* Re: Help building Pen.el (GPT for emacs) [not found] ` <CACT87JriMaF1kFjEE_8=8FEQpAi6sxr3x3vZT3rafjY=4mQgZg@mail.gmail.com> @ 2021-07-19 17:00 ` Jean Louis 2021-07-23 6:51 ` Shane Mulligan 0 siblings, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-19 17:00 UTC (permalink / raw) To: Shane Mulligan; +Cc: Eli Zaretskii, emacs-tangents, Stefan Kangas, rms * Shane Mulligan <mullikine@gmail.com> [2021-07-18 11:01]: > Pen.el stands for Prompt Engineering in emacs. > Prompt Engineering is the art of describing what you would > like a language model (transformer) to do. It is a new type of programming, > example oriented; > like literate programming, but manifested automatically. Sounds like a replacement for a programmer's mind. > A transformer takes some text (called a prompt) and continues > it. However, the continuation is the superset of all NLP tasks, Where is definition of the abbreviation NLP? > as the generation can also be a classification, for instance. Those > NLP tasks extend beyond world languages and into programming > languages (whatever has been 'indexed' or 'learned') from these > large LMs. What is definition of the abbreviation LM? > Pen.el is an editing environment for designing 'prompts' to LMs. It > is better than anything that exists, even at OpenAI or at > Microsoft. I have been working on it and preparing for this for a > long time. Good, but is there a video to show what it really does? > These prompts are example- based tasks. There are a number of design > patterns which Pen.el is seeking to encode into a domain-specific > language called 'examplary' for example- oriented programming. Do you mean "exemplary" or "examplary", is it spelling mistake? I have to ask as your description is still pretty abstract without particular example. > Pen.el creates functions 1:1 for a prompt to an emacs lisp function. The above does not tell me anything. > Emacs is Grammarly, Google Translate, Copilot, Stackoveflow and > infinitely many other services all rolled into one and allows you to > have a private parallel to all these services that is completely > private and open source -- that is if you have downloaded the > EleutherAI model locally. I understand that it is kind of fetching information, but that does not solve licensing issues, it sounds like licensing hell. > ** Response to Jean Louis > - And I do not think it should be in GNU ELPA due to above reasons. > > I am glad I have forewarned you guys. This is my current goal. Help > in my project would be appreciated. I cannot do it alone and I > cannot convince all of you. Why don't you tell about licensing issues? Taking code without proper licensing compliance is IMHO, not an option. It sounds as problem generator. > > Why don't you simply make an Emacs package as .tar as described in Emacs > Lisp manual? > Thank you for taking a look at my emacs package. It's not ready net > for Melpa merge. I hope that I will be able to find some help in > order to prepare it, but the rules are very strict and this may not > happen. I did not say to put it in Melpa. Package you can make for yourself and users so that users can M-x package-install-file That is really not related to any online Emacs package repository. It is way how to install Emacs packages no matter where one gets it. > > How does that solves the licensing problems? > The current EleutherAI model which competes with GPT-3 is GPT-Neo. > It is MIT licensed. That is good. But the code that is generated and injected requires proper contribution. > Also the data it has been trained on is MIT licensed. Yes, and then the program should also solve the proper contributions automatically. You cannot just say "MIT licensed", this has to be proven, source has to be found and proper attributions applied. Why don't you implement proper licensing? Please find ONE license that you are using from code that is being used as database for generation of future code and provide link to it. Then show how is license complied to. > The current EleutherAI model which competes with Codex is GPT-j. > It is licensed with Apache-2.0 License That is good, but I am referring to the generated code. -- Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-19 17:00 ` Help building Pen.el (GPT for emacs) Jean Louis @ 2021-07-23 6:51 ` Shane Mulligan 2021-07-23 10:12 ` Jean Louis 2021-07-25 1:06 ` Richard Stallman 0 siblings, 2 replies; 46+ messages in thread From: Shane Mulligan @ 2021-07-23 6:51 UTC (permalink / raw) To: Jean Louis; +Cc: Eli Zaretskii, emacs-tangents, Stefan Kangas, rms [-- Attachment #1: Type: text/plain, Size: 10102 bytes --] Hi Jean and GNU friends, GPT is potentially the best thing to happen to emacs in a very long time. It will bring back power from the corporations and save it to your computer, open source and transparent, and offline. Please consider including a collaborative, open source prompts repository in the next version of emacs. So far I'm yet to see anything like it, but I see in commercial products everywhere that they have full domain over this new type of code. I am trying to build up relationships in my project Pen.el with others who value open source. gptprompts.org, for example. This is to create a catalogue for pen.el. One thing we have just introduced is a field to specify a licence for each prompt. However, I must say that prompts are more like functions. *Soft prompts* are very granular prompts as they have been reduced to a minimal number of characters using optimisation. https://arxiv.org/abs/2104.08691 Therefore, there must be support for prompting at the syntax level of emacs, in my opinion. And it is also clear now that since a prompt looks more like binary code, that this is a new type of function definition and a new type of programming is emerging. A prompt function is a function defined by a version of a Language Model (LM) and a prompt (input), but as is the case in haskell, every function may be reduced to one that takes a single input and returns a single output. In other words, most prompt functions will be parameterised and have an arity greater than one. I am building a collaborative imaginary programming environment in emacs. This is an editing environment where people can integrate LMs into emacs, extending emacs with prompt functions. The power of this is profound and beyond belief. I have coined the term "prompt functions", so don't expect to be able to find it online if you go searching. Here is a new corporation which is creating a prompt engineering environment. However, they do not have their own operating system to integrate prompting into. That's why emacs is years ahead, potentially. A prompt is merely a function with a language model as a parameter. Without integration, it's quite useless. https://gpt3demo.com/apps/mantiumai I think a prompts database -- something like Datomic or other RDF-like, immutable storage must be added into GNU organisation to store selected prompts and generations, and a GPL or EleutherAI GPT model is ultimately integrated into core emacs via some low level syntax through partnership with EleutherAI. I would expect in the future to download emacs along with an open- source GPT model, and be able to create prompt functions as easily as creating macros. A 1:1 prompt:function database of sorts is a good starting point in my opinion, but remembering the generations is also important. But the scale is immense. This is why a p2p database that can remember immutably is important, in my opinion. If this seems too grand of scale, then at the very least consider a GNU prompts repository. > Sounds like a replacement for a programmer's mind. Yes it is. It trivialises the implementation and requires that programmers now be more imaginative, and will be supported by the language model. Rather than writing an implementation, function is defined by the input types and a Language Model and version of the language model. > Where is definition of the abbreviation NLP? NLP stands for Natural Language Processing. Until recently, code was not considered part of that domain, but the truth is NLP algorithms are extremely useful for code generation, code search and code understanding. > What is definition of the abbreviation LM? LM stands for Language Model. It is a statistical model of language, rather than use formal grammars. Emacs lisp functions and macros do not have a syntax for stochastic/probabilistic programming. Good, but is there a video to show what it really does? Here is an online catalogue of GPT tools. Pen.el is among the developer tools. https://gpt3demo.com/category/developer-tools =Pen.el= and emacs has the potential to do all the things for all of the products in =gpt3demo.com=. > I would like to demonstrate Pen.el with this particular video which I have created to demonstrate a new type of programming -- collaborative within a language model. https://mullikine.github.io/posts/caching-and-saving-results-of-prompt-functions-in-pen-el/ https://asciinema.org/a/MhOU0eMnJsRpXf2Ak9YStPlz8 > Do you mean "exemplary" or "examplary", is it spelling mistake? I am building a DSL for encoding prompt design patterns to generate prompt functions for emacs. http://github.com/semiosis/pen.el/blob/master/src/pen-examplary.el > Pen.el creates functions 1:1 for a prompt to an emacs lisp function. What this means is that a prompt may be parameterized to define a relation (i.e. function) and therefore code and I have chosen to create one parameterized function per prompt. The prompt text once associated to a LM becomes a type of query (i.e. code), so prompts should not be discounted as being any less than such, and qualify for the GPL3 license. > I understand that it is kind of fetching information, but that does not solve licensing issues, it sounds like licensing hell. This is exactly why a GPL LM or compatible LM is absolutely crucial and needs to be integrated, otherwise all imaginary code will be violating and harvesting open source for the foreseeable future as there is no alternative. Sincerely, Shane . Shane Mulligan How to contact me: 🇦🇺 00 61 421 641 250 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> mullikine@gmail.com On Tue, Jul 20, 2021 at 5:04 AM Jean Louis <bugs@gnu.support> wrote: > * Shane Mulligan <mullikine@gmail.com> [2021-07-18 11:01]: > > Pen.el stands for Prompt Engineering in emacs. > > Prompt Engineering is the art of describing what you would > > like a language model (transformer) to do. It is a new type of > programming, > > example oriented; > > like literate programming, but manifested automatically. > > Sounds like a replacement for a programmer's mind. > > > A transformer takes some text (called a prompt) and continues > > it. However, the continuation is the superset of all NLP tasks, > > Where is definition of the abbreviation NLP? > > > as the generation can also be a classification, for instance. Those > > NLP tasks extend beyond world languages and into programming > > languages (whatever has been 'indexed' or 'learned') from these > > large LMs. > > What is definition of the abbreviation LM? > > > Pen.el is an editing environment for designing 'prompts' to LMs. It > > is better than anything that exists, even at OpenAI or at > > Microsoft. I have been working on it and preparing for this for a > > long time. > > Good, but is there a video to show what it really does? > > > These prompts are example- based tasks. There are a number of design > > patterns which Pen.el is seeking to encode into a domain-specific > > language called 'examplary' for example- oriented programming. > > Do you mean "exemplary" or "examplary", is it spelling mistake? > > I have to ask as your description is still pretty abstract without > particular example. > > > Pen.el creates functions 1:1 for a prompt to an emacs lisp function. > > The above does not tell me anything. > > > Emacs is Grammarly, Google Translate, Copilot, Stackoveflow and > > infinitely many other services all rolled into one and allows you to > > have a private parallel to all these services that is completely > > private and open source -- that is if you have downloaded the > > EleutherAI model locally. > > I understand that it is kind of fetching information, but that does > not solve licensing issues, it sounds like licensing hell. > > > ** Response to Jean Louis > > - And I do not think it should be in GNU ELPA due to above reasons. > > > > I am glad I have forewarned you guys. This is my current goal. Help > > in my project would be appreciated. I cannot do it alone and I > > cannot convince all of you. > > Why don't you tell about licensing issues? Taking code without proper > licensing compliance is IMHO, not an option. It sounds as problem > generator. > > > > Why don't you simply make an Emacs package as .tar as described in > Emacs > > Lisp manual? > > > Thank you for taking a look at my emacs package. It's not ready net > > for Melpa merge. I hope that I will be able to find some help in > > order to prepare it, but the rules are very strict and this may not > > happen. > > I did not say to put it in Melpa. Package you can make for yourself > and users so that users can M-x package-install-file > > That is really not related to any online Emacs package repository. It > is way how to install Emacs packages no matter where one gets it. > > > > How does that solves the licensing problems? > > The current EleutherAI model which competes with GPT-3 is GPT-Neo. > > It is MIT licensed. > > That is good. > > But the code that is generated and injected requires proper > contribution. > > > Also the data it has been trained on is MIT licensed. > > Yes, and then the program should also solve the proper contributions > automatically. You cannot just say "MIT licensed", this has to be > proven, source has to be found and proper attributions applied. > > Why don't you implement proper licensing? > > Please find ONE license that you are using from code that is being > used as database for generation of future code and provide link to > it. Then show how is license complied to. > > > The current EleutherAI model which competes with Codex is GPT-j. > > It is licensed with Apache-2.0 License > > That is good, but I am referring to the generated code. > > -- > Jean > > Take action in Free Software Foundation campaigns: > https://www.fsf.org/campaigns > > In support of Richard M. Stallman > https://stallmansupport.org/ > [-- Attachment #2: Type: text/html, Size: 13458 bytes --] ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 6:51 ` Shane Mulligan @ 2021-07-23 10:12 ` Jean Louis 2021-07-23 10:54 ` Eli Zaretskii 2021-07-25 1:06 ` Richard Stallman 1 sibling, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-23 10:12 UTC (permalink / raw) To: Shane Mulligan; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, rms * Shane Mulligan <mullikine@gmail.com> [2021-07-23 10:07]: > Hi Jean and GNU friends, > > GPT is potentially the best thing to happen to emacs in a very long time. > It will bring back power from the corporations and save it to your > computer, open source and transparent, and offline. Description is too apstract. Show me how it brought power from corporations and saved it to your computer? The term "open source" is vague, you know it? So I do not know what is meant with it. There is no clear licensing that you are proposing, neither so far you explained how is licensing solved. If you have not read the licenses please let me know, as then you most probably do not know to what I am referring. Example: You are using artificial intelligence, but in fact pieces of codes are in chunks copied from other sources without attribution and without knowing if licenses are compatible. If that issue is not solve I do not see why would anybody serious use AI to create code as that would potentially generate so many legal problems. And it does so now. So many people gave up on Github because Github does not comply to licenses when using Copilot. You are more or less proposing the same conflict to come to Emacs and I did not see where is your solution? > I understand that it is kind of fetching information, but that does > not solve licensing issues, it sounds like licensing hell. This is > exactly why a GPL LM or compatible LM is absolutely crucial and > needs to be integrated, otherwise all imaginary code will be > violating and harvesting open source for the foreseeable future as > there is no alternative. So how? That should be first to start with as one cannot even experiment in public without it. Experimenting at home is fine, but as soon as anything is published in public without compliance to licenses it generates problems. -- Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 10:12 ` Jean Louis @ 2021-07-23 10:54 ` Eli Zaretskii 2021-07-23 11:32 ` Jean Louis 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-23 10:54 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, emacs-tangents, mullikine, rms > Date: Fri, 23 Jul 2021 13:12:12 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: Eli Zaretskii <eliz@gnu.org>, Stefan Kangas <stefan@marxist.se>, > emacs-tangents@gnu.org, rms@gnu.org > > You are using artificial intelligence, but in fact pieces of codes are > in chunks copied from other sources without attribution and without > knowing if licenses are compatible. That's not what happens with these services: they don't _copy_ code from other software (that won't work, because the probability of the variables being called by other names is 100%, and thus such code, if pasted into your program, will not compile). What they do, they extract ideas and algorithms from those other places, and express them in terms of your variables and your data types. So licenses are not relevant here. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 10:54 ` Eli Zaretskii @ 2021-07-23 11:32 ` Jean Louis 2021-07-23 11:51 ` Eli Zaretskii 2021-07-24 1:14 ` Richard Stallman 0 siblings, 2 replies; 46+ messages in thread From: Jean Louis @ 2021-07-23 11:32 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-23 13:55]: > > Date: Fri, 23 Jul 2021 13:12:12 +0300 > > From: Jean Louis <bugs@gnu.support> > > Cc: Eli Zaretskii <eliz@gnu.org>, Stefan Kangas <stefan@marxist.se>, > > emacs-tangents@gnu.org, rms@gnu.org > > > > You are using artificial intelligence, but in fact pieces of codes are > > in chunks copied from other sources without attribution and without > > knowing if licenses are compatible. > > That's not what happens with these services: they don't _copy_ code > from other software (that won't work, because the probability of the > variables being called by other names is 100%, and thus such code, if > pasted into your program, will not compile). What they do, they > extract ideas and algorithms from those other places, and express them > in terms of your variables and your data types. So licenses are not > relevant here. According to online reviews chunks of code is copied even verbatim and people find from where. Even if modified, it still requires licensing compliance. If code compiles or not is irrelevant. If one runs it or not is also irrelevant, code need not even run. I do not believe that any of the AI-s so far "extract ideas". I never heard of it. Which algorithms is there on this planet that may extract idea? I also do not believe that AI extract algorithms, though AI has its own algorithms to generate the relevant possible code. If newly generated code is modification from other code, what we know now that it is, and is based on, that requires licensing attributions. That licenses are relevant one can see from online discussions related to Github Copilot: https://fossa.com/blog/analyzing-legal-implications-github-copilot/ https://medium.com/geekculture/githubs-ai-copilot-might-get-you-sued-if-you-use-it-c1cade1ea229 https://fosspost.org/github-copilot/ https://artificialintelligence-news.com/2021/07/06/experts-debate-github-latest-ai-tool-violates-copyright-law/ https://www.opensourceforu.com/2021/07/github-copilot-sparks-debates-around-open-source-licenses/ That is just a small excerpt from a very large debat online related to licensing issues. