In the past, many developers were against copyright law because they saw it as a way for big corps to stifle competition and curb creativity in order to increase their profits. A lot of people right now invoke the violation of the same copyright law because the tide has changed and now companies, by ignoring copyright law, are hurting artists/smaller companies and/or not contributing back or unlawfully closing the code in the case of GPL.
I don't see any kind of hypocritical stance here honestly. All this time the criticism of the enforcement of copyright law or now the lack of it just reflects the fact that some people are genuinely concerned that bad actors(big corps) are using the law to damage society in order to pursue their own interests.
Well, no, not really - there's no "they" in general. Copyright law is constructed by design for large institutions, lobbying, case law, and attendant legislation allow it to be abused by data hoarders. It was never built to protect individual creators or authors or artists, despite the PR campaigns and marketing for it.
There are just companies big enough to ignore those institutions for which copyright law was created, like Google, etc, and the fair use exceptions Google carved out empowered AI companies to make similar moves. To an extent.
What's hurting artists and smaller companies and various licensing schemes intended to push back is the fundamental structure of the laws.
All copyrights need to be nuked and replaced. If we want to support individuals and maximize protections of individuals, and we want to disincentivize data hoarders that do nothing but recycle old content and IP in perpetual rent-seeking schemes, we should implement a 5 year copyright system.
The first 5 years, you get total copyright, any commercial use has to be licensed explicitly, fair use remains largely as it is now. From year 6-10, fair use gets extended - you have to credit the creator, pay a 15% royalty direct to the creator, but otherwise you can use it for anything. Year 10-20, you must credit the creator, but otherwise the media is in the public domain.
We should be pushing for and incentivizing creative use of data, empowering as many people as possible to use it and riff on it and make the culture vibrant and active and free from centralizing, manipulative actors.
99% of commercial profits come from the first 5 years after any piece of media gets published - book, music, film, artwork, etc. Copyright should protect that, but after that 5 years, things open up so the price you pay in order to participate in the marketplace which the US fosters is that your content thereafter becomes available for use by anyone, and they have to pay a fair markup for the use. You don't get to deny anyone the use of the media. You'll get credited, paid, and then after 10 years, it's public domain + mandatory credits, kinda like an MIT license style. After 20 years, it's fully public domain.
Throw in things like "if you're not paid the royalty, you can sue for up to half of the total revenues generated by the offending work" or something appropriately scaled to prohibit casual abuses, but not totally explode someone's life over honest mistakes, and scale between the two extremes accordingly.
Things like Sony and Disney and Hollywood studios are evil. They're effectively data cartels and hoarders, rarely producing anything, gatekeeping access and socializing, imposing obscene contracts on naive artists and creators, exploiting everything and everyone they touch without returning concurrent value to society. They don't deserve consideration or protection under a sane copyright system, especially in a world with gigabit internet everywhere. Screw the MAFIAA and all the people responsible for things ending up like they have.
Until then, pirate everything. If you feel an ethical obligation to pay, then do the research and send some crypto or a $20 bill in the mail to the author or creator. All sorts of people have crypto wallets, these days.
Most people in my social circles are various flavours of anti-AI, and it drives me crazy how many of them, who were once stridently anti-copyright, are now using copyright as one of the great pillars of AI opposition
I'm heavily anti-copyright. I don't think it should exist. However, as long as it exists, I want it to be applied consistently across the board: AI shouldn't get to use Open Source code while ignoring its license, until Open Source developers get to use proprietary code while ignoring its license. Ditto art, movies, books, etc.
Usually when I see this opinion (yours), it leans on an uncharitable coloring of everybody who sees problems with copyright as "anti-copyright", when really those people largely are happy with the concept of protecting an individual's work. I.e. it is the age-old "those people" argument, where "those people" are a made-up conglomerate of opinions that are real, but come from slightly different contexts and from different people, throwing away those variables to create the illusion of a hypocrite.
