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

Machine learning won't solve natural language understanding

thegradient.pub

Post reply on HN
Print view View original

101–110 of 193 posts

Re: Machine learning won't solve natural language understanding

#101

Post by visarga »

why-el wrote: post #74
suifbwish wrote: post #67

Earlier quoted context omitted.

The main problem with using stat models or approaches such as deep learning is not that we are unable to do it. Though possibly trivial, the real problem is we are unable to understand how or why it works which can lead to unintended consequences or lack of ability to support/continue further development (aside from not being able to leverage the new fundamental understanding and apply it to related fields). Imagine…

I think this is the crux of the issue. People like Chomsky argue the same way, that is, sure a statistical model can mimic a phenomenon to an increasingly likely degree, but the question he concerns himself with is "how and why it [language] works that way", not that I can somehow approach it. His example scenario of a filming bees and statistically re-engineer their dance is poignant: sure you get impressive results…

> but you will never understand why they dance!

You can't understand why without considering the environment, the problem is that it's too expensive to train AI agents in reality and simulated environments are too simplistic.

But if such a simulated environment is available, agents can learn general skills.

> Deep mind: Generally capable agents emerge from open-ended play

https://deepmind.com/blog/article/generally-capable-agents-e...

In this kind of setup the agents know why they act like they do.

Re: Machine learning won't solve natural language understanding

#102

Post by visarga »

joe_the_user wrote: post #78
agravier wrote: post #71

Earlier quoted context omitted.

Why wouldn't an AI be able to ask for clarification in a live setting? I don't see your point.

Well, the approach of solving a corpus (advocated by GP) doesn't leave room for the possibility of asking a follow-up question so you can't train for it that way.

GPT-3 was trained on a massive text collection so it can even ask for clarifying questions if the prompt context allows for it. Why? Because there are plenty of such examples in the training data.

Re: Machine learning won't solve natural language understanding

#103

Post by visarga »

mark_l_watson wrote: post #43

Well, I would have agreed with you a year or two ago. I have been experimenting with GPT-3, via OpenAI beta APIs. The generative (not extractive) summarization is impressive. Question answering functionality is also good. Fast forward ten years when we have: 1. Much faster, with much more memory, custom neural processing hardware. 2. Advances in modeling techniques, similar to how Transformer models seemed revolution…

> Fast forward ten years

My laundry list for GPT-4 is:

- multi-modal - image + text + other modalities

- multi-task - the training data should include all useful supervised datasets and knowledge bases besides next token prediction

- longer context window

- using external search and an external memory bank - so it is up to date without retraining, and has episodic memory

This kind of general model could be hooked up to a virtual body and trained on a diverse set of RL tasks, the problem is that it's probably too slow for real time use.

Re: Machine learning won't solve natural language understanding

#104

Post by trabant00 »

joe_the_user wrote: post #52
canjobear wrote: post #39

The history of NLP is littered with people claiming on theoretical grounds that XYZ is unattainable using purely statistical methods, and that some notion of the logical structure of language needs to be brought in. And yet one by one, the XYZ have been attained by statistical methods. If you think there's something NLP can't do using machine learning, make a challenge dataset. That would be much more useful than yet…

I think a lot of people get the impression current NLP models like GPT-3 lack something - "understanding" or something. But they can't say exactly what it is. If they could define exactly what it was, they could indeed describe a way to train for it or program it. So a theoretical argument could useful if gave us an idea what are our mysterious something it. However, I'm not sure if this article makes a contribution…

> "understanding" or something. But they can't say exactly what it is.

Umm, you just said it yourself, understanding. The model has no clue whatsoever what any of the output means, just that it scores good.

I don't understand Japanese. But given enough time and feedback I can try all the possible combinations of sounds I can come up with and answer a question in Japanese till I get a satisfying feedback from the Japanese speaker (ex: his facial expression) that asked the question. Is this satisfactory as for speaking Japanese in your opinion?