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 11:32 ` Jean Louis @ 2021-07-23 11:51 ` Eli Zaretskii 2021-07-23 12:47 ` Jean Louis 2021-07-24 1:14 ` Richard Stallman 1 sibling, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-23 11:51 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, emacs-tangents, mullikine, rms > Date: Fri, 23 Jul 2021 14:32:00 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: mullikine@gmail.com, stefan@marxist.se, emacs-tangents@gnu.org, > rms@gnu.org > > > That's not what happens with these services: they don't _copy_ code > > from other software (that won't work, because the probability of the > > variables being called by other names is 100%, and thus such code, if > > pasted into your program, will not compile). What they do, they > > extract ideas and algorithms from those other places, and express them > > in terms of your variables and your data types. So licenses are not > > relevant here. > > According to online reviews chunks of code is copied even verbatim and > people find from where. That cannot be true. It is nonsense to copy unrelated code into a program and tell people this is what they should use. > If code compiles or not is irrelevant. If one runs it or not is also > irrelevant, code need not even run. A feature or service that is based on this idea will never fly, believe me. Which program would want to have code pasted into his/her program that would cause compilation errors or, worse, break it at run time? > I do not believe that any of the AI-s so far "extract ideas". I never > heard of it. Which algorithms is there on this planet that may extract > idea? That's a very general question, it is impossible to answer it in a post to a mailing list. If you are really interested, you will have to read up on that. But you are wrong in your beliefs. > If newly generated code is modification from other code, what we know > now that it is, and is based on, that requires licensing > attributions. Once again, your assumptions are all wrong, so your conclusions are also wrong. Why not try one of these services and see what they actually do, before you pass your (quite harsh) judgment on them, and on the modern state of AI in general? > That licenses are relevant one can see from online discussions related > to Github Copilot: That people ask these questions and discuss this doesn't mean the problem is real. many people don't really understand what copyright means and how to apply it to program code. People also ask questions about the GPL, and there's a vociferous group of people who think the copyright assignment of code to the FSF means you give up all your rights in the code you've written. None of that is true, but still the rumors and the heated discussions go on and on. Their existence proves nothing, except that some people misunderstand something. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 11:51 ` Eli Zaretskii @ 2021-07-23 12:47 ` Jean Louis 2021-07-23 13:39 ` Shane Mulligan 2021-07-23 19:33 ` Eli Zaretskii 0 siblings, 2 replies; 46+ messages in thread From: Jean Louis @ 2021-07-23 12:47 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-23 14:51]: > > According to online reviews chunks of code is copied even verbatim and > > people find from where. > > That cannot be true. It is nonsense to copy unrelated code into a > program and tell people this is what they should use. I wonder how sure you are in that, did you do the online research? It is not about related or unrelated, I do believe that AI finds and generates related code. But Here are references disputing how "it cannot be true": https://hacker-news.news/post/27710287 https://mmacvicar.medium.com/it-is-best-if-copilot-copies-everything-d84506128e5a https://loudlabs.nl/news/githubs-commercial-ai-tool-was-built-from-open-source-code/ > > If code compiles or not is irrelevant. If one runs it or not is also > > irrelevant, code need not even run. > > A feature or service that is based on this idea will never fly, > believe me. Which program would want to have code pasted into his/her > program that would cause compilation errors or, worse, break it at run > time? Of course people want code to fun. Just that copyright laws don't handle technical functionality. It is irrelevant if program works or does not work. There are thousands of copyrighted programs that cannot work any more as devices are not on the market, they are still under copyright. > > I do not believe that any of the AI-s so far "extract ideas". I never > > heard of it. Which algorithms is there on this planet that may extract > > idea? > > That's a very general question, it is impossible to answer it in a > post to a mailing list. If you are really interested, you will have > to read up on that. But you are wrong in your beliefs. > > > If newly generated code is modification from other code, what we know > > now that it is, and is based on, that requires licensing > > attributions. > > Once again, your assumptions are all wrong, so your conclusions are > also wrong. Why not try one of these services and see what they > actually do, before you pass your (quite harsh) judgment on them, and > on the modern state of AI in general? I can hear you how I am wrong, conclusions are wrong, though I gave you references enough to research it on Internet that will tell that there are possible serious licensing problems with such generated code. > > That licenses are relevant one can see from online discussions related > > to Github Copilot: > > That people ask these questions and discuss this doesn't mean the > problem is real. many people don't really understand what copyright > means and how to apply it to program code. Well said! Though that is not relevant. Question is very particular, specific and concrete: ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ How does Pen.el and background AI services ensure of licensing compliance? I would appreciate if you find solution to that or stay on that subject, as if I am wrong or right is not relevant, what I wish is to have assurance that it is free software. Prove me wrong by providing exact references in not only on country's law but also other countries' laws, the lows that make it legal, or how otherwise the legality of such code is justified and how users may get free software. For example you may wish to mention "fair use" and on the other hand similar laws must be found in other countries that would justify it to be free software. As long as you don't tackle those subjects there is no legal solution for Pen.el and background AI to be used with assurance that software is truly free software. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 12:47 ` Jean Louis @ 2021-07-23 13:39 ` Shane Mulligan 2021-07-23 14:39 ` Jean Louis 2021-07-26 0:16 ` Richard Stallman 2021-07-23 19:33 ` Eli Zaretskii 1 sibling, 2 replies; 46+ messages in thread From: Shane Mulligan @ 2021-07-23 13:39 UTC (permalink / raw) To: Jean Louis; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, rms [-- Attachment #1: Type: text/plain, Size: 9744 bytes --] Hi Jean, Eli, GNU, > "open source" I am referring to free software in the spirit of GNU. Free as in freedom, from oppression, from an attack against creative and cognitive intelligence. > GPT is potentially the best thing to happen to emacs in a very long time. > It will bring back power from the corporations and save it to your > computer, open source and transparent, and offline. The way this will work is you will download the free GPT model, such as GPT-j, GPT-neo or GPT-neox and then you will have an offline and private alternative to many things previously you would go online for. I have been working 5 months on demonstrations this whole time and I have informed you guys via emails, using specific demonstrations. I've even hand picked for you. Now you ask again and I'll give you another, but you are missing the point by focusing on one example when the possibilities are infinite. If I was a computing pioneer and I had to convince you of the importance of AI and all I had was the lambda calculus, would you see it? And you ask me for another cherry-picked demo, but it is very much beyond your current understanding. It's so much beyond what you believe is possible that you ask for an example. I have shown you 10+ already. GPT turns emacs into something very powerful beyond your current comprehension. It's so profound that it will replace many of the online and offline services you may have come to take for granted. It goes way beyond that too. This is the telos and purpose of emacs. It can save free software by absorbing GPT. GPT is not a toy. Here is another demo. The below instructions were given to me by the tutor in Pen.el when I asked it for help. There are two ways to quit Emacs, the hard way, and the easy way. In the hard way, you type M-x kill-emacs, and press enter. In the easy way, you press C-x C-c. The following is a prompt that created this interactive function. #+BEGIN_SRC yaml prompt: | This is a conversation between a human and a brilliant AI. The topic is "<2>". Human: Hello, are you my <1> tutor? ### AI: Hi there. Yes I am. How can I help you? ### Human: Thanks. I have a question. <2> ### AI: I would be happy to answer your question. #+END_SRC Here is the recording of me doing that: https://asciinema.org/a/SCUhm3l11N3w5eilUfewBDCiP In the future the core prompts will be only a few bytes and most prompts we use will be defined using types alone, or subjective to previously executed prompt functions, collaboratively. A prompt may be defined by type names alone, plus the version of a LM; the rest is inferred or subjective to peer to peer prompts: #+BEGIN_SRC emacs-lisp (defprompt ("short lines of code" "regex")) #+END_SRC You haven't yet understood the profundity of GPT and doing a great disservice to free software by stifling imaginary programming inside emacs. Eli is right in that these LMs are not copy pasting. They are indeed generating from a statistical model. I know many people who are creating their own models on themselves. There will be infinitely many of these models in the future. GNU needs one. LMs may be trained on subsets of the code from GitHub to ensure there is no possibility for derivative works from conflicting licensed code. > You are more or less proposing the same conflict to come to Emacs and > I did not see where is your solution? Emacs is dead without GPT. That's why I raised the issue. It's dead because it can be imagined by LMs and will lose its power. It'll be just another imagined environment. Software is changing and Emacs can't miss out because it represents freedom. > So how? As I have said countless times in the past and is clearly visible from their website, EleutherAI has MIT and Apache licensed alternative models which you can freely download. Here is a completer you can freely use: https://github.com/samrawal/emacs-secondmate This will not replace everything from =gpt3demo.com=. You're failing to see the full picture here. It's absolutely vital for emacs' survival to have GPT incorporated. Make it happen. > As soon as anything is published in public > without compliance to licenses it generates > problems. Prompts are completely at the license of the person who created it, even if they are queries to GPT3. Just as I can write SQL to Microsoft SQL Server and license my SQL as GPL3. Prompts are code and I'm talking about a new type of programming that is nonexistent in emacs. My suggestions: - Create prompts database - So people can collaborate on open source prompts - So people can extend emacs with language models - So people know it's ok to use their imagination and emacs supports creative intelligence - Integrate prompt functions into emacs somehow - defprompt - Optionally ship GPT-neo and GPT-j with emacs - Consider creating a prompting server - Consider a database for saving generations =Pen.el= is GPL3. There's nothing wrong with typing on a keyboard so it's fully compliant with licensing. =Pen.el= allows you to select the completion engine and you may use a libre completion engine such as GPT-j, GPT-neo or GPT-neox. > "Prove me wrong" Do me a favour and do some research yourself. I have too much to do. Sincerely, Shane Shane Mulligan How to contact me: 🇦🇺 00 61 421 641 250 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> mullikine@gmail.com On Sat, Jul 24, 2021 at 12:50 AM Jean Louis <bugs@gnu.support> wrote: > * Eli Zaretskii <eliz@gnu.org> [2021-07-23 14:51]: > > > According to online reviews chunks of code is copied even verbatim and > > > people find from where. > > > > That cannot be true. It is nonsense to copy unrelated code into a > > program and tell people this is what they should use. > > I wonder how sure you are in that, did you do the online research? It > is not about related or unrelated, I do believe that AI finds and > generates related code. But > > Here are references disputing how "it cannot be true": > > https://hacker-news.news/post/27710287 > > > https://mmacvicar.medium.com/it-is-best-if-copilot-copies-everything-d84506128e5a > > > https://loudlabs.nl/news/githubs-commercial-ai-tool-was-built-from-open-source-code/ > > > > If code compiles or not is irrelevant. If one runs it or not is also > > > irrelevant, code need not even run. > > > > A feature or service that is based on this idea will never fly, > > believe me. Which program would want to have code pasted into his/her > > program that would cause compilation errors or, worse, break it at run > > time? > > Of course people want code to fun. Just that copyright laws don't > handle technical functionality. It is irrelevant if program works or > does not work. There are thousands of copyrighted programs that cannot > work any more as devices are not on the market, they are still under > copyright. > > > > I do not believe that any of the AI-s so far "extract ideas". I never > > > heard of it. Which algorithms is there on this planet that may extract > > > idea? > > > > That's a very general question, it is impossible to answer it in a > > post to a mailing list. If you are really interested, you will have > > to read up on that. But you are wrong in your beliefs. > > > > > If newly generated code is modification from other code, what we know > > > now that it is, and is based on, that requires licensing > > > attributions. > > > > Once again, your assumptions are all wrong, so your conclusions are > > also wrong. Why not try one of these services and see what they > > actually do, before you pass your (quite harsh) judgment on them, and > > on the modern state of AI in general? > > I can hear you how I am wrong, conclusions are wrong, though I gave > you references enough to research it on Internet that will tell that > there are possible serious licensing problems with such generated > code. > > > > That licenses are relevant one can see from online discussions related > > > to Github Copilot: > > > > That people ask these questions and discuss this doesn't mean the > > problem is real. many people don't really understand what copyright > > means and how to apply it to program code. > > Well said! Though that is not relevant. > > Question is very particular, specific and concrete: > ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ > > How does Pen.el and background AI services ensure of licensing > compliance? > > I would appreciate if you find solution to that or stay on that > subject, as if I am wrong or right is not relevant, what I wish is to > have assurance that it is free software. Prove me wrong by providing > exact references in not only on country's law but also other > countries' laws, the lows that make it legal, or how otherwise the > legality of such code is justified and how users may get free > software. > > For example you may wish to mention "fair use" and on the other hand > similar laws must be found in other countries that would justify it to > be free software. > > As long as you don't tackle those subjects there is no legal solution > for Pen.el and background AI to be used with assurance that software > is truly free software. > > > > Jean > > Take action in Free Software Foundation campaigns: > https://www.fsf.org/campaigns > > In support of Richard M. Stallman > https://stallmansupport.org/ > > > > > [-- Attachment #2: Type: text/html, Size: 13153 bytes --] ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 13:39 ` Shane Mulligan @ 2021-07-23 14:39 ` Jean Louis 2021-07-26 0:16 ` Richard Stallman 1 sibling, 0 replies; 46+ messages in thread From: Jean Louis @ 2021-07-23 14:39 UTC (permalink / raw) To: Shane Mulligan; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, rms * Shane Mulligan <mullikine@gmail.com> [2021-07-23 16:40]: > Hi Jean, Eli, GNU, > > > "open source" > I am referring to free software in the spirit > of GNU. Free as in freedom, from oppression, > from an attack against creative and cognitive > intelligence. Alright, though in GNU project we don't use the term "Open Source" as it was never about it. The term is "free software". Open Source is today vague, people use the term "Open" for things which are not so open, including source code, there is open source from Microsoft which is proprietary and yet called "open source", there we have now OpenAI which is not free software, and so on. See "Open": https://www.gnu.org/philosophy/words-to-avoid.html#Open > Now you ask again and I'll give you another, but you are missing the > point by focusing on one example when the possibilities are > infinite. I hope that generated code will not take longer time to verify then writing it by hand. Not to mistake me, if the databases are free as in the definition of free software and your code is free, then I am definitely for that, and I like AI, we have too little of the artificial intelligence in 21st century. We are under developed civilization in that regard. The movie 2001: A Space Odyssey was made in 1968, prediction was already that we would have space ships with Hal AI that guides us, but we are not far from bedroom with Amazon spying "AI" devices. It is very important to have all parts free as in free software definition. Review again: https://www.gnu.org/philosophy/free-sw.html > GPT turns emacs into something very powerful beyond your current > comprehension. It's so profound that it will replace many of the > online and offline services you may have come to take for > granted. It goes way beyond that too. Yes, I am asking for less abstract, more practical examples. I have no use from the hype. > Here is another demo. > The below instructions were given to me by the > tutor in Pen.el when I asked it for help. > > There are two ways to quit Emacs, the hard way, and the easy way. > In the hard way, you type M-x kill-emacs, and press enter. > In the easy way, you press C-x C-c. I have to look on it realistically from my angle, so I do not see any use in this example. I see that AI guessed something and generated text. I would not agree that M-x kill-emacs is hard way, and that C-x C-c is easy way. I would rather say that easy way is to choose File and Quit menu options. > The following is a prompt that created this interactive function. > > #+BEGIN_SRC yaml > prompt: | > This is a conversation between a human and a brilliant AI. > The topic is "<2>". > > Human: Hello, are you my <1> tutor? > ### > AI: Hi there. > Yes I am. > How can I help you? > ### > Human: Thanks. I have a . <2> > ### > AI: I would be happy to answer your question. > #+END_SRC I am sorry, I wish to see example of usefulness. I will go over your previous examples. It is definitely possible that I neglected it, but you know from beginning that I am interested in this. I have my reasons why I am interested as I do generate a lot of text and I wish to spare my writings. In the above quote I do not see that prompt, and how is relevant to how to quit Emacs. > Here is the recording of me doing that: > > https://asciinema.org/a/SCUhm3l11N3w5eilUfewBDCiP I have clicked on that link and could not find exact reference. I found "haskell lsp with HIE", something about "htop" and "stackexchange". > A prompt may be defined by type names alone, > plus the version of a LM; the rest is inferred > or subjective to peer to peer prompts: > > #+BEGIN_SRC emacs-lisp > (defprompt ("short lines of code" "regex")) > #+END_SRC > > You haven't yet understood the profundity of GPT and > doing a great disservice to free software by > stifling imaginary programming inside emacs. The above hype paragraphs are suspiciously AI-looking. > > You are more or less proposing the same conflict to come to Emacs and > > I did not see where is your solution? > Emacs is dead without GPT. That's why I raised the issue. It's dead > because it can be imagined by LMs and will lose its power. It'll be > just another imagined environment. Software is changing and Emacs > can't miss out because it represents freedom. That was not context of my question. Did you read last email to Eli about it? It seem like you either ignored my question or you keep using AI to generate hype about it. Julia Reda from Germany is at least trying to answer my question related to licensing compliance here: https://juliareda.eu/2021/07/github-copilot-is-not-infringing-your-copyright/ So there are at least ways to go to understand how it complies or could comply to licenses or be liberated from licenses. > You're failing to see the full picture here. > It's absolutely vital for emacs' survival to > have GPT incorporated. Make it happen. I'm having a hard time following your proposal. I'm not sure what you're asking. But I am too old to understand it. Though it is not my age that matters. It is your experience, your knowledge, your education, your understanding, your wisdom. In all of these you and me have to be the best judge of what is good for you and me and what is not. The above paragraph was created by AI with small corrections. It says nothing just as the above quoted paragraph says nothing essential. "full picture", "absolutely vital", "Emacs survival", "Make it happen" -- that is sales pitch. And I am sales manager btw. That AI is useful in general, no need to convince me. Question was about licensing and I see that some activists like Julia Reda, which I have contacted previously in relation to copyright issues in EU, have found legal justifications for licensing compliance. That is what I wanted to know. The issue is however not closed with the assumptions of Julia Reda, as she may know EU laws, but not all jurisdictions are quite aligned, so we have still to be vigilant and follow up on that. IMHO, you should incorporate justifications used by Mrs. Reda in your commentary or README files with references so that licensing becomes clear for future readers. > > As soon as anything is published in public > > without compliance to licenses it generates > > problems. > Prompts are completely at the license of the > person who created it, even if they are > queries to GPT3. If prompts never appear in published works those are not relevant. If they do appear, their licensing compliance is assured by programmer or author. Even prompts could be coming from proprietary software. > My suggestions: > - Create prompts database > - So people can collaborate on open source prompts > - So people can extend emacs with language models > - So people know it's ok to use their imagination and emacs supports > creative intelligence > - Integrate prompt functions into emacs somehow? > - defprompt > - Optionally ship GPT-neo and GPT-j with emacs > - Consider creating a prompting server > - Consider a database for saving generations I find it all goods ideas, I just wish I could see more practical example: - prompts database, should be on the website? Where? Hosted by which party? Is it centralized or decentralized? - collaboration by which means? Email, chat? Website forum? How exactly? - I understand your 3rd point. - GPT-neo and GPT-j is how big? -- Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 13:39 ` Shane Mulligan 2021-07-23 14:39 ` Jean Louis @ 2021-07-26 0:16 ` Richard Stallman 2021-07-26 0:28 ` Shane Mulligan 1 sibling, 1 reply; 46+ messages in thread From: Richard Stallman @ 2021-07-26 0:16 UTC (permalink / raw) To: Shane Mulligan; +Cc: eliz, stefan, emacs-tangents, bugs [[[ To any NSA and FBI agents reading my email: please consider ]]] [[[ whether defending the US Constitution against all enemies, ]]] [[[ foreign or domestic, requires you to follow Snowden's example. ]]] > GPT turns emacs into something very powerful > beyond your current comprehension. It's so > profound that it will replace many of the > online and offline services you may have come > to take for granted. It goes way beyond that too. Unfortunately, telling me that something is "powerful beyond [my] current comprehension" does not help me start to comprehend any of it. Would you like to name some of the services that GPT would replace? I might learn something concrete from that. > Here is the recording of me doing that: > https://asciinema.org/a/SCUhm3l11N3w5eilUfewBDCiP I looked at that page, but I have no idea what it means. The page shows three boxes side by side. Each seems to contain some code, or maybe parameter specs, in a language I don't know. I clicked on the first box and it brought me to a similar page with three other boxes. It tasks about "asciicasts" but I don't know what that means. If it is something to be viewed, how can I do so? -- Dr Richard Stallman (https://stallman.org) Chief GNUisance of the GNU Project (https://gnu.org) Founder, Free Software Foundation (https://fsf.org) Internet Hall-of-Famer (https://internethalloffame.org) ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-26 0:16 ` Richard Stallman @ 2021-07-26 0:28 ` Shane Mulligan 2021-07-30 3:20 ` Shane Mulligan 0 siblings, 1 reply; 46+ messages in thread From: Shane Mulligan @ 2021-07-26 0:28 UTC (permalink / raw) To: rms; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, Jean Louis [-- Attachment #1: Type: text/plain, Size: 2511 bytes --] Hey Richard and all. I have just participated in the Augment Minds unconference and have a recorded demo of Pen.el I will also be presenting the demo to Nat Friedman. I have made some references to the new codex model and how it has stolen the inspiration from Free software. The point I'm making is this: Pen.el and software which combines GPT into the operating system is the future and I'm alerting GNU to this first but I'm also showing GitHub. This is for the following reasons - The Copilot/codex model is a disgrace - We need an free repository of prompts and prompt functions for emacs I hope the demo which I will send in the next day or two (or whenever it becomes available) will be informative. It will be easier than the asciicast. Thank you. Shane Mulligan How to contact me: 🇦🇺 00 61 421 641 250 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> mullikine@gmail.com On Mon, Jul 26, 2021 at 12:16 PM Richard Stallman <rms@gnu.org> wrote: > [[[ To any NSA and FBI agents reading my email: please consider ]]] > [[[ whether defending the US Constitution against all enemies, ]]] > [[[ foreign or domestic, requires you to follow Snowden's example. ]]] > > > GPT turns emacs into something very powerful > > beyond your current comprehension. It's so > > profound that it will replace many of the > > online and offline services you may have come > > to take for granted. It goes way beyond that too. > > Unfortunately, telling me that something is "powerful beyond [my] > current comprehension" does not help me start to comprehend any of it. > > Would you like to name some of the services that GPT would replace? > I might learn something concrete from that. > > > Here is the recording of me doing that: > > > https://asciinema.org/a/SCUhm3l11N3w5eilUfewBDCiP > > I looked at that page, but I have no idea what it means. The page > shows three boxes side by side. Each seems to contain some code, or > maybe parameter specs, in a language I don't know. I clicked on the > first box and it brought me to a similar page with three other boxes. > > It tasks about "asciicasts" but I don't know what that means. > If it is something to be viewed, how can I do so? > > > > -- > Dr Richard Stallman (https://stallman.org) > Chief GNUisance of the GNU Project (https://gnu.org) > Founder, Free Software Foundation (https://fsf.org) > Internet Hall-of-Famer (https://internethalloffame.org) > > > [-- Attachment #2: Type: text/html, Size: 4961 bytes --] ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-26 0:28 ` Shane Mulligan @ 2021-07-30 3:20 ` Shane Mulligan 2021-07-30 6:55 ` Jean Louis 0 siblings, 1 reply; 46+ messages in thread From: Shane Mulligan @ 2021-07-30 3:20 UTC (permalink / raw) To: rms; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, Jean Louis [-- Attachment #1.1: Type: text/plain, Size: 3705 bytes --] Hey guys. In the last week I have been writing a thesis for Imaginary Programming, which aims to make all of this clear and formalised. I am very sorry if I have sounded frustrated, but I think that this is so important for free software and a GPL-4 may be required to protect people, but also that Copilot and OpenAI's Codex and GPT-3 models infringe upon the spirit of GPT-3 code. I will attach the thesis into this email. https://github.com/semiosis/imaginary-programming-thesis/blob/master/thesis.org I am working around the clock to finish this thesis and have it published, but it's really important to have these protections in place before the huge suite of SASS services and Microsoft Apps hit the market which are using Copilot and Codex to generate derivative works and applications built upon the backs of free software developers. Thank you. Shane Mulligan How to contact me: 🇦🇺 00 61 421 641 250 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> mullikine@gmail.com On Mon, Jul 26, 2021 at 12:28 PM Shane Mulligan <mullikine@gmail.com> wrote: > Hey Richard and all. > > I have just participated in the Augment Minds unconference and have a > recorded demo of Pen.el > > I will also be presenting the demo to Nat Friedman. I have made some > references to the new codex model and how it has stolen the inspiration > from Free software. > > The point I'm making is this: Pen.el and software which combines GPT into > the operating system is the future > and I'm alerting GNU to this first but I'm also showing GitHub. This is > for the following reasons > > - The Copilot/codex model is a disgrace > - We need an free repository of prompts and prompt functions for emacs > > I hope the demo which I will send in the next day or two (or whenever it > becomes available) will be informative. It will be easier than the > asciicast. > > Thank you. > > Shane Mulligan > > How to contact me: > 🇦🇺 00 61 421 641 250 > 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> > mullikine@gmail.com > > > On Mon, Jul 26, 2021 at 12:16 PM Richard Stallman <rms@gnu.org> wrote: > >> [[[ To any NSA and FBI agents reading my email: please consider ]]] >> [[[ whether defending the US Constitution against all enemies, ]]] >> [[[ foreign or domestic, requires you to follow Snowden's example. ]]] >> >> > GPT turns emacs into something very powerful >> > beyond your current comprehension. It's so >> > profound that it will replace many of the >> > online and offline services you may have come >> > to take for granted. It goes way beyond that too. >> >> Unfortunately, telling me that something is "powerful beyond [my] >> current comprehension" does not help me start to comprehend any of it. >> >> Would you like to name some of the services that GPT would replace? >> I might learn something concrete from that. >> >> > Here is the recording of me doing that: >> >> > https://asciinema.org/a/SCUhm3l11N3w5eilUfewBDCiP >> >> I looked at that page, but I have no idea what it means. The page >> shows three boxes side by side. Each seems to contain some code, or >> maybe parameter specs, in a language I don't know. I clicked on the >> first box and it brought me to a similar page with three other boxes. >> >> It tasks about "asciicasts" but I don't know what that means. >> If it is something to be viewed, how can I do so? >> >> >> >> -- >> Dr Richard Stallman (https://stallman.org) >> Chief GNUisance of the GNU Project (https://gnu.org) >> Founder, Free Software Foundation (https://fsf.org) >> Internet Hall-of-Famer (https://internethalloffame.org) >> >> >> [-- Attachment #1.2: Type: text/html, Size: 8049 bytes --] [-- Attachment #2: thesis.org --] [-- Type: application/octet-stream, Size: 26754 bytes --] * Imaginary programming is a new programming paradigm based on language models ** Abstract Imaginary code is code who's behaviour is influenced by LMs. The side effects or return values of imaginary code, therefore, are imagined by a LM, but may also be used to facilitate the imagination of the programmer and may be considered to be a bicycle for the imagination. This is very obvious when interacting with an imaginary REPL. I will attempt to formalise imaginary programming, make some demonstrations of programming within this paradigm and explore some useful data structures and algorithms that are both impurely and purely imaginary. I'll also give an example of an imaginary programming language that I have created (perhaps the first of its kind), examplary. The motivation for formalising imaginary programming is not purely academic. Imaginary code needs to be recognised as code so that it may be protected by GPL. I also posit that models of NL, if trained on source code, create a holographic representation of the software, which I argue is a derivative work and a reflection of the original code. I argue that a holographic representation of software both within (author inspiration) and without (how it is used) is just another representation of the software, alongside the original source code, just as functions may be represented differently. ** Introduction The recently burgeoning and soon to diminish programming paradigm of prompt engineering is about to be superseded by prompt-tuning and the fine-tuning of LMs which will further occlude the way that software works. Prompt engineering has barely had it's time in the spotlight and as a result has not established itself as a sovereign programming paradigm. However, imaginary programming is a broader definition that encapsulates all programming that solicits LMs and uses their output to effect change in a program's logic and will outlast prompt engineering as a useful concept. In contrast with imaginary code, ordinary code has not yet been contaminated by a LM, and we say that it has no imaginary dimension to it. Impure imaginary code is where ordinary code intersects with pure imaginary code. An impure imaginary function is a function that queries a LM to directly affect its own logic or output. We say that an impure imaginary function is grounded [to reality] because it's connecting base reality to a LM. The output and behaviour of an impure imaginary function is directly influenced by base reality plus a query to a LM. The query [or prompt] to the LM may be in part constructed manually through prompt engineering, or in part constructed automatically via prompt tuning, or in part constructed or eliminated by the fine-tuning of a LM. Even after fine-tuning, there is still a query to be formulated to the LM, and that query may indeed be the empty string. Considering that large LMs such as GPT-3 can perform multiple tasks, the process of refining a query through prompt-engineering, prompt-tuning or fine-tuning also characterises the expected output from the LM. All that is left is to map a prompt along with its associated LM to a function and then you have a prompt function. Prompt functions reconcile LMs with programming languages. A prompt function is just a function that prompts a LM and may optionally be parameterized with template variables that are substituted into the prompt or also contain hyperparameters to affect the LM's operation. Such functions are the basis for services such as GitHub Copilot. ** Impure imaginary code is useful Impure imaginary code is very obviously useful as such code utilises LMs that are trained to perform useful tasks. GPT-3, for example, is a generalist and only requires a tiny amount of prompt design and/or fine-tuning to direct it to perform the task you want. Following are some demonstrations of using impure imaginary code to construct part of an imaginary programming environment, perform code generation, transpile code and translate world languages. *** Useful impure imaginary functions **** With one =Pen.el= system The following prompt function definition function associates a prompt with a LM (OpenAI's GPT-3 davinci) and defines the parameters for a function in emacs lisp. #+BEGIN_SRC yaml -n :async :results verbatim code title: bash one liner generator on OS from natural language doc: Get a bash one liner on OS from natural language notes: - "rlprompt is used here outside of pen.el" rlprompt: nlsh <1> prompt: | # List of one-liner shell commands for <1>. # Language: Shell # Operating System: <1> Input: Print the current directory Output: pwd ### Input: List files Output: ls -l ### Input: Change directory to /tmp Output: cd /tmp ### repeater: | Input: {} Output: lm-command: "openai-complete.sh" engine: davinci temperature: 0.8 max-tokens: 60 top-p: 1 stop-sequences: - "###" vars: - Operating System - command examples: - Arch Linux - Install package postprocessor: 'sed ''s/^Output: //''' conversation-mode: true #+END_SRC The following is the generated documentation for the interactive prompt function in emacs. #+BEGIN_SRC text -n :async :results verbatim code pf-bash-one-liner-generator-from-natural-language is an interactive function defined in pen-example-config.el. Signature (pf-bash-one-liner-generator-from-natural-language &optional TASK-DESCRIPTION &key NO-SELECT-RESULT) Documentation bash one liner generator from natural language Get a bash one liner from natural language path: - /home/shane/source/git/spacemacs/prompts/prompts/bash-one-liner.prompt examples: - shift last argument Key Bindings This command is not in any keymaps. References pf-bash-one-liner-generator-from-natural-language is unused in pen-example-config.el. #+END_SRC Below is the generated interactive function in emacs lisp. #+BEGIN_SRC emacs-lisp -n :async :results verbatim code (lambda (&optional task-description &rest --cl-rest--) "bash one liner generator from natural language\nGet a bash one liner from natural language\n\npath:\n- /home/shane/source/git/spacemacs/prompts/prompts/bash-one-liner.prompt\n\nexamples:\n- shift last argument\n\n(fn &optional TASK-DESCRIPTION &key NO-SELECT-RESULT)" (interactive (list (if mark-active (pen-selected-text) (if nil (etv "shift last argument") (read-string-hist "task-description: " "shift last argument"))))) (let* ((no-select-result (car (cdr (plist-member --cl-rest-- ':no-select-result))))) (progn (let ((--cl-keys-- --cl-rest--)) (while --cl-keys-- (cond ((memq (car --cl-keys--) '(:no-select-result :allow-other-keys)) (setq --cl-keys-- (cdr (cdr --cl-keys--)))) ((car (cdr (memq ':allow-other-keys --cl-rest--))) (setq --cl-keys-- nil)) (t (error "Keyword argument %s not one of (:no-select-result)" (car --cl-keys--)))))) (cl-block pf-bash-one-liner-generator-from-natural-language (let* ((final-prompt "The following is a list of one-liners for the linux command-line:\n\n# get newest file in directory bash\n$ ls -t * | head -1\n###\n# Find with invert match - e.g. find every file that is not mp3\n$ find . -name '*' -type f -not -path '*.mp3'\n###\n# Recursively remove all \"node_modules\" folders\n$ find . -name \"node_modules\" -exec rm -rf '{}' +\n###\n# <1>\n$\n") (final-max-tokens (str (if (variable-p 'max-tokens) (eval 'max-tokens) 60))) (final-stop-sequences (if (variable-p 'stop-sequences) (eval 'stop-sequences) '("###"))) (vals (mapcar 'str (if (not (interactive-p)) (progn (cl-loop for sym in '(task-description) for iarg in '((if mark-active (pen-selected-text) (if nil (etv "shift last argument") (read-string-hist "task-description: " "shift last argument")))) collect (let* ((initval (eval sym))) (if (and (not initval) iarg) (eval iarg) initval)))) (cl-loop for v in '(task-description) until (eq v '&key) collect (eval v))))) (vals (cl-loop for tp in (-zip-fill nil vals 'nil) collect (let* ((v (car tp)) (pp (cdr tp))) (if pp (pen-sn pp v) v)))) (i 1) (final-prompt (pen-expand-template final-prompt vals)) (prompt-end-pos (or (byte-string-search "<:pp>" "The following is a list of one-liners for the linux command-line:\n\n# get newest file in directory bash\n$ ls -t * | head -1\n###\n# Find with invert match - e.g. find every file that is not mp3\n$ find . -name '*' -type f -not -path '*.mp3'\n###\n# Recursively remove all \"node_modules\" folders\n$ find . -name \"node_modules\" -exec rm -rf '{}' +\n###\n# <1>\n$\n") (string-bytes final-prompt))) (final-prompt (string-replace "<:pp>" "" final-prompt)) (final-prompt (if nil (sor (pen-snc nil final-prompt) (concat "prompt-filter " nil " failed.")) final-prompt)) (pen-sh-update (or pen-sh-update (>= (prefix-numeric-value current-global-prefix-arg) 4))) (shcmd (pen-log (concat (sh-construct-envs `(("PEN_PROMPT" ,(pen-encode-string final-prompt)) ("PEN_LM_COMMAND" ,"openai-complete.sh") ("PEN_ENGINE" ,"davinci") ("PEN_MAX_TOKENS" ,(pen-expand-template final-max-tokens vals)) ("PEN_TEMPERATURE" ,(pen-expand-template (str 0.8) vals)) ("PEN_STOP_SEQUENCE" ,(pen-encode-string (str (if (variable-p 'stop-sequence) (eval 'stop-sequence) "###")))) ("PEN_TOP_P" ,1) ("PEN_CACHE" ,nil) ("PEN_N_COMPLETIONS" ,5) ("PEN_END_POS" ,prompt-end-pos))) " " "upd lm-complete"))) (resultsdirs (cl-loop for i in (number-sequence 1 1) collect (progn (message (concat "pf-bash-one-liner-generator-from-natural-language" " query " (int-to-string i) "...")) (let ((ret (pen-prompt-snc shcmd i))) (message (concat "pf-bash-one-liner-generator-from-natural-language" " done " (int-to-string i))) ret)))) (results (-uniq (flatten-once (cl-loop for rd in resultsdirs collect (if (sor rd) (->> (glob (concat rd "/*")) (mapcar 'e/cat) (mapcar (lambda (r) (if (and nil (sor nil)) (pen-sn nil r) r))) (mapcar (lambda (r) (if (and (variable-p 'prettify) prettify nil (sor nil)) (pen-sn nil r) r))) (mapcar (lambda (r) (if (not nil) (s-trim-left r) r))) (mapcar (lambda (r) (if (not nil) (s-trim-right r) r))) (mapcar (lambda (r) (cl-loop for stsq in final-stop-sequences do (let ((matchpos (string-search stsq r))) (if matchpos (setq r (s-truncate matchpos r ""))))) r))) (list (message "Try UPDATE=y or debugging"))))))) (result (if no-select-result (length results) (cl-fz results :prompt (concat "pf-bash-one-liner-generator-from-natural-language" ": ") :select-only-match t)))) (if no-select-result results (if (interactive-p) (cond ((>= (prefix-numeric-value current-prefix-arg) 4) (etv result)) ((and nil mark-active) (replace-region result)) ((or nil nil) (insert result)) (t (etv result))) result))))))) #+END_SRC The above function creates a NL shell. This enables you to generate shell commands based on NL and it is parameterized to enable you to specify the operating system that the commands generated should run on. #+BEGIN_SRC emacs-lisp -n :async :results raw (list2str (pf-bash-one-liner-generator-on-os-from-natural-language "Arch Linux" "Disable firewall" :no-select-result t)) #+END_SRC Here is a list of suggestions generated from the above prompt function. #+BEGIN_SRC text -n :async :results verbatim code iptables -F iptables -P OUTPUT DROP sed -i 's/^[ \t]*firewall=.*$/firewall=0/' /etc/sysconfig/iptables systemctl stop iptables.service sudo systemctl stop iptables sudo ufw disable #+END_SRC You may also run it as a REPL. https://semiosis.github.io/posts/imaginary-programming-with-gpt-3/ #+BEGIN_SRC yaml -n :async :results verbatim code title: Code interpreter kickstarter future-titles: - Code interpreter kickstarter doc: Given a line of code, infer the result of running that code prompt-version: 4 prompt: | Code examples: Language: Python Input: print(random.randint(0,9)) Output: 5 ### Language: Bash Input: Str="Learn Linux from LinuxHint"; subStr=${Str:6:5} Output: Linux ### repeater: | Language: <1> Input: {} Output: issues: engine: davinci temperature: 0.8 max-tokens: 60 top-p: 1 stop-sequences: - "##" - "\n" vars: - language - code examples: - haskell - '"Hello" ++ " " ++ "World"' prefer-external: true external: iol similarity-test: string-equal quality-script: levenshtein -s conversation-mode: true n-test-runs: 5 #+END_SRC #+BEGIN_SRC emacs-lisp -n :async :results raw (car (pf-code-interpreter-kickstarter "Haskell" "\"Hello\" ++ \" \" ++ \"World\"" :no-select-result t)) #+END_SRC #+BEGIN_SRC text -n :async :results verbatim code Hello World #+END_SRC **** With two =Pen.el= systems ***** Using a common language model Translating communications with a world language translation prompt function. #+BEGIN_SRC yaml -n :async :results verbatim code title: Translate from world language X to Y prompt-version: 3 doc: This prompt translates English text to any world langauge prompt: | ### # English: Hello # Russian: Zdravstvuyte # Italian: Salve # Japanese: Konnichiwa # German: Guten Tag # French: Bonjour # Spanish: Hola ### # English: Happy birthday! # French: Bon anniversaire ! # German: Alles Gute zum Geburtstag! # Italian: Buon compleanno! # Indonesian: Selamat ulang tahun! ### # <1>: <3> # <2>: engine: davinci temperature: 0.5 max-tokens: 200 top-p: 1 stop-sequences: - "#" vars: - from-language - to-language - phrase preprocessors: - cat - cat - pen-s onelineify postprocessor: pen-s unonelineify examples: - English - French - Goodnight var-defaults: - "(or (sor (nth 0 (pf-get-language (pen-selected-text) :no-select-result t))) (read-string-hist \"Pen From language: \"))" - "(read-string-hist \"Pen To language: \")" - "(pen-selected-text)" filter: on #+END_SRC A demonstration of two people who understand different world languages using a common LM to understand one another. #+NAME: fromenglish #+BEGIN_SRC text -n :async :results verbatim code Happy birthday To you #+END_SRC #+BEGIN_SRC emacs-lisp -n :async :results code raw ;; Alice translates into french for Bob (car (pf-translate-from-world-language-x-to-y "English" "French" "Happy birthday\nTo you" :no-select-result t)) #+END_SRC #+NAME: fromfrench #+BEGIN_SRC text -n :async :results verbatim code Bon anniversaire A vous #+END_SRC #+BEGIN_SRC text -n :async :results verbatim code Merci beaucoup #+END_SRC #+BEGIN_SRC emacs-lisp -n :async :results code raw ;; Bob translates back into English for Alice (car (pf-translate-from-world-language-x-to-y "French" "English" "Merci\nbeaucoup" :no-select-result t)) #+END_SRC #+BEGIN_SRC text -n :async :results verbatim code Thank you! #+END_SRC https://asciinema.org/a/7YnSnrrLgbiFlyMyYxBgaZYUb #+BEGIN_EXPORT html <!-- Play on asciinema.com --> <!-- <a title="asciinema recording" href="https://asciinema.org/a/7YnSnrrLgbiFlyMyYxBgaZYUb" target="_blank"><img alt="asciinema recording" src="https://asciinema.org/a/7YnSnrrLgbiFlyMyYxBgaZYUb.svg" /></a> --> <!-- Play on the blog --> <script src="https://asciinema.org/a/7YnSnrrLgbiFlyMyYxBgaZYUb.js" id="asciicast-7YnSnrrLgbiFlyMyYxBgaZYUb" async></script> #+END_EXPORT ***** With different language models - GPT-neo and GPT-3? - curie vs davinci? - Generate a story about a meeting with one prompt - Summarize with bullet points - meeting-bullets-to-summary.prompt *** An impure imaginary data structure **** With one =Pen.el= system - Natural language database entry **** With two =Pen.el= systems - Database prompt **** With three =Pen.el= systems - Database prompt *** TODO Find a useful impure imaginary algorithm **** With one =Pen.el= system - Translate from X to Y - Backtranslate from Y to X Find a better prompt? **** With two =Pen.el= systems **** With three =Pen.el= systems ** Pure imaginary code is useful Pure imaginary programming is a type of programming where the original language models may not even be known. I demonstate that collaborative pure imaginary programming is useful. *** Translation between two =Pen.el= systems with different language models A common library of pure imaginary functions. #+BEGIN_SRC emacs-lisp -n :async :results verbatim code ("translate" "prose" "from" "to") #+END_SRC Pure imaginary functions can be composed. #+BEGIN_SRC emacs-lisp -n :async :results verbatim code ("translate" ("make analogy about" "topic") "from" "to") #+END_SRC ** Imaginary programming languages are required to work with language models *** Examplary - Part of it is task-oriented, which defers imagination to a language model to understand what it means. - Part of it is example-oriented, which is pure-imaginary. *** Example-oriented #+BEGIN_SRC emacs-lisp -n :async :results verbatim code ;; Convert lines to regex. (xl-defprompt ("lines of code" regex) ;; :task "Convert lines to regex" ;; Generate input with this ;; :gen "examplary-edit-generator shane" :gen examplary-edit-generator :filter "grex" ;; The third argument (if supplied) should be incorrect output (a counterexample). ;; If the 2nd argument is left out, it will be generated by the command specified by :external :examples (("example 1\nexample2") ("example 2\nexample3" "^example [23]$") ("pi4\npi5" "^pi[45]$" "pi4\npi5")) :lm-command "openai-complete.sh") #+END_SRC *** Task oriented #+BEGIN_SRC emacs-lisp -n :async :results verbatim code ("translate" ("make analogy about" "topic") "from" "to") #+END_SRC ** Projecting the code back to the starting LM is possible - Semantic search on existing documents - Semantic search on existing functions in emacs ** Language models encode holographic representations of software It's important to avoid mixing training data of varying licenses when training LMs. One risk is that in the future, as holographic representations of software are used more in place of running original source code (i.e. as LMs are used more to simulate software), a software's hologram is more likely to be used in ways that violate the original license or the spirit of the license. LMs bring with them understanding of the way software is used, and also an understanding of the inspiration that went into designing that software. The issue is that this is all automated and right now new software companies are staking their future on LMs and using said models to their fullest. Therefore, the inexorable conclusion is that software that has been used to train these models will be used holographically, perhaps more than even from their original software and their holographic representation that encodes the value of the software (the way it's used as opposed to written) is what's more important and that's is what is being exploited. If the original code of an example of free software was part of the training data of a NN alongside software of other conflicting licenses then that effectively relicences the same software without consent, going forward into the future. *** Generating parts of emacs with GPT-3 I am able to generate parts of GPL protected software using LMs and can query the LMs as to how they are used. Therefore, the software exists now in the latent space of a language model in the form of a hologram, within and without the source code. Language models encode contrived associations made between different pieces of software in order to create an accurate model that is useful for simulation, code generation, code understanding and modelling the usage of software. - The holographic representation *** =0.9 / 1= is still stealing ** Counter arguments *** It's not imaginary, it's just... English? more like, stochastic programming? Imaginary programming is more of an activity and a style of programming and is not really concerned with the amount of uncertainty. Your code might take a trip through someone else's LM along the way and be projected back to your own. That means that some of the logic is completely obscured and you have to make assumptions. You may collaborate on a user interface or program with others and since that code can't be fully understood by one person because of the veil then you are compelled to imagine in order to create something useful. A person must build their own interface from the pure imaginary functions that are shared. It's a paradigm completely made up so it's useful as far as it's useful. All this is based on this idea that we will have many finetuned and completely different transformer models and we must learn to communicate. The NeverEnding story also influenced my thoughts. Once everyone stops believing in Fantasia it ceases to exist, as does the utility of applications built in pure imaginary code. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-30 3:20 ` Shane Mulligan @ 2021-07-30 6:55 ` Jean Louis 0 siblings, 0 replies; 46+ messages in thread From: Jean Louis @ 2021-07-30 6:55 UTC (permalink / raw) To: Shane Mulligan; +Cc: Eli Zaretskii, emacs-tangents, Stefan Kangas, rms * Shane Mulligan <mullikine@gmail.com> [2021-07-30 06:20]: > Hey guys. > > In the last week I have been writing a thesis for Imaginary Programming, > which aims to make all of this clear and formalised. > > I am very sorry if I have sounded frustrated, but I think that this is so > important for free software and a GPL-4 may be required to protect people, > but also that Copilot and OpenAI's Codex and GPT-3 models infringe upon the > spirit of GPT-3 code. > > I will attach the thesis into this email. > > https://github.com/semiosis/imaginary-programming-thesis/blob/master/thesis.org Please try to see if you can help on this: FSF-funded call for white papers on philosophical and legal questions around Copilot https://www.fsf.org/blogs/licensing/fsf-funded-call-for-white-papers-on-philosophical-and-legal-questions-around-copilot -- Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 12:47 ` Jean Louis 2021-07-23 13:39 ` Shane Mulligan @ 2021-07-23 19:33 ` Eli Zaretskii 2021-07-24 3:07 ` Jean Louis 1 sibling, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-23 19:33 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, emacs-tangents, mullikine, rms > Date: Fri, 23 Jul 2021 15:47:21 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: mullikine@gmail.com, stefan@marxist.se, emacs-tangents@gnu.org, > rms@gnu.org > > * Eli Zaretskii <eliz@gnu.org> [2021-07-23 14:51]: > > > According to online reviews chunks of code is copied even verbatim and > > > people find from where. > > > > That cannot be true. It is nonsense to copy unrelated code into a > > program and tell people this is what they should use. > > I wonder how sure you are in that, did you do the online research? It > is not about related or unrelated, I do believe that AI finds and > generates related code. But > > Here are references disputing how "it cannot be true": You take everything you read in these blogs for granted? Did you actually see the original code which these allude to? > > > If code compiles or not is irrelevant. If one runs it or not is also > > > irrelevant, code need not even run. > > > > A feature or service that is based on this idea will never fly, > > believe me. Which program would want to have code pasted into his/her > > program that would cause compilation errors or, worse, break it at run > > time? > > Of course people want code to fun. Just that copyright laws don't > handle technical functionality. It is irrelevant if program works or > does not work. For copyright purposes, it doesn't. But for the programmer who uses the code it very much does. So if these services give them code that doesn't work they will not use it, and eventually will stop using the services. > > Once again, your assumptions are all wrong, so your conclusions are > > also wrong. Why not try one of these services and see what they > > actually do, before you pass your (quite harsh) judgment on them, and > > on the modern state of AI in general? > > I can hear you how I am wrong, conclusions are wrong, though I gave > you references enough to research it on Internet that will tell that > there are possible serious licensing problems with such generated > code. See above. > Question is very particular, specific and concrete: > ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ > > How does Pen.el and background AI services ensure of licensing > compliance? > > I would appreciate if you find solution to that or stay on that > subject, as if I am wrong or right is not relevant, what I wish is to > have assurance that it is free software. Prove me wrong by providing > exact references in not only on country's law but also other > countries' laws, the lows that make it legal, or how otherwise the > legality of such code is justified and how users may get free > software. I'm sorry, but I don't work for you. If you have problems with using code from these services, then the onus is on you to do the research and make up your own mind. The discussion here is not about the code these services give their users, it's whether and how Emacs can make use of those services. Emacs allows the user to write proprietary code, and there's no legal issues when the user does that. Emacs also allows the user to copy someone else's code without permission, and that's not a problem for Emacs when the user does that. > As long as you don't tackle those subjects there is no legal solution > for Pen.el and background AI to be used with assurance that software > is truly free software. You confuse "free software" with "software being used to write free programs". They are not the same. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 19:33 ` Eli Zaretskii @ 2021-07-24 3:07 ` Jean Louis 2021-07-24 7:32 ` Eli Zaretskii 2021-07-25 1:09 ` Richard Stallman 0 siblings, 2 replies; 46+ messages in thread From: Jean Louis @ 2021-07-24 3:07 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-23 22:34]: > > Here are references disputing how "it cannot be true": > > You take everything you read in these blogs for granted? Did you > actually see the original code which these allude to? In case of Copilot, Github admits: https://docs.github.com/en/github/copilot/research-recitation "This investigation demonstrates that GitHub Copilot can quote a body of code verbatim, but that it rarely does so, and when it does, it mostly quotes code that everybody quotes, and mostly at the beginning of a file, as if to break the ice.This investigation demonstrates that GitHub Copilot can quote a body of code verbatim, but that it rarely does so, and when it does, it mostly quotes code that everybody quotes, and mostly at the beginning of a file, as if to break the ice." And fact that it is "rare" does not make it a less problem for copyright purposes as the new author cannot know which part of the code has used "rare" verbatim. > > Question is very particular, specific and concrete: > > ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ > > > > How does Pen.el and background AI services ensure of licensing > > compliance? > > > > I would appreciate if you find solution to that or stay on that > > subject, as if I am wrong or right is not relevant, what I wish is to > > have assurance that it is free software. Prove me wrong by providing > > exact references in not only on country's law but also other > > countries' laws, the lows that make it legal, or how otherwise the > > legality of such code is justified and how users may get free > > software. > > I'm sorry, but I don't work for you. If you have problems with using > code from these services, then the onus is on you to do the research > and make up your own mind. The discussion here is not about the code > these services give their users, it's whether and how Emacs can make > use of those services. Emacs allows the user to write proprietary > code, and there's no legal issues when the user does that. Emacs also > allows the user to copy someone else's code without permission, and > that's not a problem for Emacs when the user does that. If you don't wish to correspond, don't, you are free. I have never said nor implied "you work for me" and I cannot see how is that relevant to the question. If you participate in discussion and respond to my question relating to licensing compliance, then provide a reference justifying its legality. Or simply say you don't have such. Your employment is not subject of my question nor relevant. I am not user of proprietary software and I don't consider options of writing proprietary software. Neither I am participating in discussion to foster ideas of creation of proprietary software. I am free software user and for that specific case I am interested how the licensing issue is solved. However, my question is at least answered by my online research as I have already found the refrences: 1. Julia Reda's reference; and 2. OpenAI_RFC-84-FR-58141.pdf https://www.uspto.gov/sites/default/files/documents/OpenAI_RFC-84-FR-58141.pdf Conclusions are: - legal justifications exists for US jurisdiction as the companies providing the AI are strong enough to find their ways, they are playing on the card as given in references above; as somebody already said, I doubt they would use "fair use" doctrine if the AI would be trained on proprietary software such as Windows; - conflict is serious and it is out there among the people and remains unsolved; AI has been trained on GPL and other free software and is used by corporations to generate new code without attributions; people complain that it is misuse of intentions of authors; - overall international legal situation is thus unclear, especially considering that free software spans the whole world, not just the US jurisdiction, as what may work within US is not same among all jurisdictions; > > As long as you don't tackle those subjects there is no legal solution > > for Pen.el and background AI to be used with assurance that software > > is truly free software. > > You confuse "free software" with "software being used to write free > programs". They are not the same. Maybe I have expressed myself in such way as not to get the point understood. It must be so, as I have finally found the first legal references myself. For Pen.el I have never made any relevance to legality question I made, and I have the pen.el repository over here and license is clear. Never mentioned it. I have not made reference to "software being used to write free programs" as a server side service I did not tackle that, it is most probably proprietary software or some versions could be free software. But that is not relevant. What is at hand is: ━━━━━━━━━━━━━━━━━━━ 1. There is pool of GPL and other free software which authors expect compliance to their licenses; 2. Large corporation is trying to use "fair use" doctrine on the pool of software to create a service; 3. Service generates new software, sometimes duplicating verbatim code; Question was and still remains largely unsolved is how authors who use newly generated code can be sure that generated software is free software and to comply to GPL and other free software licenses? Conclusion as of 2021-07-24 is that authors cannot be sure as there are legal uncertainties. