▲ 756 ▼ Anthropic's Claude AI blocks video game cloning and reverse-engineering — sweeping new restrictions also hit legitimate developers and basic troubleshooting tasks. (www.tomshardware.com) submitted 1 week ago by eicker@lemmy.world to c/technology@lemmy.world 188 comments fedilink hide all child comments
[–] riskable@programming.dev 40 points 1 week ago (19 children) Why? Reverse engineering is neither unethical or illegal. permalink fedilink source parent hideshow 19 child comments replies: [–] r00ty@kbin.life 28 points 1 week ago (8 children) I never said it was. But, I'm also not an investor. I'm just saying the kind of things that scare off investors. And this would be a big one. permalink fedilink source parent hideshow 8 child comments replies: [–] schipelblorp@sh.itjust.works 11 points 1 week ago (7 children) What? The literal extermination of the human race is not enough to scare off investors. permalink fedilink source parent hideshow 7 child comments replies: [–] thejml@sh.itjust.works 18 points 1 week ago (1 child) At this point, I think they believe it’s a good indicator of potential. A way of saying just how awesome and magical and powerful the AI is. Otherwise the "AI went off the rails and breached X companies servers" would also have been met with contempt and negatively impacted them. permalink fedilink source parent hideshow 1 child comment replies: [–] RedstoneValley@sh.itjust.works 2 points 1 week ago I still can't wrap my head around how this strategy ("buy our models, they are so powerful they might kill all humans") would work in their favor, apart from one thing: restrictions would hurt Chinese models probably more than themselves. But I find it very unlikely that characters like Altman or Amodei would be capable of playing that kind of 4d chess. It's a weird flex permalink fedilink source parent [–] cardfire@sh.itjust.works 8 points 1 week ago Killing the entire species is absolutely less of a concern for investors than copyright violations. permalink fedilink source parent [–] avidamoeba@lemmy.ca 7 points 1 week ago (1 child) Have you seen the investment in fossil fuels. 😄 permalink fedilink source parent hideshow 1 child comment replies: [–] schipelblorp@sh.itjust.works 7 points 1 week ago 🤣😂🥲🙂😐😔🥺😭 permalink fedilink source parent [–] LostCarcosan@lemmy.today 4 points 1 week ago No, but if you were an investor, would you invest in a product that would make it so people can easiy recreate your other products so that they stop buying your other products? Ending humanity doesn't scare them but making $0.75 less this year does permalink fedilink source parent [–] r00ty@kbin.life 2 points 1 week ago Well, to be blunt, it depends. Would the eventual extermination of the human race cause them to lose money in the short term? permalink fedilink source parent [–] CosmicDebris@lemmy.dbzer0.com 4 points 1 week ago (9 children) It's dangerous for IP, though. permalink fedilink source parent hideshow 9 child comments replies: [–] Waterpumpee@lemmus.org -3 points 1 week ago (8 children) Only if it is protected.Which code is afaik not. permalink fedilink source parent hideshow 8 child comments replies: [–] Kirp123@lemmy.world 19 points 1 week ago In the US code is copyrighted the same as literary works. But that only covers the actual lines of code and not the ideas, functions or algorithms they implement. This means that if someone copy pastes your code word for word they are infringing your copyright but if they write their own code that does the same function as yours they are fine. permalink fedilink source parent [–] frongt@lemmy.zip 2 points 1 week ago (6 children) It absolutely is. permalink fedilink source parent hideshow 6 child comments replies: [–] beejjorgensen@lemmy.sdf.org 5 points 1 week ago (5 children) Human-generated code absolutely is, correct. But AI-generated code is still untested. AI-generated art does not have copyright protection currently. My money is that AI-generated code will have copyright protection through some mental and legal gymnastics. Because money. permalink fedilink source parent hideshow 5 child comments replies: [–] frongt@lemmy.zip 3 points 1 week ago (3 children) Oh, AI-generated code may not be, but like you said we don't have a court ruling yet. I don't think it can't be, but we'd first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. permalink fedilink source parent hideshow 3 child comments replies: [–] beejjorgensen@lemmy.sdf.org 4 points 1 week ago (2 children) but we’d first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. I'm not so convinced on that one. Copying code without permission, yes, that's illegal. Distributing copied code without permission, also illegal. But processing code (copied with permission) into weights, that doesn't feel to me like a violation. If I have a legit copy of something, I should be able to distribute computed statistics about that copy. EFF points out some hazards: https://www.eff.org/deeplinks/2026/10/rewriting-copyright-law-generative-ai-would-hurt-everyone permalink fedilink source parent hideshow 2 child comments replies: [–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent [–] Natanael@infosec.pub 1 point 1 week ago* Copyright generally requires human expression regardless of what the type of work is, it's not about what you were inspired by If you did derive your work from something else and yours is similar enough, then now both you and the original author has separate copyrights on the result which both needs to be licensed to sell the work (prime example is book translations) permalink fedilink source parent
