6 ms·

>Yes, it's completely possible that we get to some sort of limit and the scaling laws fall down, loss stops dropping even with more data, parameters, compute. B

>Yes, it's completely possible that we get to some sort of limit and the scaling laws fall down, loss stops dropping even with more data, parameters, compute. But we have zero reason to believe that is the case, currently, it's only hypothetical [Edit:- and not really a good hypothetical, because there's no theoretical backing that explains current progress but predicts a future discontinuity]. In terms of what we've experimentally seen so far, it just keeps getting better as you scale them up, so to find out if you're right and there's some future roadblock we don't know about we just have to keep scaling these things to see if it works. Also I have no idea why you think Occam's Razor supports "There will be a discontinuity in this currently accurate observed trend, near enough in the future to be relevant" when the trend is talking about the effects of an algorithm applied to data. By analogy, it's something like claiming Moore's Law won't hold because Occam's Razor suggests at some point there's a physical breakdown, except instead of saying it now you're saying it in 1978, & you're saying it about an algorithmic output rather than a physical thing with an obvious limit (atomic size). Once we see any evidence of all of such a discontinuity, we can adjust what we're doing based on that. Until then, if we want to know whether scaling laws hold we should scale the models.

tsimionescu 3y ago
First of all, there is no theory at all which explains the current progress of LLMs. It's not even a quantifiable progress - it's clearly there, but no one can say with any true precision whether GPT2 to GPT3 to GPT4 is following linear or exponential or whatever else precise function of progress in capabilities. We can clearly see that they are improving, but we have no idea how to accurately quantify this space. Secondly, we can see that there are obvious quantifiable physical increases in the requirements of training and inference for GPT4 compared to GPT3 and to GPT2. So, we can absolutely make a physical argument: at some point, you'll need too much energy to keep scaling up GPTs for it to be feasible, even if capability would keep improving with model size. Thirdly, the GPTs require ever more data to train, and there is a limited amount of text data in the world, of which they are already using a significant percent. So, there is also an argument that we might run out of data to improve even before we run out of physical resources. And making an argument that Moore's law will stop at some point due to physical constraints was just as correct in 1978 as it is in 2024. Predicting when that is is obviously hard - and the same is true for LLM scaling. We may be at the peek, or there may be 100 more generations before we even approach the limits I'm contemplating. But the limits are certainly there.
- lucubratory 3y ago
  This conversation started with someone saying that despite LLMs having a theoretical basis for how they could emulate human intelligence, they won't actually reach that point because the existing scaling laws won't hold. I am pointing out that we don't know that at all, scaling laws are so far fantastically accurate, and there is no sign of undershooting existing those laws so far. Therefore, it's very premature to declare that LLMs will be unable to reach any given milestone because scaling laws will hit a discontinuity before we get there. Basically: "There isn't even a theoretical basis for how LLMs could achieve human level intelligence" "There is, a neural network trained to predict the next token isn't restricted in the algorithms it can implement internally to achieve that goal, and eventually for loss to be low enough you need to successfully emulate human minds. We have reason to believe loss will continue to go lower because we have very accurate scaling laws which predict that & haven't been wrong so far." "But an LLM won't be able to reach that point because the scaling laws will break down." "Maybe, there's no evidence for that yet but we should scale them up if we want to find out for sure." You are here -> "But in the physical extreme LLMs will break down as an architecture." To which my response is: Yes, of course this specific architecture will break down in the physical extreme. By the time something is building computronium to run the next AI I strongly doubt it will be running matrix multiplication, let alone LLMs specifically. There is no guarantee or reason to believe that we will hit those limits before we are able to achieve human level intelligence or more; the only reason to believe it will happen is an empirically observed trend which could totally break down, so I wouldn't bet the house on it happening, but it's wrong to just rule out the outcome.
  - tsimionescu 3y ago
    No, I am at the point where I'm saying the scaling "laws" will break down. Especially since there are no scaling laws, it's an empirical observation based on essentially 3 examples - GPT2 to GPT3 to GPT4 (and their rough equivalents from Facebook and Google and a handful of others). This all happened in the span of 4 years or less. We don't even fully know how much larger GPT4 is compared to GPT3. Not to mention, we already know OpenAI has spent more than a hundred millions of dollars training GPT-4, and they are quite probably selling it at a loss (each query may well cost them more in compute power than they charge for it). So, if GPT4 is ten times bigger than GPT3(.5) as sometimes reported, economics suggest they may already not be able to scale GPT5 to the same extent, and are definitely not going to be able to reach GPT7 (1000 times GPT4, or roughly 100 billion dollars). So, unless you believe compute power will go down in cost massively, or that we are just a factor 10-100 away from human-level intelligence, odds are good we won't see it from this line of AIs in the next 10-20 years.