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 3:07 ` Jean Louis @ 2021-07-24 7:32 ` Eli Zaretskii 2021-07-24 7:54 ` Jean Louis 2021-07-25 1:09 ` Richard Stallman 1 sibling, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 7:32 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, emacs-tangents, mullikine, rms > Date: Sat, 24 Jul 2021 06:07:18 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: mullikine@gmail.com, stefan@marxist.se, emacs-tangents@gnu.org, > rms@gnu.org > > * Eli Zaretskii <eliz@gnu.org> [2021-07-23 22:34]: > > > Here are references disputing how "it cannot be true": > > > > You take everything you read in these blogs for granted? Did you > > actually see the original code which these allude to? > > In case of Copilot, Github admits: > https://docs.github.com/en/github/copilot/research-recitation Yes, and there are cases of real code stealing out there. The only thing that proves is that some mistaken or dishonest operators can do this. > > > I would appreciate if you find solution to that or stay on that > > > subject, as if I am wrong or right is not relevant, what I wish is to > > > have assurance that it is free software. Prove me wrong by providing > > > exact references in not only on country's law but also other > > > countries' laws, the lows that make it legal, or how otherwise the > > > legality of such code is justified and how users may get free > > > software. > > > > I'm sorry, but I don't work for you. If you have problems with using > > code from these services, then the onus is on you to do the research > > and make up your own mind. The discussion here is not about the code > > these services give their users, it's whether and how Emacs can make > > use of those services. Emacs allows the user to write proprietary > > code, and there's no legal issues when the user does that. Emacs also > > allows the user to copy someone else's code without permission, and > > that's not a problem for Emacs when the user does that. > > If you don't wish to correspond, don't, you are free. > > I have never said nor implied "you work for me" and I cannot see how > is that relevant to the question. You consistently take the stance that implies, and many times explicitly states, that (a) you represent the views of the GNU project, and (b) the GNU project should or should not do this and that. Then, when people like me object, you demand that they prove something to you, or else. But no one here is under any obligation of proving anything to you, and your views and opinions (which are quite radical, I must say) are your own and no one else's. They are your own responsibility, and if you want them to be proven or dis-proven, you should do that yourself. > If you participate in discussion and respond to my question relating > to licensing compliance, then provide a reference justifying its > legality. Or simply say you don't have such. Your employment is not > subject of my question nor relevant. I could ask you to do the same. You never provided any reference justifying the legality, just a lot of blogs that spread FUD (whose motivation, which many times is struggle against Free Software, I described in my previous message). If you demand something of your correspondents, please live up to the same high standards, or stop demanding that others do. Quoting a random selection of blog postings is NOT research and does NOT justify anything, except that the issue is being "discussed" by some people. It doesn't even mean that those discussions are serious, let alone that whoever posts those opinions doesn't have an agenda. > I am not user of proprietary software and I don't consider options of > writing proprietary software. Neither I am participating in discussion > to foster ideas of creation of proprietary software. > > I am free software user and for that specific case I am interested how > the licensing issue is solved. You are free to do whatever you like in your work; that is your prerogative and no one else's. But here we discuss what the Emacs project should or should not do about this technology, not your private decisions. > 2. OpenAI_RFC-84-FR-58141.pdf > https://www.uspto.gov/sites/default/files/documents/OpenAI_RFC-84-FR-58141.pdf > > Conclusions are: > > - legal justifications exists for US jurisdiction as the companies > providing the AI are strong enough to find their ways, they are > playing on the card as given in references above; as somebody > already said, I doubt they would use "fair use" doctrine if the AI > would be trained on proprietary software such as Windows; > > - conflict is serious and it is out there among the people and remains > unsolved; AI has been trained on GPL and other free software and is > used by corporations to generate new code without attributions; > people complain that it is misuse of intentions of authors; > > - overall international legal situation is thus unclear, especially > considering that free software spans the whole world, not just the > US jurisdiction, as what may work within US is not same among all > jurisdictions; That's not what the above document concludes. Quote: Conclusion We submit that: I. Under current law, training AI systems constitutes fair use. II. Policy considerations underlying fair use doctrine support the finding that training AI systems constitute fair use. III. Nevertheless, legal uncertainty on the copyright implications of training AI systems imposes substantial costs on AI developers and so should be authoritatively resolved. > Conclusion as of 2021-07-24 is that authors cannot be sure as there > are legal uncertainties. Those are your personal conclusions. They don't follow, and sometimes directly contradict, the references you yourself posted (sometimes so much so that I wonder whether we really read the same text). My opinions differ substantially, for the reasons I explained above and in other messages. (Full disclosure: part of my daytime job is development of sophisticated AI-based algorithms that use machine learning technologies for various practical purposes, including analysis of "natural language" text.) ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 7:32 ` Eli Zaretskii @ 2021-07-24 7:54 ` Jean Louis 2021-07-24 8:50 ` Eli Zaretskii 0 siblings, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-24 7:54 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-24 10:33]: > > I have never said nor implied "you work for me" and I cannot see how > > is that relevant to the question. > > You consistently take the stance that implies, and many times > explicitly states, that (a) you represent the views of the GNU > project I have never in my life said so, you please stop with it. > and (b) the GNU project should or should not do this and that. I have always beeing following GNU project as such because I agree to its principles and guidances as published on GNU website. I will definitely compare issues at hand with the already well known GNU guidelines just as you and other people do. For example when there is recommendation of proprietary software I will say that GNU project does not endorse such. > Then, when people like me object, you demand that they prove > something to you, or else. The issues are totally unrelated. I am sorry for your misunderstandings. I was interested to find out how is legality solved when re-using the code generated by AI. Nothing else beyond that. And I have found references and made conclusions. > > If you participate in discussion and respond to my question relating > > to licensing compliance, then provide a reference justifying its > > legality. Or simply say you don't have such. Your employment is not > > subject of my question nor relevant. > > I could ask you to do the same. You never provided any reference > justifying the legality, just a lot of blogs that spread FUD (whose > motivation, which many times is struggle against Free Software, I > described in my previous message). Take it easy. I asked simple question, refined the question, and found references, made conclusions of legality justification in US jurisdiction and that there are unsolved global problem. There is no need to expand discussion into directions which is purely irrelevant to the question about licensing compliance. > > 2. OpenAI_RFC-84-FR-58141.pdf > > https://www.uspto.gov/sites/default/files/documents/OpenAI_RFC-84-FR-58141.pdf > > > > Conclusions are: > > > > - legal justifications exists for US jurisdiction as the companies > > providing the AI are strong enough to find their ways, they are > > playing on the card as given in references above; as somebody > > already said, I doubt they would use "fair use" doctrine if the AI > > would be trained on proprietary software such as Windows; > > > > - conflict is serious and it is out there among the people and remains > > unsolved; AI has been trained on GPL and other free software and is > > used by corporations to generate new code without attributions; > > people complain that it is misuse of intentions of authors; > > > > - overall international legal situation is thus unclear, especially > > considering that free software spans the whole world, not just the > > US jurisdiction, as what may work within US is not same among all > > jurisdictions; > > That's not what the above document concludes. Quote: My conclusions are not what document conclude. I never said so. The document is related exclusively to US jurisdiction and its final determination is vague. It is their proposal, and not a court decision. They have openly said that they are financially strong enough and will try to defend any cases by using "fair use" doctrine in the US. This however, does not solve all jurisdictions. "fair use" doctrine is also not finally solved in the US. > Conclusion > > We submit that: > > I. Under current law, training AI systems constitutes fair use. That is their opinion, as that is a corporation that has the strength to submit such document and of course that they found some legal defense. If anybody starts complaining it will be a court case that will give final judgments. > II. Policy considerations underlying fair use doctrine support the > finding that training AI systems constitute fair use. > > III. Nevertheless, legal uncertainty on the copyright implications > of training AI systems imposes substantial costs on AI developers > and so should be authoritatively resolved. Everything from that document relates to US jurisdiction only. It is one-sided thus biased document, clearly opposing the views of many GPL authors. It is the corporations defense argument for ripping off the GPL software. They found the way and wish to play on that card. It is document of conflict, not document of friendship or collaboration. It is document of one-sidede defense, not a document that contributes to free software. It is document that defend proprietary software, not document that fosters free software. Thus it does not resolve anything in the community. It serves to one party only. Would that corporation release all software as free software that would bring or make a new leap forward. They are making one big step backwards. One can see that by number of aware free software developers canceling the Github accounts. > > Conclusion as of 2021-07-24 is that authors cannot be sure as there > > are legal uncertainties. > > Those are your personal conclusions. Personal definitely, but not as the only one with the same opinion, which should be clear from references which left probably unread. Legality of free software on the planet was ensured by the GPL license. Maybe the license was never planned to be international, but it does function well internationally. Legality of AI generated code and "free use" doctrine in the US is at this point of development yet far from functioning well internationally. > (Full disclosure: part of my daytime job is development of > sophisticated AI-based algorithms that use machine learning > technologies for various practical purposes, including analysis of > "natural language" text.) That is great, it is technical part of software. Keep doing. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 7:54 ` Jean Louis @ 2021-07-24 8:50 ` Eli Zaretskii 2021-07-24 16:16 ` Jean Louis 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 8:50 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, emacs-tangents, mullikine, rms > Date: Sat, 24 Jul 2021 10:54:07 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: mullikine@gmail.com, stefan@marxist.se, emacs-tangents@gnu.org, > rms@gnu.org > > * Eli Zaretskii <eliz@gnu.org> [2021-07-24 10:33]: > > > I have never said nor implied "you work for me" and I cannot see how > > > is that relevant to the question. > > > > You consistently take the stance that implies, and many times > > explicitly states, that (a) you represent the views of the GNU > > project > > I have never in my life said so, you please stop with it. Please re-read your postings. They say otherwise. I realize that you didn't intend that, but that's how your words sound, here and elsewhere. If you want to avoid such interpretation, please take great care to tone down your categorical statements, use qualifiers like IMO and AFAIK and "I think", and generally make sure your words say that's your opinion, not an absolute truth, let alone something the GNU project decided to do or is doing. > For example when there is recommendation of proprietary software I > will say that GNU project does not endorse such. Please consider adding "AFAIK" or somesuch, otherwise this sounds like you are speaking for the project. > I was interested to find out how is legality solved when re-using the > code generated by AI. Nothing else beyond that. And I have found > references and made conclusions. And I challenged your conclusions. Nothing beyond that. > > That's not what the above document concludes. Quote: > > My conclusions are not what document conclude. I never said so. You didn't say otherwise, either. Someone who didn't have time to read the document could think that it's what the document concluded. That's why I posted the actual quotation, so that people could draw their own conclusions, or decide they do want to read the document itself. > The document is related exclusively to US jurisdiction and its final > determination is vague. It is their proposal, and not a court > decision. They have openly said that they are financially strong > enough and will try to defend any cases by using "fair use" doctrine > in the US. All true, but the same can be said about all the other posts and blogs you quoted. Which leaves us none the wiser about the problem. We still need to assess the original facts and data and make our own conclusions. I did, and my conclusions are starkly different from yours. > Everything from that document relates to US jurisdiction only. It is > one-sided thus biased document, clearly opposing the views of many GPL > authors. > > It is the corporations defense argument for ripping off the GPL > software. They found the way and wish to play on that card. > > It is document of conflict, not document of friendship or > collaboration. It is document of one-sidede defense, not a document > that contributes to free software. > > It is document that defend proprietary software, not document that > fosters free software. > > Thus it does not resolve anything in the community. It serves to one > party only. And yet you draw conclusions from it about how GNU and Emacs should behave about this technology? How does that make sense? > > > Conclusion as of 2021-07-24 is that authors cannot be sure as there > > > are legal uncertainties. > > > > Those are your personal conclusions. > > Personal definitely, but not as the only one with the same opinion, How does that make any difference? Should GNU and Emacs take the fact that several people expressed this opinion as meaning it is the truth for our purposes? Especially since some (perhaps many) of them are driven by motivation that is explicitly anti-Free Software? > which should be clear from references which left probably unread. Please don't assume I didn't read those references. You have no basis for making such nasty assumptions about me, let alone expressing them publicly here. > Legality of free software on the planet was ensured by the GPL > license. Maybe the license was never planned to be international, but > it does function well internationally. > > Legality of AI generated code and "free use" doctrine in the US is at > this point of development yet far from functioning well > internationally. Which, to me, says that we should carefully examine this issue by ourselves, not draw any premature conclusions from the hoop-la out there. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 8:50 ` Eli Zaretskii @ 2021-07-24 16:16 ` Jean Louis 2021-07-24 16:44 ` Eli Zaretskii 0 siblings, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-24 16:16 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-24 11:51]: > Please re-read your postings. They say otherwise. I realize that you > didn't intend that, but that's how your words sound, here and > elsewhere. If you want to avoid such interpretation, please take > great care to tone down your categorical statements, use qualifiers > like IMO and AFAIK and "I think", and generally make sure your words > say that's your opinion, not an absolute truth, let alone something > the GNU project decided to do or is doing. > > > For example when there is recommendation of proprietary software I > > will say that GNU project does not endorse such. > Please consider adding "AFAIK" or somesuch, otherwise this sounds like > you are speaking for the project. Thanks, but no, all opinions are private. What is official for GNU project is on the GNU website, in the GNU manuals and other documentation. Though, that is not the subject of this thread. You have said your opinion though you did not mention not even one legal reference to the questions about licensing compliance that I have mentioned. So I keep it as your differing opinion though without references to legalities I do not find it relevant. Julia Reda's opinion I find relevant, and I found it online. I wish I could find it as answer on this mailing list sooner than online, but it is not so. > And yet you draw conclusions from it about how GNU and Emacs should > behave about this technology? How does that make sense? I never said so, that is misunderstanding. Once again, my question related GPL licensing compliance is relevant to GNU project, and to GPL licensed software authors. Licensing is legality. It is not related to technological parts. I have never mentioned technology and how GNU and Emacs should behave. In the thread of Pen.el subject I wanted to find out how is compliance to licenses solved. Don't make fuss about the simple questions. Maybe is better to wait and let maybe somebody else jump in and answer it. You seem to personally chase me that I stop asking questions? It does not really seem welcoming, it seems like I did something bad to you and you are pushing with force to stop me asking such simple banal question. We talk about GPL licensing compliance for years in various GNU related discussion within GNU project and without GNU project. I was asking German companies about licensing compliance to GNU GPL software and had such a nice conversation with them and they agreed to comply to it, and provided sources. You please make it easy, as I am asking logical question, don't call me radical as that has negative connotations. > > Legality of AI generated code and "free use" doctrine in the US is at > > this point of development yet far from functioning well > > internationally. > > Which, to me, says that we should carefully examine this issue by > ourselves, not draw any premature conclusions from the hoop-la out > there. Remember that it was me who first responded to original poster and installed pen.el and tried to run it, at that time I did not have the OpenAI key, but now I have it. From that, it should be obvious that I am interested in the technology. Without even looking online (due to my limited Internet) I have asked about licensing compliance, there was no answer until I found it today from online sources. That question is related to adopting the technology, not to rejecting it. If you wish to adopt anything into private use one has to have permissions, or in this case "fair use" exemption granted by US government. It should be obvious that I have referenced legal advisors, attorneys who made that document, including Julia Reda, known as activist in Germany, and it was me who found references and listed it here. Beside those really deficient expressions, if you have a constructive references on how how each jurisdiction would accept "fair use" let me know, otherwise leave this discussion in peace and myself. Stay on subject, don't call me words as me and you didn't graze the sheep together. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 16:16 ` Jean Louis @ 2021-07-24 16:44 ` Eli Zaretskii 2021-07-24 18:01 ` Jean Louis 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 16:44 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, emacs-tangents, mullikine, rms > Date: Sat, 24 Jul 2021 19:16:34 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: mullikine@gmail.com, stefan@marxist.se, emacs-tangents@gnu.org, > rms@gnu.org > > > > For example when there is recommendation of proprietary software I > > > will say that GNU project does not endorse such. > > > Please consider adding "AFAIK" or somesuch, otherwise this sounds like > > you are speaking for the project. > > Thanks, but no, all opinions are private. Again, yours don't sound private, and many people will become confused at best. > You have said your opinion though you did not mention not even one > legal reference to the questions about licensing compliance that I > have mentioned. So I keep it as your differing opinion though without > references to legalities I do not find it relevant. This is a public list. It is important for me to state my differing opinions so that people could make up their own minds. > You seem to personally chase me that I stop asking questions? No. I'm trying to correct the wrong impression your postings could have made on people reading them. > It does not really seem welcoming, it seems like I did something bad > to you and you are pushing with force to stop me asking such simple > banal question. I said exactly what I think was wrong with your postings. I wish you would change the style and wording when you speak on these issues, but I have no real control on what you will do. > You please make it easy, as I am asking logical question, don't call > me radical as that has negative connotations. I explained in detail why I said that. I'm sorry to conclude that you disregard that. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 16:44 ` Eli Zaretskii @ 2021-07-24 18:01 ` Jean Louis 0 siblings, 0 replies; 46+ messages in thread From: Jean Louis @ 2021-07-24 18:01 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms Eli, my question is solved, there was subject "Pen.el" and I found what is the legal justification from third parties. Thus my question is solved. I would like to try Pen.el but I have here technical problems. Thank for suggestions, I will think about it. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 3:07 ` Jean Louis 2021-07-24 7:32 ` Eli Zaretskii @ 2021-07-25 1:09 ` Richard Stallman 1 sibling, 0 replies; 46+ messages in thread From: Richard Stallman @ 2021-07-25 1:09 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, eliz, mullikine, emacs-tangents [[[ To any NSA and FBI agents reading my email: please consider ]]] [[[ whether defending the US Constitution against all enemies, ]]] [[[ foreign or domestic, requires you to follow Snowden's example. ]]] > And fact that it is "rare" does not make it a less problem for > copyright purposes as the new author cannot know which part of the > code has used "rare" verbatim. I think that is correct. Falling into this pit may be unusual, but that doesn't make it painless. -- Dr Richard Stallman (https://stallman.org) Chief GNUisance of the GNU Project (https://gnu.org) Founder, Free Software Foundation (https://fsf.org) Internet Hall-of-Famer (https://internethalloffame.org) ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 11:32 ` Jean Louis 2021-07-23 11:51 ` Eli Zaretskii @ 2021-07-24 1:14 ` Richard Stallman 2021-07-24 2:10 ` Shane Mulligan 2021-07-24 6:49 ` Eli Zaretskii 1 sibling, 2 replies; 46+ messages in thread From: Richard Stallman @ 2021-07-24 1:14 UTC (permalink / raw) To: Jean Louis; +Cc: stefan, eliz, mullikine, emacs-tangents [[[ To any NSA and FBI agents reading my email: please consider ]]] [[[ whether defending the US Constitution against all enemies, ]]] [[[ foreign or domestic, requires you to follow Snowden's example. ]]] > > That's not what happens with these services: they don't _copy_ code > > from other software (that won't work, because the probability of the > > variables being called by other names is 100%, and thus such code, if > > pasted into your program, will not compile). What they do, they > > extract ideas and algorithms from those other places, and express them > > in terms of your variables and your data types. So licenses are not > > relevant here. > According to online reviews chunks of code is copied even verbatim and > people find from where. Even if modified, it still requires licensing > compliance. From what I have read, it seems that the behavior of copilot runs on a spectrum from the first description to the second description. I expect that in many cases, nothing copyrightable has been copied, but in some cases copilot does copy a substantial amount from a copyrighted work. -- Dr Richard Stallman (https://stallman.org) Chief GNUisance of the GNU Project (https://gnu.org) Founder, Free Software Foundation (https://fsf.org) Internet Hall-of-Famer (https://internethalloffame.org) ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 1:14 ` Richard Stallman @ 2021-07-24 2:10 ` Shane Mulligan 2021-07-24 2:34 ` Shane Mulligan 2021-07-24 6:49 ` Eli Zaretskii 1 sibling, 1 reply; 46+ messages in thread From: Shane Mulligan @ 2021-07-24 2:10 UTC (permalink / raw) To: rms; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, Jean Louis [-- Attachment #1: Type: text/plain, Size: 2665 bytes --] It's a bit like whitewashing because it's reconstructing generatively by finding artificial/contrived associations between different works that the author had not intended but may have been part of their inspiration inspiration, and it compresses the information based on these assocations. It's a bit like running a lossy 'zip' on the internet and then decompressing probabilistically. When run deterministically (set the temperature of GPT to 0), you may actually see 'snippets' from various places, every time, with the same input generating the same snippets. So the source material is important. What GitHub did was very, very bad but they did it anyway. That doesn't mean GPT is bad, it just means they zipped up content they should not have and created this language 'index' or ('codex' is what they call it). What they really should do, if they are honest people, is train the model on subsets of GitHub code by separate licence and release the models with the same license. Shane Mulligan How to contact me: 🇦🇺 00 61 421 641 250 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> mullikine@gmail.com On Sat, Jul 24, 2021 at 1:14 PM Richard Stallman <rms@gnu.org> wrote: > [[[ To any NSA and FBI agents reading my email: please consider ]]] > [[[ whether defending the US Constitution against all enemies, ]]] > [[[ foreign or domestic, requires you to follow Snowden's example. ]]] > > > > That's not what happens with these services: they don't _copy_ code > > > from other software (that won't work, because the probability of the > > > variables being called by other names is 100%, and thus such code, if > > > pasted into your program, will not compile). What they do, they > > > extract ideas and algorithms from those other places, and express > them > > > in terms of your variables and your data types. So licenses are not > > > relevant here. > > > According to online reviews chunks of code is copied even verbatim and > > people find from where. Even if modified, it still requires licensing > > compliance. > > From what I have read, it seems that the behavior of copilot runs on a > spectrum from the first description to the second description. I > expect that in many cases, nothing copyrightable has been copied, but > in some cases copilot does copy a substantial amount from a > copyrighted work. > > -- > Dr Richard Stallman (https://stallman.org) > Chief GNUisance of the GNU Project (https://gnu.org) > Founder, Free Software Foundation (https://fsf.org) > Internet Hall-of-Famer (https://internethalloffame.org) > > > [-- Attachment #2: Type: text/html, Size: 4946 bytes --] ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 2:10 ` Shane Mulligan @ 2021-07-24 2:34 ` Shane Mulligan 2021-07-24 3:14 ` Shane Mulligan 0 siblings, 1 reply; 46+ messages in thread From: Shane Mulligan @ 2021-07-24 2:34 UTC (permalink / raw) To: rms; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, Jean Louis [-- Attachment #1: Type: text/plain, Size: 3939 bytes --] This is why the technology is a bit like a personal Google search, Stackoverflow, which you can store offline because it's an index of the internet that is capable of reconstruction. But it's not limited to code generation. Codex is nothing. Emacs + GPT would carve a large piece out of M$. Codex is a model trained for the purpose of generating code, but GPT models will become abundant for all tasks, including image and audio synthesis and understanding. Emacs is a complete operating system. VSCode is geared towards programming. Emacs can do infinitely more things with GPT than VSCode can because it's holistic. Even the 'eliza' in emacs can pass the turing test with GPT. GPT can run sequences of commands in emacs to automate entire workflows with natural language. But the future is in collaborative GPT. The basis/base truth would become versions of LMs or ontologies. Right now that's EleutherAI. Shane Mulligan How to contact me: 🇦🇺 00 61 421 641 250 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> mullikine@gmail.com On Sat, Jul 24, 2021 at 2:10 PM Shane Mulligan <mullikine@gmail.com> wrote: > It's a bit like whitewashing because it's > reconstructing generatively by finding > artificial/contrived associations between > different works that the author had not > intended but may have been part of their > inspiration inspiration, and it compresses the > information based on these assocations. > > It's a bit like running a lossy 'zip' on the > internet and then decompressing > probabilistically. > > When run deterministically (set the temperature of GPT to 0), you may > actually > see 'snippets' from various places, every time, with the same input > generating > the same snippets. > > So the source material is important. > > What GitHub did was very, very bad but they > did it anyway. > > That doesn't mean GPT is bad, it just means > they zipped up content they should not have > and created this language 'index' or ('codex' > is what they call it). > > What they really should do, if they are honest > people, is train the model on subsets of > GitHub code by separate licence and release > the models with the same license. > > Shane Mulligan > > How to contact me: > 🇦🇺 00 61 421 641 250 > 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> > mullikine@gmail.com > > > On Sat, Jul 24, 2021 at 1:14 PM Richard Stallman <rms@gnu.org> wrote: > >> [[[ To any NSA and FBI agents reading my email: please consider ]]] >> [[[ whether defending the US Constitution against all enemies, ]]] >> [[[ foreign or domestic, requires you to follow Snowden's example. ]]] >> >> > > That's not what happens with these services: they don't _copy_ code >> > > from other software (that won't work, because the probability of the >> > > variables being called by other names is 100%, and thus such code, >> if >> > > pasted into your program, will not compile). What they do, they >> > > extract ideas and algorithms from those other places, and express >> them >> > > in terms of your variables and your data types. So licenses are not >> > > relevant here. >> >> > According to online reviews chunks of code is copied even verbatim and >> > people find from where. Even if modified, it still requires licensing >> > compliance. >> >> From what I have read, it seems that the behavior of copilot runs on a >> spectrum from the first description to the second description. I >> expect that in many cases, nothing copyrightable has been copied, but >> in some cases copilot does copy a substantial amount from a >> copyrighted work. >> >> -- >> Dr Richard Stallman (https://stallman.org) >> Chief GNUisance of the GNU Project (https://gnu.org) >> Founder, Free Software Foundation (https://fsf.org) >> Internet Hall-of-Famer (https://internethalloffame.org) >> >> >> [-- Attachment #2: Type: text/html, Size: 7947 bytes --] ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 2:34 ` Shane Mulligan @ 2021-07-24 3:14 ` Shane Mulligan 0 siblings, 0 replies; 46+ messages in thread From: Shane Mulligan @ 2021-07-24 3:14 UTC (permalink / raw) To: rms; +Cc: Eli Zaretskii, Stefan Kangas, emacs-tangents, Jean Louis [-- Attachment #1: Type: text/plain, Size: 4856 bytes --] Proprietary code from within the M$ ecosystem is uninspired and bad code by comparison. Open source code is the gold mine so M$ will not like being told they cannot use open source to compile codex. It's a complete r*pe of open source. GPT is trained on public language and language belongs to people generally, not some select group. It's not meant to be a tool for controlling people. GPT is literally the soul of a billion people and should be public domain and not feared by GNU but instead rescued. Sorry for the rhetoric! On Sat, Jul 24, 2021 at 2:34 PM Shane Mulligan <mullikine@gmail.com> wrote: > This is why the technology is a bit like a > personal Google search, Stackoverflow, which > you can store offline because it's an index of the internet that is > capable of reconstruction. > > But it's not limited to code generation. Codex > is nothing. Emacs + GPT would carve a large > piece out of M$. > > Codex is a model trained for the purpose of > generating code, but GPT models will become > abundant for all tasks, including image and > audio synthesis and understanding. > > Emacs is a complete operating system. > VSCode is geared towards programming. > > Emacs can do infinitely more things with GPT > than VSCode can because it's holistic. > > Even the 'eliza' in emacs can pass the turing > test with GPT. GPT can run sequences of commands in emacs to automate > entire workflows with natural language. > > But the future is in collaborative GPT. > > The basis/base truth would become versions of > LMs or ontologies. > > Right now that's EleutherAI. > > Shane Mulligan > > How to contact me: > 🇦🇺 00 61 421 641 250 > 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> > mullikine@gmail.com > > > On Sat, Jul 24, 2021 at 2:10 PM Shane Mulligan <mullikine@gmail.com> > wrote: > >> It's a bit like whitewashing because it's >> reconstructing generatively by finding >> artificial/contrived associations between >> different works that the author had not >> intended but may have been part of their >> inspiration inspiration, and it compresses the >> information based on these assocations. >> >> It's a bit like running a lossy 'zip' on the >> internet and then decompressing >> probabilistically. >> >> When run deterministically (set the temperature of GPT to 0), you may >> actually >> see 'snippets' from various places, every time, with the same input >> generating >> the same snippets. >> >> So the source material is important. >> >> What GitHub did was very, very bad but they >> did it anyway. >> >> That doesn't mean GPT is bad, it just means >> they zipped up content they should not have >> and created this language 'index' or ('codex' >> is what they call it). >> >> What they really should do, if they are honest >> people, is train the model on subsets of >> GitHub code by separate licence and release >> the models with the same license. >> >> Shane Mulligan >> >> How to contact me: >> 🇦🇺 00 61 421 641 250 >> 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> >> mullikine@gmail.com >> >> >> On Sat, Jul 24, 2021 at 1:14 PM Richard Stallman <rms@gnu.org> wrote: >> >>> [[[ To any NSA and FBI agents reading my email: please consider ]]] >>> [[[ whether defending the US Constitution against all enemies, ]]] >>> [[[ foreign or domestic, requires you to follow Snowden's example. ]]] >>> >>> > > That's not what happens with these services: they don't _copy_ code >>> > > from other software (that won't work, because the probability of >>> the >>> > > variables being called by other names is 100%, and thus such code, >>> if >>> > > pasted into your program, will not compile). What they do, they >>> > > extract ideas and algorithms from those other places, and express >>> them >>> > > in terms of your variables and your data types. So licenses are >>> not >>> > > relevant here. >>> >>> > According to online reviews chunks of code is copied even verbatim >>> and >>> > people find from where. Even if modified, it still requires licensing >>> > compliance. >>> >>> From what I have read, it seems that the behavior of copilot runs on a >>> spectrum from the first description to the second description. I >>> expect that in many cases, nothing copyrightable has been copied, but >>> in some cases copilot does copy a substantial amount from a >>> copyrighted work. >>> >>> -- >>> Dr Richard Stallman (https://stallman.org) >>> Chief