[–] r00ty@kbin.life 28 points 1 week ago (8 children) I never said it was. But, I'm also not an investor. I'm just saying the kind of things that scare off investors. And this would be a big one. permalink fedilink source parent hideshow 8 child comments replies: [–] schipelblorp@sh.itjust.works 11 points 1 week ago (7 children) What? The literal extermination of the human race is not enough to scare off investors. permalink fedilink source parent hideshow 7 child comments replies: [–] thejml@sh.itjust.works 18 points 1 week ago (1 child) At this point, I think they believe it’s a good indicator of potential. A way of saying just how awesome and magical and powerful the AI is. Otherwise the "AI went off the rails and breached X companies servers" would also have been met with contempt and negatively impacted them. permalink fedilink source parent hideshow 1 child comment replies: [–] RedstoneValley@sh.itjust.works 2 points 1 week ago I still can't wrap my head around how this strategy ("buy our models, they are so powerful they might kill all humans") would work in their favor, apart from one thing: restrictions would hurt Chinese models probably more than themselves. But I find it very unlikely that characters like Altman or Amodei would be capable of playing that kind of 4d chess. It's a weird flex permalink fedilink source parent [–] cardfire@sh.itjust.works 8 points 1 week ago Killing the entire species is absolutely less of a concern for investors than copyright violations. permalink fedilink source parent [–] avidamoeba@lemmy.ca 7 points 1 week ago (1 child) Have you seen the investment in fossil fuels. 😄 permalink fedilink source parent hideshow 1 child comment replies: [–] schipelblorp@sh.itjust.works 7 points 1 week ago 🤣😂🥲🙂😐😔🥺😭 permalink fedilink source parent [–] LostCarcosan@lemmy.today 4 points 1 week ago No, but if you were an investor, would you invest in a product that would make it so people can easiy recreate your other products so that they stop buying your other products? Ending humanity doesn't scare them but making $0.75 less this year does permalink fedilink source parent [–] r00ty@kbin.life 2 points 1 week ago Well, to be blunt, it depends. Would the eventual extermination of the human race cause them to lose money in the short term? permalink fedilink source parent
[–] schipelblorp@sh.itjust.works 11 points 1 week ago (7 children) What? The literal extermination of the human race is not enough to scare off investors. permalink fedilink source parent hideshow 7 child comments replies: [–] thejml@sh.itjust.works 18 points 1 week ago (1 child) At this point, I think they believe it’s a good indicator of potential. A way of saying just how awesome and magical and powerful the AI is. Otherwise the "AI went off the rails and breached X companies servers" would also have been met with contempt and negatively impacted them. permalink fedilink source parent hideshow 1 child comment replies: [–] RedstoneValley@sh.itjust.works 2 points 1 week ago I still can't wrap my head around how this strategy ("buy our models, they are so powerful they might kill all humans") would work in their favor, apart from one thing: restrictions would hurt Chinese models probably more than themselves. But I find it very unlikely that characters like Altman or Amodei would be capable of playing that kind of 4d chess. It's a weird flex permalink fedilink source parent [–] cardfire@sh.itjust.works 8 points 1 week ago Killing the entire species is absolutely less of a concern for investors than copyright violations. permalink fedilink source parent [–] avidamoeba@lemmy.ca 7 points 1 week ago (1 child) Have you seen the investment in fossil fuels. 😄 permalink fedilink source parent hideshow 1 child comment replies: [–] schipelblorp@sh.itjust.works 7 points 1 week ago 🤣😂🥲🙂😐😔🥺😭 permalink fedilink source parent [–] LostCarcosan@lemmy.today 4 points 1 week ago No, but if you were an investor, would you invest in a product that would make it so people can easiy recreate your other products so that they stop buying your other products? Ending humanity doesn't scare them but making $0.75 less this year does permalink fedilink source parent [–] r00ty@kbin.life 2 points 1 week ago Well, to be blunt, it depends. Would the eventual extermination of the human race cause them to lose money in the short term? permalink fedilink source parent
