MikeB wrote: ↑Fri Dec 30, 2022 1:00 amAt the end of the day, whether legal or illegal, it did appear to be attempt by Chessbase to make money off of the backs who had volunteered their time to the Stockfish project.
But this is allowed and fully in the spirit of the GPL, provided that any product covered by the SF copyright is released under the terms of the GPL.
No single contributor to SF should feel sour about this because SF would not have been what it is without the GPL. (In fact, SF might never have existed if Glaurung had not been released und the GPL, although it is conceivable that Marco would have picked another starting point.)
edit: just noticed I am reproducing what others have already written.
Because of the functional nature of the mask geometry, the designs cannot be effectively protected under copyright law (except perhaps as decorative art).
I already wrote above that a particular graphical representation of a neural network (with creative choice for colors and layout etc.) could be protected. But the set of weights does not contain such choices.
A mask geometry is clearly a computer program. The 6502 can now be emulated directly from a photo of the chip.
Perhaps legislators will come up with a special kind of intellectual property protection for NNs, but today this does not exist.
+1
I recall that MLP neural networks solved the XOR problem with SLP NNs, from this point of view the NN must be pure functional....
A text stored in an electronic medium is just a collection of numbers. Or a JPEG encoded movie.
This counterargument goes absolutely nowhere (and I must have refuted it about 10 times by now). The numbers are not protected, but the expression they encode is. The expression encoded in electronic text or in a JPEG is not a mere idea or functionality.
[...]
And here is my question, what does a neural network for chess really encode? What do these numbers/weights represent in the real world? Is it just a function, is it chess patterns, is it an approximation of perfect chess knowledge, is it the creative process of the NN creator, what does supervised learning encode, what does reinforcement learning encode? Twaddle, twaddle
As you mentioned, you can encode a function with NNs, but you can also encode a .jpeg picture with NNs, it is both, memory and computation.
Because of the functional nature of the mask geometry, the designs cannot be effectively protected under copyright law (except perhaps as decorative art).
I already wrote above that a particular graphical representation of a neural network (with creative choice for colors and layout etc.) could be protected. But the set of weights does not contain such choices.
A mask geometry is clearly a computer program. The 6502 can now be emulated directly from a photo of the chip.
So... if you build a new chip, I can legally clone it?
This doesn't make sense: there must be some case law about this.
Let's take cars: the outside of a car is designed to look pretty, and clearly will be protected. But if Ford make a great new engine, then why don't Toyota just directly copy it in every detail? Why has that never happened in countries that respect property and copyright laws?
Human chess is partly about tactics and strategy, but mostly about memory
Because of the functional nature of the mask geometry, the designs cannot be effectively protected under copyright law (except perhaps as decorative art).
I already wrote above that a particular graphical representation of a neural network (with creative choice for colors and layout etc.) could be protected. But the set of weights does not contain such choices.
A mask geometry is clearly a computer program. The 6502 can now be emulated directly from a photo of the chip.
So... if you build a new chip, I can legally clone it?
This doesn't make sense: there must be some case law about this.
Let's take cars: the outside of a car is designed to look pretty, and clearly will be protected. But if Ford make a great new engine, then why don't Toyota just directly copy it in every detail? Why has that never happened in countries that respect property and copyright laws?
smatovic wrote: ↑Sun Jan 01, 2023 10:45 amAnd here is my question, what does a neural network for chess really encode? What do these numbers/weights represent in the real world? Is it just a function, is it chess patterns, is it an approximation of perfect chess knowledge, is it the creative process of the NN creator, what does supervised learning encode, what does reinforcement learning encode? Twaddle, twaddle
A mathematical function. Perhaps "knowledge" or "ideas". Not something expressive that we can perceive.
Moreover, whatever it encodes does not reflect any human author's personality, it is not the author's own intellectual creation, there is no personal stamp. It is the result of backpropagation calculations.
As you mentioned, you can encode a function with NNs, but you can also encode a .jpeg picture with NNs, it is both, memory and computation.
If an NN is somehow used to encode an image (such that the image can be reproduced), it will be covered by the copyright on the image.
Because of the functional nature of the mask geometry, the designs cannot be effectively protected under copyright law (except perhaps as decorative art).
I already wrote above that a particular graphical representation of a neural network (with creative choice for colors and layout etc.) could be protected. But the set of weights does not contain such choices.
A mask geometry is clearly a computer program. The 6502 can now be emulated directly from a photo of the chip.