I think this is a good explanation, and even if this isn't what the OP says, I see arguments like this frequently.
In the case of copyright, think of it as anti-current-implementation of copyright rather than anti-copyright. For example, you could oppose the current copyright term, but that doesn't mean you are anti-copyright. Quite the opposite, in fact.
Don't worry, they will flip and become staunch supporters of AI too once it starts to benefit them. Then they will immediately forget the copyright issue.
> Most people in my social circles are various flavours of anti-AI, and it drives me crazy how many of them, who were once stridently anti-copyright, are now using copyright as one of the great pillars of AI opposition
As the article has pointed out, it's not the principle that has changed, but the scale. Lots of things that are tolerable at small scale (e.g. lying, stealing) become disruptive to society at larger scale.
Copyright has been used in the past as a way for corporations to rent-seek and limit innovation. Now it may be the only legal means to stop them from doing that.
It's not illegal for me to drive from New York to San Francisco on Interstate 80. But if I were able to endlessly duplicate myself and my car, about a million of me could hog the entire country-wide highway. Not sure if even that would be "illegal", but it would sure be annoying, and I suspect there would need to be laws updated / rewritten to account for my ability to duplicate myself without limit. I doubt society would just agree, "yep, one of you on I-80 is okay, and so a million of you on I-80 is also okay."
Were they really 100% anti-copyright though? By and large, copyright is the reason most of us have a job and get paid. There are things that shouldn't necessarily be copyright-able like APIs, and copyright probably exists for too long in certain cases, but a world without copyright doesn't really work with our current economic model.
Everyone is anti-copyright until they understand what copyright means.
In the matter of reforming things, as distinct from deforming them, there is one plain and simple principle; a principle which will probably be called a paradox. There exists in such a case a certain institution or law; let us say, for the sake of simplicity, a fence or gate erected across a road. The more modern type of reformer goes gaily up to it and says, "I don't see the use of this; let us clear it away." To which the more intelligent type of reformer will do well to answer: "If you don't see the use of it, I certainly won't let you clear it away. Go away and think. Then, when you can come back and tell me that you do see the use of it, I may allow you to destroy it."
There is a difference between distributing pirated copies of popular media by already rich artists who you know get paid anyway, and the systematic art theft of AI machines who “create” new art based on artists works who may or may not have been paid for it, and definitely didn’t get credited.
Both are copyright infringements, but only the latter is art theft.
Copyright infringement use to be the absolute worst crime imaginable if you asked the tech industry, that is until they started doing it themselves at scale and now they are claiming it’s the law that’s broken, it’s crazy.
All of that last one really says is that broadly speaking the average person has no idea what free speech actually is and the kinds of things that it covers. I put it in the same bucket as like the young kids uploading to YouTube with the comment no copyright infringement intended thinking it's like plagiarism.
> Copyright infringement use to be the absolute worst crime imaginable if you asked the tech industry
Every day hundreds of links to archive.is are posted[1] to this website to get around paywalls. Technologists built file sharing tools to subvert copyright. It has never been one of the worst crimes imaginable in tech circles.
"Tech circles" was never the claim. The original phrase was "tech industry", and that seems to be accurate. The post replying to it may have misread or misinterpreted what "tech industry" means. (Or perhaps the term is simply ambiguous and each person who reads it comes away with a different meaning!)
> I've never met someone who wasn't on Aaron's side on that one.
This rather says something about the people by who you are surrounded. I know quite a lot of people who are on Aaron Swartz side here, for example people who are in academia or those who left academia but are still deeply interested in scientific topics.
Maybe the time horizon for a statement like that shouldn’t include the decades before most current tech companies existed, much less at this scale even for the few still kicking around from 50 years ago.
The tech industry? I don’t remember that being the case, at least not in general. Content owners yes— and there’s tech overlap there with Sony and some others. Beside that it’s never been a major hill for tech to die on, except in having to implement systems to deal with DMCA takedowns, and that only as half baked as they could get away with. Which unfortunately has meant “not in favor of users” when it comes to to failure modes and where and how to default actions.