[–] thejml@sh.itjust.works 18 points 1 week ago (1 child) At this point, I think they believe it’s a good indicator of potential. A way of saying just how awesome and magical and powerful the AI is. Otherwise the "AI went off the rails and breached X companies servers" would also have been met with contempt and negatively impacted them. permalink fedilink source parent hideshow 1 child comment replies: [–] RedstoneValley@sh.itjust.works 2 points 1 week ago I still can't wrap my head around how this strategy ("buy our models, they are so powerful they might kill all humans") would work in their favor, apart from one thing: restrictions would hurt Chinese models probably more than themselves. But I find it very unlikely that characters like Altman or Amodei would be capable of playing that kind of 4d chess. It's a weird flex permalink fedilink source parent
[–] RedstoneValley@sh.itjust.works 2 points 1 week ago I still can't wrap my head around how this strategy ("buy our models, they are so powerful they might kill all humans") would work in their favor, apart from one thing: restrictions would hurt Chinese models probably more than themselves. But I find it very unlikely that characters like Altman or Amodei would be capable of playing that kind of 4d chess. It's a weird flex permalink fedilink source parent
[–] cardfire@sh.itjust.works 8 points 1 week ago Killing the entire species is absolutely less of a concern for investors than copyright violations. permalink fedilink source parent
[–] avidamoeba@lemmy.ca 7 points 1 week ago (1 child) Have you seen the investment in fossil fuels. 😄 permalink fedilink source parent hideshow 1 child comment replies: [–] schipelblorp@sh.itjust.works 7 points 1 week ago 🤣😂🥲🙂😐😔🥺😭 permalink fedilink source parent
[–] LostCarcosan@lemmy.today 4 points 1 week ago No, but if you were an investor, would you invest in a product that would make it so people can easiy recreate your other products so that they stop buying your other products? Ending humanity doesn't scare them but making $0.75 less this year does permalink fedilink source parent
[–] r00ty@kbin.life 2 points 1 week ago Well, to be blunt, it depends. Would the eventual extermination of the human race cause them to lose money in the short term? permalink fedilink source parent
[–] CosmicDebris@lemmy.dbzer0.com 4 points 1 week ago (9 children) It's dangerous for IP, though. permalink fedilink source parent hideshow 9 child comments replies: [–] Waterpumpee@lemmus.org -3 points 1 week ago (8 children) Only if it is protected.Which code is afaik not. permalink fedilink source parent hideshow 8 child comments replies: [–] Kirp123@lemmy.world 19 points 1 week ago In the US code is copyrighted the same as literary works. But that only covers the actual lines of code and not the ideas, functions or algorithms they implement. This means that if someone copy pastes your code word for word they are infringing your copyright but if they write their own code that does the same function as yours they are fine. permalink fedilink source parent [–] frongt@lemmy.zip 2 points 1 week ago (6 children) It absolutely is. permalink fedilink source parent hideshow 6 child comments replies: [–] beejjorgensen@lemmy.sdf.org 5 points 1 week ago (5 children) Human-generated code absolutely is, correct. But AI-generated code is still untested. AI-generated art does not have copyright protection currently. My money is that AI-generated code will have copyright protection through some mental and legal gymnastics. Because money. permalink fedilink source parent hideshow 5 child comments replies: [–] frongt@lemmy.zip 3 points 1 week ago (3 children) Oh, AI-generated code may not be, but like you said we don't have a court ruling yet. I don't think it can't be, but we'd first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. permalink fedilink source parent hideshow 3 child comments replies: [–] beejjorgensen@lemmy.sdf.org 4 points 1 week ago (2 children) but we’d first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. I'm not so convinced on that one. Copying code without permission, yes, that's illegal. Distributing copied code without permission, also illegal. But processing code (copied with permission) into weights, that doesn't feel to me like a violation. If I have a legit copy of something, I should be able to distribute computed statistics about that copy. EFF points out some hazards: https://www.eff.org/deeplinks/2026/10/rewriting-copyright-law-generative-ai-would-hurt-everyone permalink fedilink source parent hideshow 2 child comments replies: [–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent [–] Natanael@infosec.pub 1 point 1 week ago* Copyright generally requires human expression regardless of what the type of work is, it's not about what you were inspired by If you did derive your work from something else and yours is similar enough, then now both you and the original author has separate copyrights on the result which both needs to be licensed to sell the work (prime example is book translations) permalink fedilink source parent