So... if you build a new chip, I can legally clone it?
This doesn't make sense: there must be some case law about this.
Let's take cars: the outside of a car is designed to look pretty, and clearly will be protected. But if Ford make a great new engine, then why don't Toyota just directly copy it in every detail? Why has that never happened in countries that respect property and copyright laws?
I think the line draws similar to a dictionary, like you can't copyright individual words, their meanings or explanations, however you can copyright the specific instance of their collections, which involves a fair amount of human involvement.
So for NNs, you can't copyright numbers, but if you have a specific collection of numbers that also involved a fair amount of human involvement, then it should be copyrightable. Although they can be lame ones since those NNs(collections) can be independently and trivially produced by anyone, they are unlikely to end up having exactly the same numbers. The ones that resulted in exactly the same numbers will render any copyrights meaningless anyways.
It also does not contradict the steam engine example, the idea can't be copyrighted, the specific designs can be covered by patents, the specific engine(instance) can be copyrighted.
Consider this, I can write a hello world program, which then I would have my copyright over it, a lame one of course(but I do not agree with people who would argue that what I wrote is uncopyrightable), then I write hello world programs for 100 different programming languages, and my copyright over the collection of 100 hello world programs becomes somewhat less lamer, it also doesn't stop anyone else from doing the same and even for that I'm not sure everyone who attempted it would result in exactly the same code as I wrote. So why is there a problem?
Last edited by noobpwnftw on Sun Jan 01, 2023 5:38 pm, edited 1 time in total.
syzygy wrote: ↑Sun Jan 01, 2023 5:15 amIf you manage to use NNs as a functional alternative to JPEG, then sure, an NN encoding a specific image will be covered by the copyright on the image. If you manage to use an NN to encode the source code of a computer program, then same.
Well, that confirms the only point I wanted to make: the reasoning"these are weights of an NN, so they cannot be protected by copyright" is not sound. It depends on what the NN does, and how it was taught or designed to do that.
Now in the case of Chess we have of course to realize that there is a court ruling that declares the moves of a chess game cannot be copyrighted. This makes it less likely that even the data fed to the NN for training could carry any copyrights that the NN might inherit (as a derived work).
From what I'm reading here, if a company sells machines with no patents in a pretty container, I cannot copy the container because that has "artistic human expression", but I can copy the machine EXACTLY.
If that was the law, then I would have expected it to have happened many times. I know there are allegations of developing countries doing this in the pre-internet era (I've actually seen this done to me: I wrote an article for a magazine, and years later saw my article, with very few changes, in a Brazilian magazine - credited to somebody I'd never heard of!), but that's not what I'm taking about: here's what I'm talking about:
* a factory near to where I live produces a useful machine in a pretty container, and it has no patents
* I commission a different factory to make the EXACT SAME MACHINE (no differences at all) and package it in a different container
If there's no legal reason for me to not do that, why doesn't it happen all the time?
Human chess is partly about tactics and strategy, but mostly about memory
towforce wrote: ↑Sun Jan 01, 2023 5:55 pm
This issue simply cannot be unique to NNs.
From what I'm reading here, if a company sells machines with no patents in a pretty container, I cannot copy the container because that has "artistic human expression", but I can copy the machine EXACTLY.
If that was the law, then I would have expected it to have happened many times. I know there are allegations of developing countries doing this in the pre-internet era (I've actually seen this done to me: I wrote an article for a magazine, and years later saw my article, with very few changes, in a Brazilian magazine - credited to somebody I'd never heard of!), but that's not what I'm taking about: here's what I'm talking about:
* a factory near to where I live produces a useful machine in a pretty container, and it has no patents
* I commission a different factory to make the EXACT SAME MACHINE (no differences at all) and package it in a different container
If there's no legal reason for me to not do that, why doesn't it happen all the time?
I guess cos nobody has interest in an steam machine or an decades old CPU design, patents are usually 20 years, mask work 10 to 15, Copyright somewhere between 50 to 100, you know probable the .mp3 and .jpeg patents (technical solution), Frauenhofer Institute collected millions in fees with .mp3, nowadays everybody is free to implement encoder/decoder without fees, but meanwhile new codecs offer new features, project https://www.gutenberg.org/ has thousands of free books online, and there were legal issues with this cos countries differ when Copyright does expire...the Copyright on the original Mickey Mouse character expired recently, but Disney changes the design over the years, even if you are now free to use the original Mickey, nowadays kids won't recognize him...