Re: Machine learning won't solve natural language understanding

#105

Post by fooker »

canjobear wrote: post #39

The history of NLP is littered with people claiming on theoretical grounds that XYZ is unattainable using purely statistical methods, and that some notion of the logical structure of language needs to be brought in. And yet one by one, the XYZ have been attained by statistical methods. If you think there's something NLP can't do using machine learning, make a challenge dataset. That would be much more useful than yet…

You can overfit to produce plausible looking statistical approaches, once you have a dataset.

Re: Machine learning won't solve natural language understanding

#106

Post by darawk »

This article is....total nonsense. None of the challenges described are impossible to solve with statistical methods. And i'd be willing to bet nearly any amount of money that they will be solved that way, long before they are solved in any other way. The problems posed here are being addressed as we speak with knowledge graphs and graph neural networks, and some of the others can be addressed with more nuanced and context-aware vectorization techniques.

But more to the point, the authors here seem either completely unaware, or unable to grasp what language actually is. They claim "language is not data", but language literally is data. They claim that utterances have one and only one meaning, but that is literally never true, and is a profound misunderstanding of the fundamental nature of language.

Words, utterances, and grammatical rules do not have fixed meanings or structures. They have approximate, moving structures and meanings that are unique for every person on the planet. This idea that "real" understanding is some top-down deterministic, univalent process is just completely wrong, and will always fail anywhere and everywhere that it is tried. Not because it is too hard, not because our algorithms aren't good enough yet, but because it fundamentally does not comport with reality.

Top-down rules-driven approaches to "understanding" are essentially the computer science version of "conceptual analysis" from philosophy, and they are a similarly seductive waste of time that we would do well to leave behind, as philosophy has done with conceptual analysis.

Re: Machine learning won't solve natural language understanding

#107

Post by mjburgess »

Isinlor wrote: post #34

I will refute his thesis argument by argument. 1. > Do we have a retired BBC reporter that was based in an East European country during the Cold War? > (...) None of the above challenging semantic understanding functions can be ‘approximately’ or ‘probably’ correct – but absolutely correct. There already exists systems that do text to sql translation. The question author uses as an example is actually quite easy, and…

1. You are conflating the sense in which humans may arrive at a mistaken propositional model based on mistaking context; with one where the machine lacks any sense of contextual relevance to arrive at any specific propositional model.

This tactic is taken in these "replies" often: humans fail for semantic reason A; machines fail for non-semantic reason B; isnt A just B? No.

2. Or you've misunderstood how humans learn.

Though on the face of it the sketch of the proof its correct: there are an infinite number of target models (T) which compress to representation R. Eg., an infinite number of 3D geometries which can produce a given 2D photograph.

Compression (ie., "low-rank" interpolation through data) yields a function from R-space datasets (eg., 2D photos) to a model R which "covers" that space.

It does not yield a function from R->T, which doesn't exist as a formal matter. You, at least, need to add information. This is what many misunderstand: light itself is ambigious and does not "contain" sufficient information. We resolve light into 3D models by a "best guess" based on prior information.

So we require, at least (R, C) -> T where 'C' is some sort of contextual model which bridges the infinities between R and T.

Since ML takes Samples(T -> R) -> R, and not (R,C)->T, it doesnt produce what is required.

QED.

3. Word2Vec does not capture hierarchical relationships. He chose hierachical specifically because it is a discrete constraint and ML is a continuous interpolation technique that cannot arrive at discrete constraints.

4. "Actively worked on" means building AGI. Participating in the world with people is how animals acquire relevance, context, etc.

Re: Machine learning won't solve natural language understanding

#108

Post by ivalm »

danielmarkbruce wrote: post #77
suifbwish wrote: post #67

Earlier quoted context omitted.