GNUisance of the GNU Project (https://gnu.org) >>> Founder, Free Software Foundation (https://fsf.org) >>> Internet Hall-of-Famer (https://internethalloffame.org) >>> >>> >>> -- Shane Mulligan How to contact me: 🇦🇺 00 61 421 641 250 🇳🇿 00 64 21 1462 759 <+64-21-1462-759> mullikine@gmail.com [-- Attachment #2: Type: text/html, Size: 10603 bytes --] ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 1:14 ` Richard Stallman 2021-07-24 2:10 ` Shane Mulligan @ 2021-07-24 6:49 ` Eli Zaretskii 2021-07-24 7:33 ` Jean Louis 2021-07-24 7:41 ` Philip Kaludercic 1 sibling, 2 replies; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 6:49 UTC (permalink / raw) To: rms; +Cc: mullikine, emacs-tangents, stefan, bugs > From: Richard Stallman <rms@gnu.org> > Date: Fri, 23 Jul 2021 21:14:23 -0400 > Cc: stefan@marxist.se, eliz@gnu.org, mullikine@gmail.com, > emacs-tangents@gnu.org > > > > That's not what happens with these services: they don't _copy_ code > > > from other software (that won't work, because the probability of the > > > variables being called by other names is 100%, and thus such code, if > > > pasted into your program, will not compile). What they do, they > > > extract ideas and algorithms from those other places, and express them > > > in terms of your variables and your data types. So licenses are not > > > relevant here. > > > According to online reviews chunks of code is copied even verbatim and > > people find from where. Even if modified, it still requires licensing > > compliance. > > From what I have read, it seems that the behavior of copilot runs on a > spectrum from the first description to the second description. I > expect that in many cases, nothing copyrightable has been copied, but > in some cases copilot does copy a substantial amount from a > copyrighted work. It cannot be a verbatim copy, because at least the variables, and sometimes also the data types, need to be renamed. Whether the result is still under the original copyright cannot be established without actually comparing the two versions of the code. So any general flat rejection of the idea of these services on these grounds is not serious, IMO. Of course, someone like Jean will not use any code until a bunch of lawyers submit an official opinion about the legal implications, but IMO that's a radical view that doesn't make a lot of sense, especially since none of the code accessible openly via the net can be proprietary, for obvious reasons. Jean could do whatever he personally likes, but his radical views don't necessarily bind the GNU project in general and Emacs in particular. Moreover, ironically Jean bases his views on opinions and issues expressed by clear opponents of Free Software. The strongest drive behind many of these blogs' aversion from these services is the fear that GPL-licensed code creeps into proprietary software produced by enterprises and their software subcontractors, because that would require them to make the sources available or at least put them at a risk of lawsuits. It is a well-known fact that most, if not all, software contracts for proprietary software nowadays include explicit prohibition of using GPL-licensed code in the product. It is those people that serve these contracts and enterprises who drive the whoop-la about licensing issues in code offered by these AI-based services. So before embracing their FUD and biased opinions, I really suggest to actually look at the code, compare it with the original, and make an independent assessment of both whether it's a "copy" from the copyright POV and of the licenses of the original code. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 6:49 ` Eli Zaretskii @ 2021-07-24 7:33 ` Jean Louis 2021-07-24 8:10 ` Eli Zaretskii 2021-07-24 7:41 ` Philip Kaludercic 1 sibling, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-24 7:33 UTC (permalink / raw) To: Eli Zaretskii; +Cc: mullikine, emacs-tangents, stefan, rms Eli, I do take care of licensing when re-using somebody's software, and when publishing software or distributing it. There is nothing "radical" about it. Concerns of other people are also not radical. Intention of authors is not respected even if there is legal circumvention in the US such as "fair use", that does not fly in other jurisdictions. I do understand you have some unsolved issues or something you cannot handle related to licensing as you are more for technical side, but please don't call it "radical" as that does not teach people about GPL licensing. Jean ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 7:33 ` Jean Louis @ 2021-07-24 8:10 ` Eli Zaretskii 2021-07-24 8:21 ` Jean Louis 2021-07-24 8:35 ` Jean Louis 0 siblings, 2 replies; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 8:10 UTC (permalink / raw) To: Jean Louis; +Cc: mullikine, emacs-tangents, stefan, rms > Date: Sat, 24 Jul 2021 10:33:57 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: rms@gnu.org, stefan@marxist.se, mullikine@gmail.com, > emacs-tangents@gnu.org > > > Eli, I do take care of licensing when re-using somebody's software, > and when publishing software or distributing it. > > There is nothing "radical" about it. Considering the license of the code is not radical, indeed. But the criteria you personally apply when considering that _are_ radical. You posted enough opinions about these matters to make that abundantly clear. There's nothing wrong with having such views, they are your personal views, and are entirely legitimate. All I'm saying is that the Emacs project should not be guided by such views, for the reasons I explained. > Concerns of other people are also not radical. No, but your interpretation of those "concerns" is. > Intention of authors is not respected even if there is legal > circumvention in the US such as "fair use", that does not fly in > other jurisdictions. So you agree that the problems you raised don't seem to exist at least in the US? > I do understand you have some unsolved issues or something you cannot > handle related to licensing No, I don't have any unsolved issues. > as you are more for technical side ??? What is that supposed to mean? > but please don't call it "radical" as that does not teach people > about GPL licensing. When I see a radical view, I call it "radical". Promoting Free Software requires healthy pragmatism, because we want the Free Software to flourish and remain relevant by picking up the advances in technology. Rejecting such new technologies just because there's some doubts expressed by someone in some blog is "radical", and IMO eventually detrimental to Free Software development. We should instead carefully and independently assess the issues and make our own judgment based on specific details of each such development. We cannot run away of every idea because some people say it might cause trouble in some cases. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 8:10 ` Eli Zaretskii @ 2021-07-24 8:21 ` Jean Louis 2021-07-24 8:35 ` Jean Louis 1 sibling, 0 replies; 46+ messages in thread From: Jean Louis @ 2021-07-24 8:21 UTC (permalink / raw) To: Eli Zaretskii; +Cc: mullikine, emacs-tangents, stefan, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-24 11:11]: > There's nothing wrong with having such views, they are your personal > views, and are entirely legitimate. All I'm saying is that the Emacs > project should not be guided by such views, for the reasons I > explained. My question has been resolved. It is over. Did you read it? ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 8:10 ` Eli Zaretskii 2021-07-24 8:21 ` Jean Louis @ 2021-07-24 8:35 ` Jean Louis 2021-07-24 8:59 ` Eli Zaretskii 1 sibling, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-24 8:35 UTC (permalink / raw) To: Eli Zaretskii; +Cc: mullikine, emacs-tangents, stefan, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-24 11:11]: > When I see a radical view, I call it "radical". Promoting Free > Software requires healthy pragmatism, because we want the Free > Software to flourish and remain relevant by picking up the advances in > technology. Rejecting such new technologies just because there's some > doubts expressed by someone in some blog is "radical", and IMO > eventually detrimental to Free Software development. We should > instead carefully and independently assess the issues and make our own > judgment based on specific details of each such development. We > cannot run away of every idea because some people say it might cause > trouble in some cases. I am totally for advances of technology as long as we foster free software and freedom in computing. We have too little of AI today in 21st century. Question is definitely not so general how presented in your paragraph above. It is very specific, related on how to solve licensing issues. There are no doubts that code may be copied verbatin, as here is authorized and official documentation by Github related to Copilot: https://docs.github.com/en/github/copilot/research-recitation It is not related to various other AIs, etc. I am not sure if the same AI is even used in Pen.el. It may not be relevant. In the Copilot documentation it says: Quote: This investigation demonstrates that GitHub Copilot can quote a body of code verbatim, but that it rarely does so, and when it does, it mostly quotes code that everybody quotes, and mostly at the beginning of a file, as if to break the ice. Additionally I have been using OpenAI and found not 0.1 percent verbatim responses, I could find those pages on Internet from where verbatim paragraphs were cited. I am still in playground. I can find paragraphs from websites from our competitors as a response. I still have to discover I in the AI in the playground of OpenAI service. Licensing issues I have made and for which I have found partial solution are in no way related to rejecting, rather to adopting it in free software. My question was how we can adopt the code generated into free software (for example by using Pen.el) as it generates code by using other GPL free software without attributions. Partially it is resolved in the US, though unproven and with great conflict with authors. It does not give assurance. I am not sure if I can generate the code and that it is really "original" and infringement free. Is the OpenAI company giving me some kind of guarantee that I will be held without liabilities if I use that code? Thus those issues may be temporarily brushed off with "fair use" in US, they remain unsolved in the US until the first few court cases or class action suite, and are not resolved on international level at all. Julia Reda's statement does not apply in all jurisdictions. At this moment there is no verified legal statement by let us say FSF attorneys or legal experts or some other organization that will confirm legal status of such generated code or text on international level. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 8:35 ` Jean Louis @ 2021-07-24 8:59 ` Eli Zaretskii 2021-07-24 16:18 ` Jean Louis 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 8:59 UTC (permalink / raw) To: Jean Louis; +Cc: mullikine, emacs-tangents, stefan, rms > Date: Sat, 24 Jul 2021 11:35:41 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: rms@gnu.org, stefan@marxist.se, mullikine@gmail.com, > emacs-tangents@gnu.org > > There are no doubts that code may be copied verbatin, as here is > authorized and official documentation by Github related to Copilot: > https://docs.github.com/en/github/copilot/research-recitation So we won't use that Github stuff. Which we won't anyway, because we avoid Github in general. How does this help us decide about general usability of this technology? It doesn't. > At this moment there is no verified legal statement by let us say FSF > attorneys or legal experts or some other organization that will > confirm legal status of such generated code or text on international > level. I suggest to leave the legal stuff to the legal experts. We should assess the data and provide them with facts, not make the decisions for them. Which means this discussion, and your suggestions that we should already stay away of this technology, are premature at best, if not in the wrong place at the wrong time. So please don't discourage independent assessment of this technology by posting half-baked "legal" opinions from people with questionable motivation (present company excluded, of course) representing that this technology is legally incompatible with Free Software. IMO, we don't yet know enough for any such definitive opinions and conclusions. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 8:59 ` Eli Zaretskii @ 2021-07-24 16:18 ` Jean Louis 2021-07-24 16:45 ` Eli Zaretskii 0 siblings, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-24 16:18 UTC (permalink / raw) To: Eli Zaretskii; +Cc: mullikine, emacs-tangents, stefan, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-24 12:00]: > So please don't discourage independent assessment of this technology > by posting half-baked "legal" opinions from people with questionable > motivation (present company excluded, of course) representing that > this technology is legally incompatible with Free Software. IMO, we > don't yet know enough for any such definitive opinions and > conclusions. Quite contrary, the GPL licensing compliance question is related to adoption and expansion of free software. I have never stated it is incompatible with free software, I have asked how is GPL licensing compliance solved. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 16:18 ` Jean Louis @ 2021-07-24 16:45 ` Eli Zaretskii 2021-07-24 17:57 ` Jean Louis 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 16:45 UTC (permalink / raw) To: Jean Louis; +Cc: mullikine, emacs-tangents, stefan, rms > Date: Sat, 24 Jul 2021 19:18:02 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: rms@gnu.org, stefan@marxist.se, mullikine@gmail.com, > emacs-tangents@gnu.org > > * Eli Zaretskii <eliz@gnu.org> [2021-07-24 12:00]: > > So please don't discourage independent assessment of this technology > > by posting half-baked "legal" opinions from people with questionable > > motivation (present company excluded, of course) representing that > > this technology is legally incompatible with Free Software. IMO, we > > don't yet know enough for any such definitive opinions and > > conclusions. > > Quite contrary, the GPL licensing compliance question is related to > adoption and expansion of free software. > > I have never stated it is incompatible with free software, I have > asked how is GPL licensing compliance solved. No, you said we shouldn't use this. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 16:45 ` Eli Zaretskii @ 2021-07-24 17:57 ` Jean Louis 2021-07-24 18:15 ` Eli Zaretskii 0 siblings, 1 reply; 46+ messages in thread From: Jean Louis @ 2021-07-24 17:57 UTC (permalink / raw) To: Eli Zaretskii; +Cc: mullikine, emacs-tangents, stefan, rms * Eli Zaretskii <eliz@gnu.org> [2021-07-24 19:46]: > > I have never stated it is incompatible with free software, I have > > asked how is GPL licensing compliance solved. > > No, you said we shouldn't use this. Sorry for misunderstandings. I don't remember I ever said we shouldn't use this. It does not appear to me logical as I remember my intention was to find out how is licensing compliance solved so that it becomes clear how it works. Question was directed to author of Pen.el and there was no clear answer neither from you, so I found myself from online research that at least in US for now it is based on "fair use" doctrine. When we take the word "fair" in its original definition, it should be obvious from online comments that many GPL authors do not really find it "fair". It is however one defense that all of present similar AI models have in common, they are to use "fair use" doctrine. We will see that. Also to mention, AI as such is not related to this particular case of using GPL and other free software, it is just one small application of overall artificial technology, there are many other applications which are totally out of this context. Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 17:57 ` Jean Louis @ 2021-07-24 18:15 ` Eli Zaretskii 0 siblings, 0 replies; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 18:15 UTC (permalink / raw) To: Jean Louis; +Cc: mullikine, emacs-tangents, stefan, rms > Date: Sat, 24 Jul 2021 20:57:42 +0300 > From: Jean Louis <bugs@gnu.support> > Cc: rms@gnu.org, stefan@marxist.se, mullikine@gmail.com, > emacs-tangents@gnu.org > > * Eli Zaretskii <eliz@gnu.org> [2021-07-24 19:46]: > > > > I have never stated it is incompatible with free software, I have > > > asked how is GPL licensing compliance solved. > > > > No, you said we shouldn't use this. > > Sorry for misunderstandings. To avoid such misunderstandings, I suggest to tone down your language when you are talking about licensing issues associated with some technologies or products, so that what you write couldn't be interpreted as saying that there are legal problems which prevent our use of those technologies and products. > Question was directed to author of Pen.el and there was no clear > answer neither from you, so I found myself from online research that > at least in US for now it is based on "fair use" doctrine. > > When we take the word "fair" in its original definition, it should be > obvious from online comments that many GPL authors do not really find > it "fair". It is however one defense that all of present similar AI > models have in common, they are to use "fair use" doctrine. We will > see that. That is a separate issue, which is IMO completely unrelated. Emacs is Free Software, and is distributed under GPL, so for Emacs it is OK to allow users to use other GPL code out there in their programs. That there are producers of proprietary software who use pieces of GPL code in their proprietary products without complying with GPL is completely unrelated to what the Emacs project can do with this technology. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 6:49 ` Eli Zaretskii 2021-07-24 7:33 ` Jean Louis @ 2021-07-24 7:41 ` Philip Kaludercic 2021-07-24 7:59 ` Eli Zaretskii 1 sibling, 1 reply; 46+ messages in thread From: Philip Kaludercic @ 2021-07-24 7:41 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms, bugs Eli Zaretskii <eliz@gnu.org> writes: >> From: Richard Stallman <rms@gnu.org> >> Date: Fri, 23 Jul 2021 21:14:23 -0400 >> Cc: stefan@marxist.se, eliz@gnu.org, mullikine@gmail.com, >> emacs-tangents@gnu.org >> >> > > That's not what happens with these services: they don't _copy_ code >> > > from other software (that won't work, because the probability of the >> > > variables being called by other names is 100%, and thus such code, if >> > > pasted into your program, will not compile). What they do, they >> > > extract ideas and algorithms from those other places, and express them >> > > in terms of your variables and your data types. So licenses are not >> > > relevant here. >> >> > According to online reviews chunks of code is copied even verbatim and >> > people find from where. Even if modified, it still requires licensing >> > compliance. >> >> From what I have read, it seems that the behavior of copilot runs on a >> spectrum from the first description to the second description. I >> expect that in many cases, nothing copyrightable has been copied, but >> in some cases copilot does copy a substantial amount from a >> copyrighted work. > > It cannot be a verbatim copy, because at least the variables, and > sometimes also the data types, need to be renamed. Whether the result > is still under the original copyright cannot be established without > actually comparing the two versions of the code. So any general > flat rejection of the idea of these services on these grounds is not > serious, IMO. Not necessarily, if it generates a pure, top-level function. Someone could type something like "Sort list of postcodes" and it generates a Radix Sort function. And if this is part of some code that was copied a lot, the model might tend to generate this verbatim even more likely. Or that is at least my understanding. -- Philip Kaludercic ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 7:41 ` Philip Kaludercic @ 2021-07-24 7:59 ` Eli Zaretskii 2021-07-24 9:31 ` Philip Kaludercic 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 7:59 UTC (permalink / raw) To: Philip Kaludercic; +Cc: stefan, emacs-tangents, mullikine, rms, bugs > From: Philip Kaludercic <philipk@posteo.net> > Cc: rms@gnu.org, mullikine@gmail.com, emacs-tangents@gnu.org, > stefan@marxist.se, bugs@gnu.support > Date: Sat, 24 Jul 2021 07:41:21 +0000 > > > It cannot be a verbatim copy, because at least the variables, and > > sometimes also the data types, need to be renamed. Whether the result > > is still under the original copyright cannot be established without > > actually comparing the two versions of the code. So any general > > flat rejection of the idea of these services on these grounds is not > > serious, IMO. > > Not necessarily, if it generates a pure, top-level function. Someone > could type something like "Sort list of postcodes" and it generates a > Radix Sort function. And if this is part of some code that was copied a > lot, the model might tend to generate this verbatim even more likely. A sort function must state at least the data type before it can be compiled. And if you are talking about pseudo-code that is data-type agnostic, then that's an algorithm, and is not copyrightable, AFAIK. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 7:59 ` Eli Zaretskii @ 2021-07-24 9:31 ` Philip Kaludercic 2021-07-24 11:19 ` Eli Zaretskii 0 siblings, 1 reply; 46+ messages in thread From: Philip Kaludercic @ 2021-07-24 9:31 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms, bugs Eli Zaretskii <eliz@gnu.org> writes: >> From: Philip Kaludercic <philipk@posteo.net> >> Cc: rms@gnu.org, mullikine@gmail.com, emacs-tangents@gnu.org, >> stefan@marxist.se, bugs@gnu.support >> Date: Sat, 24 Jul 2021 07:41:21 +0000 >> >> > It cannot be a verbatim copy, because at least the variables, and >> > sometimes also the data types, need to be renamed. Whether the result >> > is still under the original copyright cannot be established without >> > actually comparing the two versions of the code. So any general >> > flat rejection of the idea of these services on these grounds is not >> > serious, IMO. >> >> Not necessarily, if it generates a pure, top-level function. Someone >> could type something like "Sort list of postcodes" and it generates a >> Radix Sort function. And if this is part of some code that was copied a >> lot, the model might tend to generate this verbatim even more likely. > > A sort function must state at least the data type before it can be > compiled. And if you are talking about pseudo-code that is data-type > agnostic, then that's an algorithm, and is not copyrightable, AFAIK. No, I was thinking about concrete code, that depending on the language might even just rely on the standard library, especially if the language has generics. Seeing how often SO code has been found in random repositories[0], I don't think it is improbable that the trained models might notice these patterns. [0] For example https://programming.guide/worlds-most-copied-so-snippet.html -- Philip Kaludercic ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 9:31 ` Philip Kaludercic @ 2021-07-24 11:19 ` Eli Zaretskii 2021-07-24 14:16 ` Philip Kaludercic 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 11:19 UTC (permalink / raw) To: Philip Kaludercic; +Cc: stefan, emacs-tangents, mullikine, rms, bugs > From: Philip Kaludercic <philipk@posteo.net> > Cc: rms@gnu.org, mullikine@gmail.com, emacs-tangents@gnu.org, > stefan@marxist.se, bugs@gnu.support > Date: Sat, 24 Jul 2021 09:31:38 +0000 > > >> Not necessarily, if it generates a pure, top-level function. Someone > >> could type something like "Sort list of postcodes" and it generates a > >> Radix Sort function. And if this is part of some code that was copied a > >> lot, the model might tend to generate this verbatim even more likely. > > > > A sort function must state at least the data type before it can be > > compiled. And if you are talking about pseudo-code that is data-type > > agnostic, then that's an algorithm, and is not copyrightable, AFAIK. > > No, I was thinking about concrete code, that depending on the language > might even just rely on the standard library, especially if the language > has generics. Seeing how often SO code has been found in random > repositories[0], I don't think it is improbable that the trained models > might notice these patterns. Sorry, I don't understand what you have in mind. Can you show an example of useful code that could be copied verbatim into a program without at least some renaming, without breaking the program? ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 11:19 ` Eli Zaretskii @ 2021-07-24 14:16 ` Philip Kaludercic 2021-07-24 14:37 ` Eli Zaretskii 0 siblings, 1 reply; 46+ messages in thread From: Philip Kaludercic @ 2021-07-24 14:16 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms, bugs Eli Zaretskii <eliz@gnu.org> writes: >> > A sort function must state at least the data type before it can be >> > compiled. And if you are talking about pseudo-code that is data-type >> > agnostic, then that's an algorithm, and is not copyrightable, AFAIK. >> >> No, I was thinking about concrete code, that depending on the language >> might even just rely on the standard library, especially if the language >> has generics. Seeing how often SO code has been found in random >> repositories[0], I don't think it is improbable that the trained models >> might notice these patterns. > > Sorry, I don't understand what you have in mind. Can you show an > example of useful code that could be copied verbatim into a program > without at least some renaming, without breaking the program? To take the example from the article I mentioned above public static String humanReadableByteCount(long bytes, boolean si) { int unit = si ? 1000 : 1024; if (bytes < unit) return bytes + " B"; int exp = (int) (Math.log(bytes) / Math.log(unit)); String pre = (si ? "kMGTPE" : "KMGTPE").charAt(exp-1) + (si ? "" : "i"); return String.format("%.1f %sB", bytes / Math.pow(unit, exp), pre); } can be copied into a Java program, and assuming that there is no other method called humanReadableByteCount in the same class, it should compile and run without renaming or re-typing. CoPilot might generate this from a comment like, // Convert a byte count to a human-readable string since it is mentioned over 6000 times on GitHub (and this method even has a bug, as the article explains -- but that is a totally different issue). -- Philip Kaludercic ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 14:16 ` Philip Kaludercic @ 2021-07-24 14:37 ` Eli Zaretskii 2021-07-24 14:49 ` Philip Kaludercic 0 siblings, 1 reply; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 14:37 UTC (permalink / raw) To: Philip Kaludercic; +Cc: stefan, emacs-tangents, mullikine, rms, bugs > From: Philip Kaludercic <philipk@posteo.net> > Cc: rms@gnu.org, mullikine@gmail.com, emacs-tangents@gnu.org, > stefan@marxist.se, bugs@gnu.support > Date: Sat, 24 Jul 2021 14:16:55 +0000 > > > Sorry, I don't understand what you have in mind. Can you show an > > example of useful code that could be copied verbatim into a program > > without at least some renaming, without breaking the program? > > To take the example from the article I mentioned above > > public static String humanReadableByteCount(long bytes, boolean si) { > int unit = si ? 1000 : 1024; > if (bytes < unit) return bytes + " B"; > int exp = (int) (Math.log(bytes) / Math.log(unit)); > String pre = (si ? "kMGTPE" : "KMGTPE").charAt(exp-1) + (si ? "" : "i"); > return String.format("%.1f %sB", bytes / Math.pow(unit, exp), pre); > } > > can be copied into a Java program, and assuming that there is no other > method called humanReadableByteCount in the same class, it should > compile and run without renaming or re-typing. How would one know it's 'long' and not some other data type? > CoPilot might generate this from a comment like, > > // Convert a byte count to a human-readable string > > since it is mentioned over 6000 times on GitHub (and this method even > has a bug, as the article explains -- but that is a totally different > issue). That's not how AI works: it doesn't just count the number of times something is mentioned. That usually leads to unsatisfactory results. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 14:37 ` Eli Zaretskii @ 2021-07-24 14:49 ` Philip Kaludercic 2021-07-24 15:13 ` Eli Zaretskii 0 siblings, 1 reply; 46+ messages in thread From: Philip Kaludercic @ 2021-07-24 14:49 UTC (permalink / raw) To: Eli Zaretskii; +Cc: stefan, emacs-tangents, mullikine, rms, bugs Eli Zaretskii <eliz@gnu.org> writes: >> From: Philip Kaludercic <philipk@posteo.net> >> Cc: rms@gnu.org, mullikine@gmail.com, emacs-tangents@gnu.org, >> stefan@marxist.se, bugs@gnu.support >> Date: Sat, 24 Jul 2021 14:16:55 +0000 >> >> > Sorry, I don't understand what you have in mind. Can you show an >> > example of useful code that could be copied verbatim into a program >> > without at least some renaming, without breaking the program? >> >> To take the example from the article I mentioned above >> >> public static String humanReadableByteCount(long bytes, boolean si) { >> int unit = si ? 1000 : 1024; >> if (bytes < unit) return bytes + " B"; >> int exp = (int) (Math.log(bytes) / Math.log(unit)); >> String pre = (si ? "kMGTPE" : "KMGTPE").charAt(exp-1) + (si ? "" : "i"); >> return String.format("%.1f %sB", bytes / Math.pow(unit, exp), pre); >> } >> >> can be copied into a Java program, and assuming that there is no other >> method called humanReadableByteCount in the same class, it should >> compile and run without renaming or re-typing. > > How would one know it's 'long' and not some other data type? I am not sure what you mean? "long" makes sense here because Java will automatically up-cast any other type to fit. >> CoPilot might generate this from a comment like, >> >> // Convert a byte count to a human-readable string >> >> since it is mentioned over 6000 times on GitHub (and this method even >> has a bug, as the article explains -- but that is a totally different >> issue). > > That's not how AI works: it doesn't just count the number of times > something is mentioned. That usually leads to unsatisfactory results. Of course, that would be oversimplifying. At the same time, if the training samples have common patterns, a model is more likely to reproduce that behaviour. But since these are neural networks we are talking about, it is hard to determine causality to begin with, which probably makes the whole situation even more difficult (speaking as a non-lawyer). -- Philip Kaludercic ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-24 14:49 ` Philip Kaludercic @ 2021-07-24 15:13 ` Eli Zaretskii 0 siblings, 0 replies; 46+ messages in thread From: Eli Zaretskii @ 2021-07-24 15:13 UTC (permalink / raw) To: Philip Kaludercic; +Cc: stefan, emacs-tangents, mullikine, rms, bugs > From: Philip Kaludercic <philipk@posteo.net> > Cc: rms@gnu.org, mullikine@gmail.com, emacs-tangents@gnu.org, > stefan@marxist.se, bugs@gnu.support > Date: Sat, 24 Jul 2021 14:49:02 +0000 > > > How would one know it's 'long' and not some other data type? > > I am not sure what you mean? "long" makes sense here because Java will > automatically up-cast any other type to fit. So you came up with perhaps the single example that exists in the whole world where the issues I mentioned _might_ not matter, and even that only under some assumptions. A feature that aspires to be generally useful cannot possibly depend on such problematic assumptions. > >> since it is mentioned over 6000 times on GitHub (and this method even > >> has a bug, as the article explains -- but that is a totally different > >> issue). > > > > That's not how AI works: it doesn't just count the number of times > > something is mentioned. That usually leads to unsatisfactory results. > > Of course, that would be oversimplifying. At the same time, if the > training samples have common patterns, a model is more likely to > reproduce that behaviour. No, that's not it: a single example repeated in identical form many times doesn't reinforce the learned pattern. You need many similar, but different code samples, and most probably in different languages. ^ permalink raw reply [flat|nested] 46+ messages in thread
* Re: Help building Pen.el (GPT for emacs) 2021-07-23 6:51 ` Shane Mulligan 2021-07-23 10:12 ` Jean Louis @ 2021-07-25 1:06 ` Richard Stallman 1 sibling, 0 replies; 46+ messages in thread From: Richard Stallman @ 2021-07-25 1:06 UTC (permalink / raw) To: Shane Mulligan; +Cc: eliz, stefan, emacs-tangents, bugs [[[ To any NSA and FBI agents reading my email: please consider ]]] [[[ whether defending the US Constitution against all enemies, ]]] [[[ foreign or domestic, requires you to follow Snowden's example. ]]] > GPT is potentially the best thing to happen to emacs in a very long time. If GPT-3 were released as a free program, we might want to use it. Perhaps it would be very useful. Please correct me if I am mistaken, but I think GPT-3 is an unreleased program which people can use only via SaaSS. SaaSS stands for Service as a Software Substitute. It means that a "service" accepts your data, does a specific computing job, and sends you back the results. Using such a service is morally mostly equivalent to running a nonfree program -- so we cannot suggest that anyone DO that. See https://gnu.org/philosophy/who-does-that-server-really-serve.html for more explanation of this issue. > The way this will work is you will download > the free GPT model, such as GPT-j, GPT-neo or > GPT-neox and then you will have an offline and > private alternative to many things previously > you would go online for. Are you saying there is a free replacement for GPT-3 and we can run these free models with it on our own computers? That could be good news, because we could actually use it. > It will bring back power from the corporations and save it to your > computer, That sounds exciting but it is not concrete enough to think about. open source and transparent, What follows is a side issue, but it's an important side issue. "Open source" is the slogan of a campaign we don't advocate. It is partly similar to the free software movement but discards the moral foundation: the idea of freedom. We don't use the slogan "open source" because we want to advocate freedom, not forget it. See https://gnu.org/philosophy/open-source-misses-the-point.html for more explanation of the difference between free software and open source. See also https://thebaffler.com/salvos/the-meme-hustler for Evgeny Morozov's article on the same point. -- Dr Richard Stallman (https://stallman.org) Chief GNUisance of the GNU Project (https://gnu.org) Founder, Free Software Foundation (https://fsf.org) Internet Hall-of-Famer (https://internethalloffame.org) ^ permalink raw reply [flat|nested] 46+ messages in thread
[parent not found: <YN8bZEJAkWyQwjrB@protected.localdomain>]
[parent not found: <CACT87JpAcUfuRB01CcnfbL4yCTPyDoiG_WOzzxVvAW7rhj0=Mw@mail.gmail.com>]
* Re: Help building Pen.el (GPT for emacs) [not found] ` <CACT87JpAcUfuRB01CcnfbL4yCTPyDoiG_WOzzxVvAW7rhj0=Mw@mail.gmail.com> @ 2021-07-23 15:37 ` Jean Louis 0 siblings, 0 replies; 46+ messages in thread From: Jean Louis @ 2021-07-23 15:37 UTC (permalink / raw) To: Shane Mulligan; +Cc: emacs-tangents I have tried following: 1. emacs -q 2. load src/init.el 3. then add the directory src to variable load-path 4. then to load pen.el I could see that it is asking for s-join function which it cannot find, it looked like Emacs was looping, beeping, totally unresponsive. I had to kill it. I am using Emacs development version, are you also using it? In general I was thinking that init.el and init-setup.el will set it up for me. The file installation.org is unclear, what do I need to do to start using pen.el ? $ emacs -q Connection lost to X server ':0' When compiled with GTK, Emacs cannot recover from X disconnects. This is a GTK bug: https://gitlab.gnome.org/GNOME/gtk/issues/221 For details, see etc/PROBLEMS. Fatal error 6: Aborted Backtrace: emacs(+0x1516d4)[0x55ba843c36d4] emacs(+0x463d8)[0x55ba842b83d8] emacs(+0x468d3)[0x55ba842b88d3] emacs(+0x45940)[0x55ba842b7940] emacs(+0x459fe)[0x55ba842b79fe] /usr/lib/libX11.so.6(_XIOError+0x5f)[0x7fac94d5bb9f] /usr/lib/libX11.so.6(_XEventsQueued+0xa7)[0x7fac94d59217] /usr/lib/libX11.so.6(XPending+0x62)[0x7fac94d4aa92] /usr/lib/libgdk-3.so.0(+0x8d722)[0x7fac95530722] /usr/lib/libglib-2.0.so.0(g_main_context_prepare+0x1b0)[0x7fac94ed5bc0] /usr/lib/libglib-2.0.so.0(+0xa7a06)[0x7fac94f29a06] /usr/lib/libglib-2.0.so.0(g_main_context_pending+0x2a)[0x7fac94ed371a] /usr/lib/libgtk-3.so.0(gtk_events_pending+0x10)[0x7fac95779690] emacs(+0x105c5d)[0x55ba84377c5d] emacs(+0x13e932)[0x55ba843b0932] emacs(+0x13f635)[0x55ba843b1635] emacs(+0x228e9b)[0x55ba8449ae9b] emacs(+0x993fa)[0x55ba8430b3fa] emacs(+0x7d3c9)[0x55ba842ef3c9] emacs(+0x82525)[0x55ba842f4525] emacs(+0x83cf4)[0x55ba842f5cf4] emacs(+0x686f2)[0x55ba842da6f2] emacs(+0x943a9)[0x55ba843063a9] emacs(+0x14729e)[0x55ba843b929e] emacs(+0x1b4047)[0x55ba84426047] emacs(+0x137404)[0x55ba843a9404] emacs(+0x1b6773)[0x55ba84428773] emacs(+0x1373ab)[0x55ba843a93ab] emacs(+0x13cd76)[0x55ba843aed76] emacs(+0x13d0a2)[0x55ba843af0a2] emacs(+0x4e929)[0x55ba842c0929] /usr/lib/libc.so.6(__libc_start_main+0xd5)[0x7fac938bfb25] emacs(+0x4ef2e)[0x55ba842c0f2e] Aborted ~/Programming/git/pen.el $ -- Jean Take action in Free Software Foundation campaigns: https://www.fsf.org/campaigns In support of Richard M. Stallman https://stallmansupport.org/ ^ permalink raw reply [flat|nested] 46+ messages in thread
end of thread, other threads:[~2021-07-30 6:55 UTC | newest] Thread overview: 46+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- [not found] <CACT87JohxuswvDcqGOiQR7BrHoqJFG252QD6XjEuAPU2HSuWOw@mail.gmail.com> [not found] ` <CADwFkm=cN4W0Mgo_hYgwWgddoe=cXj5+WYJWnAHZmmd+rd7gKw@mail.gmail.com> [not found] ` <E1m4YYb-0005Ds-GJ@fencepost.gnu.org> [not found] ` <CACT87JqMZ+pbVDQ-5gZHMsGcfm04CoeKZn6sY5yy+1rnxCimOQ@mail.gmail.com> [not found] ` <83im1948mj.fsf@gnu.org> [not found] ` <CACT87JrCAi3Umdke6gL+_W_7k2j+21jsuT=1hq5kyOx19L2x+A@mail.gmail.com> [not found] ` <CACT87Jo41S2FJKxfPs0qP=qkXvwvcc0xnf1X6oEkjuhmAJ6w3A@mail.gmail.com> [not found] ` <YPO+bAMpqMhxDBxU@protected.localdomain> [not found] ` <83lf642jeh.fsf@gnu.org> [not found] ` <CACT87JriMaF1kFjEE_8=8FEQpAi6sxr3x3vZT3rafjY=4mQgZg@mail.gmail.com> 2021-07-19 17:00 ` Help building Pen.el (GPT for emacs) Jean Louis 2021-07-23 6:51 ` Shane Mulligan 2021-07-23 10:12 ` Jean Louis 2021-07-23 10:54 ` Eli Zaretskii 2021-07-23 11:32 ` Jean Louis 2021-07-23 11:51 ` Eli Zaretskii 2021-07-23 12:47 ` Jean Louis 2021-07-23 13:39 ` Shane Mulligan 2021-07-23 14:39 ` Jean Louis 2021-07-26 0:16 ` Richard Stallman 2021-07-26 0:28 ` Shane Mulligan 2021-07-30 3:20 ` Shane Mulligan 2021-07-30 6:55 ` Jean Louis 2021-07-23 19:33 ` Eli Zaretskii 2021-07-24 3:07 ` Jean Louis 2021-07-24 7:32 ` Eli Zaretskii 2021-07-24 7:54 ` Jean Louis 2021-07-24 8:50 ` Eli Zaretskii 2021-07-24 16:16 ` Jean Louis 2021-07-24 16:44 ` Eli Zaretskii 2021-07-24 18:01 ` Jean Louis 2021-07-25 1:09 ` Richard Stallman 2021-07-24 1:14 ` Richard Stallman 2021-07-24 2:10 ` Shane Mulligan 2021-07-24 2:34 ` Shane Mulligan 2021-07-24 3:14 ` Shane Mulligan 2021-07-24 6:49 ` Eli Zaretskii 2021-07-24 7:33 ` Jean Louis 2021-07-24 8:10 ` Eli Zaretskii 2021-07-24 8:21 ` Jean Louis 2021-07-24 8:35 ` Jean Louis 2021-07-24 8:59 ` Eli Zaretskii 2021-07-24 16:18 ` Jean Louis 2021-07-24 16:45 ` Eli Zaretskii 2021-07-24 17:57 ` Jean Louis 2021-07-24 18:15 ` Eli Zaretskii 2021-07-24 7:41 ` Philip Kaludercic 2021-07-24 7:59 ` Eli Zaretskii 2021-07-24 9:31 ` Philip Kaludercic 2021-07-24 11:19 ` Eli Zaretskii 2021-07-24 14:16 ` Philip Kaludercic 2021-07-24 14:37 ` Eli Zaretskii 2021-07-24 14:49 ` Philip Kaludercic 2021-07-24 15:13 ` Eli Zaretskii 2021-07-25 1:06 ` Richard Stallman [not found] ` <YN8bZEJAkWyQwjrB@protected.localdomain> [not found] ` <CACT87JpAcUfuRB01CcnfbL4yCTPyDoiG_WOzzxVvAW7rhj0=Mw@mail.gmail.com> 2021-07-23 15:37 ` Jean Louis
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).