9 ms·
> It seems to me bascially certain that no compressed representation of text can be an understanding of langugae, so necessarily, any statistical algorithm here
by whimsicalism 5y ago
> It seems to me bascially certain that no compressed representation of text can be an understanding of langugae, so necessarily, any statistical algorithm here is always using coincidental tricks. That it takes 500bn parameters to do it, i think, is a clue that we dont even really need.
I think your premise contains your conclusion, which while common, is something you should strive to avoid.
I do think your opinion is a good example of the prevailing sentiment on Hacker News. To me, it seems to come from a discomfort with the fact that even "we" emerge out of the basic interactions of basic building blocks. Our brain has been able to build world knowledge "merely by" analysis of electrical impulses being transmitted to it on wires.
- mjburgess 5y agoI have no discomfort with the notion that our bodies, which grow in response to direct causal contact with our environment, contain in-their-structure the generative capbaility for knoweldge, imagination, skill, growth -- and so on. I have no discomfort with the basically schiozphrenic notion that the shapes of words have something to do with the nature of the world. I just think its a kind of insantity which absolutely destroys our ability to reason carefully about the use of these systems. That "tr" occurs before "ee" says as much about "trees" as "leaves are green" says -- it is only that *we* have the relevant semantics that the latter is meaningful when interpreted in the light of our "environmental history" recorded in our bodies, and given weight and utility by our imaginations. The structure of text is not the structure of the world. This thesis is mad. Its a scientific thesis. It is trivial to test it. It is trivial to wholey discred it. It's pseudoscience. No one here is a scientist and no one treats any of this as science. Where's the criteria for the emprical adequecy of NLP systems as models of language? Specifying any, conducting actual hypothesis tests, and establishing a theory of how NLP systems model language -- this would immediately reveal the smoke-and-mirros. The work to reveal the statistical tricks underneath them takes years, and no one has much motivation to do it. The money lies in this sales pitch, and this is no science. This is no scientific method.
- rafaelero 5y agoOk, boomer.
- whimsicalism 5y agoAgree to disagree. I think you are opining about things that you are lacking fundamental knowledge on. > The structure of text is not the structure of the world. This thesis is mad. Its a scientific thesis. It is trivial to test it. It is trivial to wholey discred it. It's pseudoscience. It's unclear what you even mean by that. Are the electrical impulses coming to our brain the "structure of the world"?
- yldedly 5y agoThe structure of having X apples in Y buckets is the same as the structure in the expression "X * Y", as long as the expression exists in a context that can parse it using the rules of arithmetic, such as a human, or a calculator. These language models lack context, not just for arithmetic, but for everything. They can't parse "X * Y" for any X and Y, they've just associated the expression with the right answer for so many values of X and Y, that we get fooled into thinking they know the rules. We get fooled into thinking they've learned the structure of the world. But they've only learned the structure of text.
- whimsicalism 5y agoIt would be trivial for a network of this size to code general rules for multiplication. At a certain point, when you have enough data, finding the actual rule is actually the easier solution than memorizing each data point. This is the key insight of deep learning.
- yldedly 5y agoReally? Better inform all the researchers working on this that they're wasting their time then: https://arxiv.org/abs/2001.05016 https://arxiv.org/abs/2001.05016 More fundamentally, any finite neural net is either constant or linear outside the training sample,depending on the activation function. Unless you design special neurons like in the paper above, which solves this specific problem for arithmetic, but not the general problem of extrapolation.