The attitude of the tech industry has always been much more vague (example: Google Books), and people from the hacker culture, who often work as programmers, are traditionally rather sceptical of at least the concrete manifestation of the copyright system ("information wants to be free", circumvention of paywalls, Guerilla Open Access Manifesto, ...).
I think this is part of a recurring pattern in tech of pushing boundaries around copyright.
In the last few years, we had Google scanning books, Google threatening to shut down News in Canada rather than pay publishers, LLMs summarizing articles on social platforms, crawlers bypassing paywalls, and so on.
Each time, the industry frames it as their interpretation of the current law, which were usually not written with these specific future use cases in mind.
In my view the current discussion regarding Gen AI is similar.
Basically a rage bait. If the law was bad, does it make it okay to violate it? In fact Anthropic is literally paying $1.5B on the copyright settlement, that indicates its completely a settled issue that AI companies have been violating this law. Some have been caught and fined, others are been lucky or that influence over the government.
> Copyright Law Was Built for Human Scale
No where in the law it has this kinda scoped limits. It has a time limit and scale doesnt not matter. Scale matter in a way that its gets harder to enforces buts that not the fault of copyright law. If you steal at a big scale, its still stealing.
> Basically a rage bait. If the law was bad, does it make it okay to violate it? In fact Anthropic is literally paying $1.5B on the copyright settlement, that indicates its completely a settled issue that AI companies have been violating this law. Some have been caught and fined, others are been lucky or that influence over the government.
Yes, but they were found not liable for copying the books they purchased. They were found liable for the books they torrented.
The former is something publishers still want to address
(I know these days it feels silly to bring this up, but...)
That is not how the separation of power is supposed to work. If a law is bad, politics (preferably a democratic process representing the people) replaces the law with a better one. Until the new law comes into effect, everyone is supposed to abide by the old law, even if it's bad.
Laws are created for lobbyist not people. Nothing works in the real world like in the text book democracy. Everything that was written about democracy is as naive as young adults fiction. But I find it refreshing that the real world events finally force more and more people to realize that.
Not really. I grew up in a socialist (communist?) society. A form of democracy closer to ideal than whatever US is doing for the last 2 centuries or more arrived in my teen years. My personal freedoms were mostly unaffected. Most treasured freedom I acquired was the freedom to gtfo which I'm happily using right now.
IMHO AI generated content should be treated the same way with how human generated content and I don't see the problem.
However as with technology the problem is a bit different, e.g.: When subletting your apartment requires manual effort, this is not a problem. Automated, it became an industry and that's a huge headache.
I think this is the key point where the derived work has unlimited possibility that they want to curb it early on. In a way it's a fair effort to keep human's competitiveness but may prove to be futile.
> IMHO AI generated content should be treated the same way with how human generated content and I don't see the problem.
The problem is that AIs don't have rights. So you literally can't treat it the same.
If I "make" something in a way I'm not allowed to (e.g. copying), the law will go after me. If AI makes something I'm not allowed to, I can just say "whoopsie doodle", blame the AI, and there are no repercussions for anyone.
Hopefully, future legislation will cater less to publishers and copyright trolls. I'm not optimistic though. While certain kinds of publishers are indeed becoming less powerful, sports-related media conglomerates are successfully lobbying for more surveillance.
The general population will likely get the worst of both worlds, with copyright trolls getting to enforce unjust laws against regular people, while big tech gets to pay their way out.
I would very much like to see us return to something resembling the original copyright terms. Like: Manually file for 22 years of enforcement. Then manually file 20 years later for 22 more.
You could create some great masterpiece at 18 and live off of it until you are 62 and starting to take social security payments.
> Personally I feel that the excessive duration of copyright just weakens authors arguments against AI.