[–] Waterpumpee@lemmus.org -3 points 1 week ago (8 children) Only if it is protected.Which code is afaik not. permalink fedilink source parent hideshow 8 child comments replies: [–] Kirp123@lemmy.world 19 points 1 week ago In the US code is copyrighted the same as literary works. But that only covers the actual lines of code and not the ideas, functions or algorithms they implement. This means that if someone copy pastes your code word for word they are infringing your copyright but if they write their own code that does the same function as yours they are fine. permalink fedilink source parent [–] frongt@lemmy.zip 2 points 1 week ago (6 children) It absolutely is. permalink fedilink source parent hideshow 6 child comments replies: [–] beejjorgensen@lemmy.sdf.org 5 points 1 week ago (5 children) Human-generated code absolutely is, correct. But AI-generated code is still untested. AI-generated art does not have copyright protection currently. My money is that AI-generated code will have copyright protection through some mental and legal gymnastics. Because money. permalink fedilink source parent hideshow 5 child comments replies: [–] frongt@lemmy.zip 3 points 1 week ago (3 children) Oh, AI-generated code may not be, but like you said we don't have a court ruling yet. I don't think it can't be, but we'd first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. permalink fedilink source parent hideshow 3 child comments replies: [–] beejjorgensen@lemmy.sdf.org 4 points 1 week ago (2 children) but we’d first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. I'm not so convinced on that one. Copying code without permission, yes, that's illegal. Distributing copied code without permission, also illegal. But processing code (copied with permission) into weights, that doesn't feel to me like a violation. If I have a legit copy of something, I should be able to distribute computed statistics about that copy. EFF points out some hazards: https://www.eff.org/deeplinks/2026/10/rewriting-copyright-law-generative-ai-would-hurt-everyone permalink fedilink source parent hideshow 2 child comments replies: [–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent [–] Natanael@infosec.pub 1 point 1 week ago* Copyright generally requires human expression regardless of what the type of work is, it's not about what you were inspired by If you did derive your work from something else and yours is similar enough, then now both you and the original author has separate copyrights on the result which both needs to be licensed to sell the work (prime example is book translations) permalink fedilink source parent
[–] Kirp123@lemmy.world 19 points 1 week ago In the US code is copyrighted the same as literary works. But that only covers the actual lines of code and not the ideas, functions or algorithms they implement. This means that if someone copy pastes your code word for word they are infringing your copyright but if they write their own code that does the same function as yours they are fine. permalink fedilink source parent
[–] frongt@lemmy.zip 2 points 1 week ago (6 children) It absolutely is. permalink fedilink source parent hideshow 6 child comments replies: [–] beejjorgensen@lemmy.sdf.org 5 points 1 week ago (5 children) Human-generated code absolutely is, correct. But AI-generated code is still untested. AI-generated art does not have copyright protection currently. My money is that AI-generated code will have copyright protection through some mental and legal gymnastics. Because money. permalink fedilink source parent hideshow 5 child comments replies: [–] frongt@lemmy.zip 3 points 1 week ago (3 children) Oh, AI-generated code may not be, but like you said we don't have a court ruling yet. I don't think it can't be, but we'd first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. permalink fedilink source parent hideshow 3 child comments replies: [–] beejjorgensen@lemmy.sdf.org 4 points 1 week ago (2 children) but we’d first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. I'm not so convinced on that one. Copying code without permission, yes, that's illegal. Distributing copied code without permission, also illegal. But processing code (copied with permission) into weights, that doesn't feel to me like a violation. If I have a legit copy of something, I should be able to distribute computed statistics about that copy. EFF points out some hazards: https://www.eff.org/deeplinks/2026/10/rewriting-copyright-law-generative-ai-would-hurt-everyone permalink fedilink source parent hideshow 2 child comments replies: [–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent [–] Natanael@infosec.pub 1 point 1 week ago* Copyright generally requires human expression regardless of what the type of work is, it's not about what you were inspired by If you did derive your work from something else and yours is similar enough, then now both you and the original author has separate copyrights on the result which both needs to be licensed to sell the work (prime example is book translations) permalink fedilink source parent
