of the space LLMs have learned.

>>kracke+AG
Can LLMs actually parse human languages? Or can they react to stimuli with a trained behavioral response? Dogs can learn to sit when you say "sit", and learn to roll over when you say "roll over". But the dog doesn't parse human language; it reacts to stimuli with a trained behavioral response.

(I'm not that familiar with LLM/ML, but it seems like trained behavioral response rather than intelligent parsing. I believe this is part of why it hallucinates? It doesn't understand concepts, it just spits out words - perhaps a parrot is a better metaphor?)

>>0xbadc+GS
>Can LLMs actually parse human languages?

IMHO, no, they have nothing approaching understanding. It's Chinese Rooms[1] all the way down, just with lots of bell and whistles. Spicy autocomplete.

1. https://en.wikipedia.org/wiki/Chinese_room

>>GolfPo+8X
Go ask the operator of a Chinese room to do some math they weren't taught in school, and see if the translation guide helps.

The analogy I've used before is a bright first-grader named Johnny. Johnny stumbles across a high school algebra book. Unless Johnny's last name is von Neumann, he isn't going to get anything out of that book. An LLM will.

So much for the Chinese Room.

>>Camper+pX
> Go ask the operator of a Chinese room to do some math they weren't taught in school, and see if the translation guide helps.

That analogy only holds if LLMs can solve novel problems that can be proven to not exist in any form in their training material.

>>jmb99+jY
They do. Spend some time using a modern reasoning model. There is a class of interesting problems, nestled between trivial ones whose answers can simply be regurgitated and difficult ones that either yield nonsense or involve tool use, that transformer networks can absolutely, incontrovertibly reason about.

>>Camper+FY
Reason about: sure. Independently solve novel ones without extreme amounts of guidance: I have yet to see it.

Granted, for most language and programming tasks, you don’t need the latter, only the former.

>>hypera+711
99.9% of humans will never solve a novel problem. It's a bad benchmark to use here

>>Workac+h91
But they will solve a problem novel to them, since they haven't read all of the text that exists.