That excess exists precisely because of the industry’s clout. For decades, rightsholders successfully lobbied Congress to extend copyright term again and again. That process appears to have finally plateaued, which is why early Mickey Mouse has now entered the public domain.
And notably, since the rise of AI, the government has not been especially quick or eager to step in and defend rightsholders.
> If even WWII-era documents are still under copyright, building a model respecting that would be impossible.
"We can't do this legally, so we should be allowed to ignore the law."
If you can't build a model while respecting licenses, don't build a model.
I don't want copyright to exist, at any duration, and I certainly think it should be much shorter than it is. However, as long as it exists, AI should not get any exception to it; such an exception inherently privileges massive entities over small entities or individuals.
> If you can't do something which looks reasonable within the laws, it means that the laws have to change.
Then change the law. For everyone, consistently, not just massive AI companies. Until then, deal with it.
> I don't see how building such a model of historical documents would impact the authors
It impacts the authors of new documents, who do not get to copy those historical documents, while an AI competing with them gets to. If you want those historical documents to be freely available, make copyright stop applying to them. For everyone, not just massive AI companies.
I'm personally in favor of a heavily reduced copyright duration similar to patents and so if AI makes people realize how stupid the current laws look, I take it.
That'd be fine if AI waited for the law to change, rather than just breaking it and trying to grow fast enough to make countries scared to enforce their own laws against it.
The people campaigning for a change previously mostly were librarians, tech enthusiasts, archivists and pirate parties, now it's megacorporations worth billions of dollars.
I have a feeling that they are going to fare better odds, the world runs on money.
They might fare better odds at using their ill-gotten goods to lobby for changes that benefit them, and do not help others or are even harmful to others.
copyright is not “broken,” it was always a two-faced scam designed to protect “owners” at the expense of creators. don’t expect this blatant hypocrisy to kill copyright, either - death of copyright is a slippery slope leading straight to communist utopia, and the death of the global acqui-parasite class
I think the problem is that “transformation” has never been clearly defined in copyright law, and that ambiguity is exactly what AI companies lean on in their defense.
At a human scale, those boundaries get clarified through litigation on a case by case basis once an infringement becomes large or obvious enough. But there has always been a gray area around when a derivative work crosses the line into infringement.
And AI didn’t create that ambiguity, it just pushed it to a more extreme scale
Copyright law has always been excessively restrictive, and is long overdue for reform. The informal practices that people have been following (i.e. free creation and distribution but no monetization and no confusion with official releases) are pretty close to what a reasonable law would state.
It is true but meaningless. Derivative works are not illegal, and you don’t need authorisation to create one.
So if you draw sonic in your living room you are indeed creating an unauthorised derivative work. And someone can call it an unauthorised derivative work. And the only reaction that should induce is raising an eyebrow and replying “ok?”
LLMs are often framed as lossy compression, and surely converting some copyrighted Sonic the Hedgehog image from PNG to JPEG is considered copyright infringement, no?!
It's not true. You don't need "authorization" to create a derivative work. You do need a license to distribute copyrighted works, including derivatives. And this only matters when you are distributing it to a sizable audience.
For example, if you rent a movie, you can watch it with your family. Nobody is going to sue you for distributing the movie with 5 people in your room. That's pure nonsense. Same with music, books, etc.
If you try to play the movie in an establishment with dozens of people, then it can become a problem, because you're essentially a theater now.
I'm not a lawyer so I don't know what the law is on selling fan art on a convention or even privately commissioned fan artwork. But things aren't as draconian as people assume it is.
In the past, many developers were against copyright law because they saw it as a way for big corps to stifle competition and curb creativity in order to increase their profits. A lot of people right now invoke the violation of the same copyright law because the tide has changed and now companies, by ignoring copyright law, are hurting artists/smaller companies and/or not contributing back or unlawfully closing the code in the case of GPL.