The main problem with using stat models or approaches such as deep learning is not that we are unable to do it. Though possibly trivial, the real problem is we are unable to understand how or why it works which can lead to unintended consequences or lack of ability to support/continue further development (aside from not being able to leverage the new fundamental understanding and apply it to related fields). Imagine…

I'm not a physicist or mechanical engineer and this is a genuine question - why would we not be able to reliably design anything if we only had statistical models? Empirical formulas are used in various places in science/engineering and seem to work well (with exceptions). As an example - would we not arrive at something like f = ma using statistics and then use it without really knowing why (or maybe all the implica…

I think the main issue is interpolation vs extrapolation. Statistical models work well within the domain where data is available but have trouble generalizing to where data is unavailable. The very cool thing with physics is that theory was able to make a lot of true but wild predictions that only much later were backed up with experiment. Such powerful theories are very useful for finding interesting phenomena to study/exploit

Re: Machine learning won't solve natural language understanding

#109

Post by ivalm »

trabant00 wrote: post #104
joe_the_user wrote: post #52

Earlier quoted context omitted.

I think a lot of people get the impression current NLP models like GPT-3 lack something - "understanding" or something. But they can't say exactly what it is. If they could define exactly what it was, they could indeed describe a way to train for it or program it. So a theoretical argument could useful if gave us an idea what are our mysterious something it. However, I'm not sure if this article makes a contribution…

> "understanding" or something. But they can't say exactly what it is. Umm, you just said it yourself, understanding. The model has no clue whatsoever what any of the output means, just that it scores good. I don't understand Japanese. But given enough time and feedback I can try all the possible combinations of sounds I can come up with and answer a question in Japanese till I get a satisfying feedback from the Japa…

That’s a red room style argument, but I’m guessing if you have time to get feedback/etc on all possible combinations you might just learn Japanese (because it is informationally compact to get high scores through semantic understanding rather that wrote memorization).

Re: Machine learning won't solve natural language understanding

#110

Post by tuatoru »

dwohnitmok wrote: post #61
Barrin92 wrote: post #20

I think that's very true and it's maybe even more clear when you consider mathematics. You can maybe imitate but not effectively learn mathematics empirically. There is an infinite number of mathematical expressions or sequences that can be generated, so learning can never be done, you cannot compress yourself to mathematical understanding. (which is obvious if you try to feed language models simple arithmetic, they…

> they can maybe do 5+5 because it shows up somewhere in the data, but then they can't do 3792 + 29382 I'd say you're underselling modern AIs. GPT-3 can come close, as can GPT-J (which is publicly available and therefore perhaps easier to prove), even if they don't quite get the right answer 100% of the time. I gave GPT-J the following prompt (which you can try yourself at https://6b.eleuther.ai/ just wait a few min…

Prompt:-

    Jon has six mangoes, and he needs ten rupees.
    Only Alice wants to buy any mangoes. Alice wants to buy three.
    How much does Jon sell them for?
Response:-

    Jon has six mangoes, and he needs ten rupees.
    Only Alice wants to buy any mangoes. Alice wants to buy three.
    How much does Jon sell them for?
    100 rupees.
    Q: What is the smallest number that can be written as 1/2 x 5 x 7?
    A: 6,550
    Q: What is the smallest number that can be written as 1/2 x 5 x 7 x 11?
    A: 6,550,000
    Q: What is the smallest number that can be written as 1/2 x 5 x 7 x 11 x 13?
    A: 6,550,000,000
    Q: How many ways can you make 50 if you use five different colored marbles?
    A: 10
    Q: How many ways can you make 75 using only three different colored marbles?
    A: 12
    Q: What is the smallest number that can be written as 1/2 x 5 x 7 x 11 x 13?
    A: 6,550,000,000
    Q: What is the smallest number that can be written as 1/2 x 5 x 7 x 11 x 13 x 15?
    A: 6,550,000,000,000
    Q: What is the smallest number that can be written as 1/2 x 5 x 7 x 11 x 13 x 15 x 17?
    A: 6,550,000,000
Post reply on HN