3 ms·

>> The specific techniques we used (a variant of the transformer architecture behind most large language models) means, among other things, that you can ask the

>> The specific techniques we used (a variant of the transformer architecture behind most large language models) means, among other things, that you can ask the model questions, and get answers about what the model "knows" in return. You know, for me it is clear that the ability to retrieve information is not the same as the ability to understand the retrieved information. For example, you could store a corpus of scientific texts in a database and then retrieve pieces of it with a query, but that doesn't say the database understands the corpus. And this is relevant to your comment because with transformers we still have essentially databases of information that can only be retrieved by the right query, where the information is stored as weights in the trained model and the query is a suitable prompt. In fact, prompts to transformers are so finicky and must be tailored so carefully to retrieve the desired piece of information that they are not very different than a SQL query, albeit one in natural language. I'm offering this as a counterpoint to your assertion that "the model has learned drug biochemistry at some significant level". What I understand from your comment is that your model has learned to represent a body of text, nothing more, nothing less.

walnutclosefarm 4y ago
I'm not claiming that any of these models have achieved a generalized level of scientific understanding. They are narrow in their abilities - savants of a sort. But it seems to me to be missing an important point to reduce them to mere databases. Learned knowledge and intelligent behavior are multi-dimensional, and I think we're seeing a different sampling with the current generation of AI models than we see in our own sampling. They are highly specific, and refined for one "intelligent" task, but not in any way broadly general. We are comparing them to human intelligence, which leans heavily in the other direction.
- YeGoblynQueenne 4y ago
  I don't see it that way. What I see is systems that have behaviours that look like human behaviours - for example, generating and recognising text, or recognising images, etc; but what these systems really do is nothing like whatever humans do, and the behaviours they produce are not the same behaviours as those of humans, after all. For example, we can see that clearly in chess and other board games. Chess engines play by executing some variant of the minimax algorithm and they can out-play humans. But human players don't typcially try to run minimax in their heads when they play and in any case can play chess without trying to emulate a chess engine. The same goes for image classifiers (it is extremely unlikely that we match sets of pixels to numerical labels in our heads, to the point of absurdity) and for language models (again it's absurdly unlikely that we calculate the probabilities of sequences of tokens). I think what you say comes from an assumption that because humans exhibit certain behaviours, and humans have human intelligence, any system that exhibits similar behaviours must also have human intelligence. Which, if you look carefully, does not follow. Anyway historically people have always been quick to identify the capabilities of computers as artificial intelligence, just because we tend to think of anything to do with computation as requiring intelligence. But we had machines that could perform calculations automatically, through entirely mechanical procedures, for more than 200 years now, and most of those we don't think are artificially intelligent. For example, I don't think anyone would consider the Pascaline to be an AI, and yet the Pascaline can do better arithmetic from any non-human animal, and from many human animals, also. So, does arithmetic require intelligence, or not? How much of what we can do thanks to our intelligence can also be done by a dumb, mechanical automatism? This I think is the important question in AI research today and one that, if we could answer it (which we can't) would finally put the lid on wishful thinking about LLMs exhibiting "a different sampling" of intelligence than ours.

walnutclosefarm 4y ago

I'm not claiming that any of these models have achieved a generalized level of scientific understanding. They are narrow in their abilities - savants of a sort. But it seems to me to be missing an important point to reduce them to mere databases. Learned knowledge and intelligent behavior are multi-dimensional, and I think we're seeing a different sampling with the current generation of AI models than we see in our own sampling. They are highly specific, and refined for one "intelligent" task, but not in any way broadly general. We are comparing them to human intelligence, which leans heavily in the other direction.

YeGoblynQueenne 4y ago
I don't see it that way. What I see is systems that have behaviours that look like human behaviours - for example, generating and recognising text, or recognising images, etc; but what these systems really do is nothing like whatever humans do, and the behaviours they produce are not the same behaviours as those of humans, after all. For example, we can see that clearly in chess and other board games. Chess engines play by executing some variant of the minimax algorithm and they can out-play humans. But human players don't typcially try to run minimax in their heads when they play and in any case can play chess without trying to emulate a chess engine. The same goes for image classifiers (it is extremely unlikely that we match sets of pixels to numerical labels in our heads, to the point of absurdity) and for language models (again it's absurdly unlikely that we calculate the probabilities of sequences of tokens). I think what you say comes from an assumption that because humans exhibit certain behaviours, and humans have human intelligence, any system that exhibits similar behaviours must also have human intelligence. Which, if you look carefully, does not follow. Anyway historically people have always been quick to identify the capabilities of computers as artificial intelligence, just because we tend to think of anything to do with computation as requiring intelligence. But we had machines that could perform calculations automatically, through entirely mechanical procedures, for more than 200 years now, and most of those we don't think are artificially intelligent. For example, I don't think anyone would consider the Pascaline to be an AI, and yet the Pascaline can do better arithmetic from any non-human animal, and from many human animals, also. So, does arithmetic require intelligence, or not? How much of what we can do thanks to our intelligence can also be done by a dumb, mechanical automatism? This I think is the important question in AI research today and one that, if we could answer it (which we can't) would finally put the lid on wishful thinking about LLMs exhibiting "a different sampling" of intelligence than ours.