I don't see any kind of hypocritical stance here honestly. All this time the criticism of the enforcement of copyright law or now the lack of it just reflects the fact that some people are genuinely concerned that bad actors(big corps) are using the law to damage society in order to pursue their own interests.
Well, no, not really - there's no "they" in general. Copyright law is constructed by design for large institutions, lobbying, case law, and attendant legislation allow it to be abused by data hoarders. It was never built to protect individual creators or authors or artists, despite the PR campaigns and marketing for it.
There are just companies big enough to ignore those institutions for which copyright law was created, like Google, etc, and the fair use exceptions Google carved out empowered AI companies to make similar moves. To an extent.
What's hurting artists and smaller companies and various licensing schemes intended to push back is the fundamental structure of the laws.
All copyrights need to be nuked and replaced. If we want to support individuals and maximize protections of individuals, and we want to disincentivize data hoarders that do nothing but recycle old content and IP in perpetual rent-seeking schemes, we should implement a 5 year copyright system.
The first 5 years, you get total copyright, any commercial use has to be licensed explicitly, fair use remains largely as it is now. From year 6-10, fair use gets extended - you have to credit the creator, pay a 15% royalty direct to the creator, but otherwise you can use it for anything. Year 10-20, you must credit the creator, but otherwise the media is in the public domain.
We should be pushing for and incentivizing creative use of data, empowering as many people as possible to use it and riff on it and make the culture vibrant and active and free from centralizing, manipulative actors.
99% of commercial profits come from the first 5 years after any piece of media gets published - book, music, film, artwork, etc. Copyright should protect that, but after that 5 years, things open up so the price you pay in order to participate in the marketplace which the US fosters is that your content thereafter becomes available for use by anyone, and they have to pay a fair markup for the use. You don't get to deny anyone the use of the media. You'll get credited, paid, and then after 10 years, it's public domain + mandatory credits, kinda like an MIT license style. After 20 years, it's fully public domain.
Throw in things like "if you're not paid the royalty, you can sue for up to half of the total revenues generated by the offending work" or something appropriately scaled to prohibit casual abuses, but not totally explode someone's life over honest mistakes, and scale between the two extremes accordingly.
Things like Sony and Disney and Hollywood studios are evil. They're effectively data cartels and hoarders, rarely producing anything, gatekeeping access and socializing, imposing obscene contracts on naive artists and creators, exploiting everything and everyone they touch without returning concurrent value to society. They don't deserve consideration or protection under a sane copyright system, especially in a world with gigabit internet everywhere. Screw the MAFIAA and all the people responsible for things ending up like they have.
Until then, pirate everything. If you feel an ethical obligation to pay, then do the research and send some crypto or a $20 bill in the mail to the author or creator. All sorts of people have crypto wallets, these days.
Most people in my social circles are various flavours of anti-AI, and it drives me crazy how many of them, who were once stridently anti-copyright, are now using copyright as one of the great pillars of AI opposition
I'm heavily anti-copyright. I don't think it should exist. However, as long as it exists, I want it to be applied consistently across the board: AI shouldn't get to use Open Source code while ignoring its license, until Open Source developers get to use proprietary code while ignoring its license. Ditto art, movies, books, etc.
Usually when I see this opinion (yours), it leans on an uncharitable coloring of everybody who sees problems with copyright as "anti-copyright", when really those people largely are happy with the concept of protecting an individual's work. I.e. it is the age-old "those people" argument, where "those people" are a made-up conglomerate of opinions that are real, but come from slightly different contexts and from different people, throwing away those variables to create the illusion of a hypocrite.
> it leans on an uncharitable coloring of everybody who sees problems with copyright as "anti-copyright"
That's the charitable coloring. Owning concepts or ideas, and trying to police others' use of ideas you """own""" is absurd.
I think this is a good explanation, and even if this isn't what the OP says, I see arguments like this frequently.
In the case of copyright, think of it as anti-current-implementation of copyright rather than anti-copyright. For example, you could oppose the current copyright term, but that doesn't mean you are anti-copyright. Quite the opposite, in fact.