tsimionescu 3y ago

First of all, there is no theory at all which explains the current progress of LLMs. It's not even a quantifiable progress - it's clearly there, but no one can say with any true precision whether GPT2 to GPT3 to GPT4 is following linear or exponential or whatever else precise function of progress in capabilities. We can clearly see that they are improving, but we have no idea how to accurately quantify this space. Secondly, we can see that there are obvious quantifiable physical increases in the requirements of training and inference for GPT4 compared to GPT3 and to GPT2. So, we can absolutely make a physical argument: at some point, you'll need too much energy to keep scaling up GPTs for it to be feasible, even if capability would keep improving with model size. Thirdly, the GPTs require ever more data to train, and there is a limited amount of text data in the world, of which they are already using a significant percent. So, there is also an argument that we might run out of data to improve even before we run out of physical resources. And making an argument that Moore's law will stop at some point due to physical constraints was just as correct in 1978 as it is in 2024. Predicting when that is is obviously hard - and the same is true for LLM scaling. We may be at the peek, or there may be 100 more generations before we even approach the limits I'm contemplating. But the limits are certainly there.

lucubratory 3y ago
This conversation started with someone saying that despite LLMs having a theoretical basis for how they could emulate human intelligence, they won't actually reach that point because the existing scaling laws won't hold. I am pointing out that we don't know that at all, scaling laws are so far fantastically accurate, and there is no sign of undershooting existing those laws so far. Therefore, it's very premature to declare that LLMs will be unable to reach any given milestone because scaling laws will hit a discontinuity before we get there. Basically: "There isn't even a theoretical basis for how LLMs could achieve human level intelligence" "There is, a neural network trained to predict the next token isn't restricted in the algorithms it can implement internally to achieve that goal, and eventually for loss to be low enough you need to successfully emulate human minds. We have reason to believe loss will continue to go lower because we have very accurate scaling laws which predict that & haven't been wrong so far." "But an LLM won't be able to reach that point because the scaling laws will break down." "Maybe, there's no evidence for that yet but we should scale them up if we want to find out for sure." You are here -> "But in the physical extreme LLMs will break down as an architecture." To which my response is: Yes, of course this specific architecture will break down in the physical extreme. By the time something is building computronium to run the next AI I strongly doubt it will be running matrix multiplication, let alone LLMs specifically. There is no guarantee or reason to believe that we will hit those limits before we are able to achieve human level intelligence or more; the only reason to believe it will happen is an empirically observed trend which could totally break down, so I wouldn't bet the house on it happening, but it's wrong to just rule out the outcome.
- tsimionescu 3y ago
  No, I am at the point where I'm saying the scaling "laws" will break down. Especially since there are no scaling laws, it's an empirical observation based on essentially 3 examples - GPT2 to GPT3 to GPT4 (and their rough equivalents from Facebook and Google and a handful of others). This all happened in the span of 4 years or less. We don't even fully know how much larger GPT4 is compared to GPT3. Not to mention, we already know OpenAI has spent more than a hundred millions of dollars training GPT-4, and they are quite probably selling it at a loss (each query may well cost them more in compute power than they charge for it). So, if GPT4 is ten times bigger than GPT3(.5) as sometimes reported, economics suggest they may already not be able to scale GPT5 to the same extent, and are definitely not going to be able to reach GPT7 (1000 times GPT4, or roughly 100 billion dollars). So, unless you believe compute power will go down in cost massively, or that we are just a factor 10-100 away from human-level intelligence, odds are good we won't see it from this line of AIs in the next 10-20 years.