[–] beejjorgensen@lemmy.sdf.org 5 points 1 week ago (5 children) Human-generated code absolutely is, correct. But AI-generated code is still untested. AI-generated art does not have copyright protection currently. My money is that AI-generated code will have copyright protection through some mental and legal gymnastics. Because money. permalink fedilink source parent hideshow 5 child comments replies: [–] frongt@lemmy.zip 3 points 1 week ago (3 children) Oh, AI-generated code may not be, but like you said we don't have a court ruling yet. I don't think it can't be, but we'd first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. permalink fedilink source parent hideshow 3 child comments replies: [–] beejjorgensen@lemmy.sdf.org 4 points 1 week ago (2 children) but we’d first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. I'm not so convinced on that one. Copying code without permission, yes, that's illegal. Distributing copied code without permission, also illegal. But processing code (copied with permission) into weights, that doesn't feel to me like a violation. If I have a legit copy of something, I should be able to distribute computed statistics about that copy. EFF points out some hazards: https://www.eff.org/deeplinks/2026/10/rewriting-copyright-law-generative-ai-would-hurt-everyone permalink fedilink source parent hideshow 2 child comments replies: [–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent [–] Natanael@infosec.pub 1 point 1 week ago* Copyright generally requires human expression regardless of what the type of work is, it's not about what you were inspired by If you did derive your work from something else and yours is similar enough, then now both you and the original author has separate copyrights on the result which both needs to be licensed to sell the work (prime example is book translations) permalink fedilink source parent
[–] frongt@lemmy.zip 3 points 1 week ago (3 children) Oh, AI-generated code may not be, but like you said we don't have a court ruling yet. I don't think it can't be, but we'd first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. permalink fedilink source parent hideshow 3 child comments replies: [–] beejjorgensen@lemmy.sdf.org 4 points 1 week ago (2 children) but we’d first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. I'm not so convinced on that one. Copying code without permission, yes, that's illegal. Distributing copied code without permission, also illegal. But processing code (copied with permission) into weights, that doesn't feel to me like a violation. If I have a legit copy of something, I should be able to distribute computed statistics about that copy. EFF points out some hazards: https://www.eff.org/deeplinks/2026/10/rewriting-copyright-law-generative-ai-would-hurt-everyone permalink fedilink source parent hideshow 2 child comments replies: [–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent
[–] beejjorgensen@lemmy.sdf.org 4 points 1 week ago (2 children) but we’d first have to have models trained without violating the rights of all the authors of the works in the training data. Like if you trained a model only on GPL software, then the model and its output should also be GPL. I'm not so convinced on that one. Copying code without permission, yes, that's illegal. Distributing copied code without permission, also illegal. But processing code (copied with permission) into weights, that doesn't feel to me like a violation. If I have a legit copy of something, I should be able to distribute computed statistics about that copy. EFF points out some hazards: https://www.eff.org/deeplinks/2026/10/rewriting-copyright-law-generative-ai-would-hurt-everyone permalink fedilink source parent hideshow 2 child comments replies: [–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent
[–] frongt@lemmy.zip 1 point 1 week ago (1 child) Sure, but it's not just statistics like "30% of lines contain the letter E". It reproduces the material, sometimes verbatim. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. permalink fedilink source parent hideshow 1 child comment replies: [–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent
[–] beejjorgensen@lemmy.sdf.org 1 point 1 week ago Reproduction of the material we agree is against the law. A model is very obviously a derivative work, as it has 0% original content input. Under fair use law, it also very obviously fails all four tests. Computing the letter frequency has 0% original content input, as well, but is definitely not a derivative work. Do you disagree? permalink fedilink source parent
[–] Natanael@infosec.pub 1 point 1 week ago* Copyright generally requires human expression regardless of what the type of work is, it's not about what you were inspired by If you did derive your work from something else and yours is similar enough, then now both you and the original author has separate copyrights on the result which both needs to be licensed to sell the work (prime example is book translations) permalink fedilink source parent