I feel a stark difference if people break the law to watch movies or if a company does it for profit
Don't worry, they will flip and become staunch supporters of AI too once it starts to benefit them. Then they will immediately forget the copyright issue.
It serves to demonstrate that in most cases, "principles" are only a rhetorical device.
I suppose both copyright and AI are problematic. And the article is thought provoking.
> Most people in my social circles are various flavours of anti-AI, and it drives me crazy how many of them, who were once stridently anti-copyright, are now using copyright as one of the great pillars of AI opposition
As the article has pointed out, it's not the principle that has changed, but the scale. Lots of things that are tolerable at small scale (e.g. lying, stealing) become disruptive to society at larger scale.
Copyright has been used in the past as a way for corporations to rent-seek and limit innovation. Now it may be the only legal means to stop them from doing that.
It's not illegal for me to drive from New York to San Francisco on Interstate 80. But if I were able to endlessly duplicate myself and my car, about a million of me could hog the entire country-wide highway. Not sure if even that would be "illegal", but it would sure be annoying, and I suspect there would need to be laws updated / rewritten to account for my ability to duplicate myself without limit. I doubt society would just agree, "yep, one of you on I-80 is okay, and so a million of you on I-80 is also okay."
Were they really 100% anti-copyright though? By and large, copyright is the reason most of us have a job and get paid. There are things that shouldn't necessarily be copyright-able like APIs, and copyright probably exists for too long in certain cases, but a world without copyright doesn't really work with our current economic model.
I'm curious what the circle is, because it doesn't match up with my circle. So, I'm genuinely curious what you mean by "anti-copyright".
Anti-copyright usually means, anti big corps who hold culture hostage.
Maybe they’re more anti-blatant-unequal-and-unfair-enforcement-of-copyright-law?
So people are pitting two foes against one another and that’s supposed to… be bad?
Everyone is anti-copyright until they understand what copyright means.
There is a difference between distributing pirated copies of popular media by already rich artists who you know get paid anyway, and the systematic art theft of AI machines who “create” new art based on artists works who may or may not have been paid for it, and definitely didn’t get credited.
Both are copyright infringements, but only the latter is art theft.
Copyright infringement use to be the absolute worst crime imaginable if you asked the tech industry, that is until they started doing it themselves at scale and now they are claiming it’s the law that’s broken, it’s crazy.
I think you can replace copyright with whatever the law <insert big company here> wants to break
Last but not least, generating csam and deep fakes porn on social medias and having to see it called free speech
All of that last one really says is that broadly speaking the average person has no idea what free speech actually is and the kinds of things that it covers. I put it in the same bucket as like the young kids uploading to YouTube with the comment no copyright infringement intended thinking it's like plagiarism.
Can you elaborate on your view about "no copyright infringement intended" being related please?
> Copyright infringement use to be the absolute worst crime imaginable if you asked the tech industry
Every day hundreds of links to archive.is are posted[1] to this website to get around paywalls. Technologists built file sharing tools to subvert copyright. It has never been one of the worst crimes imaginable in tech circles.
[1] https://hn.algolia.com/?dateRange=pastWeek&page=0&prefix=tru... (85 instances in the last week)
You might be looking at a small time horizon.
What about https://en.wikipedia.org/wiki/An_Open_Letter_to_Hobbyists ?
Let's not forget the death of Aaron Swartz.
I don't think that was the "tech circles" being pro-copyright though.
I've never met someone who wasn't on Aaron's side on that one.
"Tech circles" was never the claim. The original phrase was "tech industry", and that seems to be accurate. The post replying to it may have misread or misinterpreted what "tech industry" means. (Or perhaps the term is simply ambiguous and each person who reads it comes away with a different meaning!)
> I've never met someone who wasn't on Aaron's side on that one.
This rather says something about the people by who you are surrounded. I know quite a lot of people who are on Aaron Swartz side here, for example people who are in academia or those who left academia but are still deeply interested in scientific topics.
You agree with them
Sorry, because I am not a native English speaker I misunderstood something. :-(
Jstor is a tech company?
Maybe the time horizon for a statement like that shouldn’t include the decades before most current tech companies existed, much less at this scale even for the few still kicking around from 50 years ago.
The tech industry? I don’t remember that being the case, at least not in general. Content owners yes— and there’s tech overlap there with Sony and some others. Beside that it’s never been a major hill for tech to die on, except in having to implement systems to deal with DMCA takedowns, and that only as half baked as they could get away with. Which unfortunately has meant “not in favor of users” when it comes to to failure modes and where and how to default actions.
> Copyright infringement use to be the absolute worst crime imaginable if you asked the tech industry
This is rather the opinion of the copyright-industrial complex, as Spivak implied in his comment (https://news.ycombinator.com/item?id=46874194) by referring to Hollywood.
The attitude of the tech industry has always been much more vague (example: Google Books), and people from the hacker culture, who often work as programmers, are traditionally rather sceptical of at least the concrete manifestation of the copyright system ("information wants to be free", circumvention of paywalls, Guerilla Open Access Manifesto, ...).
To be honest I don't care. Whatever helps the copyright industry burn is my ally.
I think this is part of a recurring pattern in tech of pushing boundaries around copyright.
In the last few years, we had Google scanning books, Google threatening to shut down News in Canada rather than pay publishers, LLMs summarizing articles on social platforms, crawlers bypassing paywalls, and so on.
Each time, the industry frames it as their interpretation of the current law, which were usually not written with these specific future use cases in mind.
In my view the current discussion regarding Gen AI is similar.
Hollywood sure, publishers sure, but tech? Where? The pro piracy pro game cracking open source p2p distribution people?
Basically a rage bait. If the law was bad, does it make it okay to violate it? In fact Anthropic is literally paying $1.5B on the copyright settlement, that indicates its completely a settled issue that AI companies have been violating this law. Some have been caught and fined, others are been lucky or that influence over the government.
> Copyright Law Was Built for Human Scale
No where in the law it has this kinda scoped limits. It has a time limit and scale doesnt not matter. Scale matter in a way that its gets harder to enforces buts that not the fault of copyright law. If you steal at a big scale, its still stealing.
> Basically a rage bait. If the law was bad, does it make it okay to violate it? In fact Anthropic is literally paying $1.5B on the copyright settlement, that indicates its completely a settled issue that AI companies have been violating this law. Some have been caught and fined, others are been lucky or that influence over the government.
Yes, but they were found not liable for copying the books they purchased. They were found liable for the books they torrented.
The former is something publishers still want to address
> If the law was bad, does it make it okay to violate it?
St. Augustine: "an unjust law is no law at all."
John Marshall, in Marbury v. Madison: "a law repugnant to the constitution is void."
This is actually a fairly well established principle in common law. So, yes.
> If the law was bad, does it make it okay to violate it?
Yes.
(I know these days it feels silly to bring this up, but...)
That is not how the separation of power is supposed to work. If a law is bad, politics (preferably a democratic process representing the people) replaces the law with a better one. Until the new law comes into effect, everyone is supposed to abide by the old law, even if it's bad.
Laws are created for lobbyist not people. Nothing works in the real world like in the text book democracy. Everything that was written about democracy is as naive as young adults fiction. But I find it refreshing that the real world events finally force more and more people to realize that.
Yet you, like most of us, probably grew up in a democratic society (?) with a relatively high degree of freedom.
As they say: democracy may not exist, but we'll miss it when it's gone.
Not really. I grew up in a socialist (communist?) society. A form of democracy closer to ideal than whatever US is doing for the last 2 centuries or more arrived in my teen years. My personal freedoms were mostly unaffected. Most treasured freedom I acquired was the freedom to gtfo which I'm happily using right now.
More than yes. It's commendable.
I think build for, here means was intended to work at or with human scale in mind. Not be limited to.
Laws usually don't describe the bigger societal context in which they were conceived.
IMHO AI generated content should be treated the same way with how human generated content and I don't see the problem. However as with technology the problem is a bit different, e.g.: When subletting your apartment requires manual effort, this is not a problem. Automated, it became an industry and that's a huge headache. I think this is the key point where the derived work has unlimited possibility that they want to curb it early on. In a way it's a fair effort to keep human's competitiveness but may prove to be futile.
> IMHO AI generated content should be treated the same way with how human generated content and I don't see the problem.
The problem is that AIs don't have rights. So you literally can't treat it the same.
If I "make" something in a way I'm not allowed to (e.g. copying), the law will go after me. If AI makes something I'm not allowed to, I can just say "whoopsie doodle", blame the AI, and there are no repercussions for anyone.
Always have been broken.
Hopefully, future legislation will cater less to publishers and copyright trolls. I'm not optimistic though. While certain kinds of publishers are indeed becoming less powerful, sports-related media conglomerates are successfully lobbying for more surveillance.
The general population will likely get the worst of both worlds, with copyright trolls getting to enforce unjust laws against regular people, while big tech gets to pay their way out.
Personally I feel that the excessive duration of copyright just weakens authors arguments against AI.
If even WWII-era documents are still under copyright, building a model respecting that would be impossible.
I would very much like to see us return to something resembling the original copyright terms. Like: Manually file for 22 years of enforcement. Then manually file 20 years later for 22 more.
You could create some great masterpiece at 18 and live off of it until you are 62 and starting to take social security payments.
> Personally I feel that the excessive duration of copyright just weakens authors arguments against AI.
That excess exists precisely because of the industry’s clout. For decades, rightsholders successfully lobbied Congress to extend copyright term again and again. That process appears to have finally plateaued, which is why early Mickey Mouse has now entered the public domain.
And notably, since the rise of AI, the government has not been especially quick or eager to step in and defend rightsholders.
> If even WWII-era documents are still under copyright, building a model respecting that would be impossible.
"We can't do this legally, so we should be allowed to ignore the law."
If you can't build a model while respecting licenses, don't build a model.
I don't want copyright to exist, at any duration, and I certainly think it should be much shorter than it is. However, as long as it exists, AI should not get any exception to it; such an exception inherently privileges massive entities over small entities or individuals.
If you can't do something which looks reasonable within the laws, it means that the laws have to change.
I don't see how building such a model of historical documents would impact authors (which most of them are dead anyways).
> If you can't do something which looks reasonable within the laws, it means that the laws have to change.
Then change the law. For everyone, consistently, not just massive AI companies. Until then, deal with it.
> I don't see how building such a model of historical documents would impact the authors
It impacts the authors of new documents, who do not get to copy those historical documents, while an AI competing with them gets to. If you want those historical documents to be freely available, make copyright stop applying to them. For everyone, not just massive AI companies.
I'm personally in favor of a heavily reduced copyright duration similar to patents and so if AI makes people realize how stupid the current laws look, I take it.
That'd be fine if AI waited for the law to change, rather than just breaking it and trying to grow fast enough to make countries scared to enforce their own laws against it.
The people campaigning for a change previously mostly were librarians, tech enthusiasts, archivists and pirate parties, now it's megacorporations worth billions of dollars.
I have a feeling that they are going to fare better odds, the world runs on money.
They might fare better odds at using their ill-gotten goods to lobby for changes that benefit them, and do not help others or are even harmful to others.
copyright is not “broken,” it was always a two-faced scam designed to protect “owners” at the expense of creators. don’t expect this blatant hypocrisy to kill copyright, either - death of copyright is a slippery slope leading straight to communist utopia, and the death of the global acqui-parasite class
laws that were already broken can still be broken. AI exposed how broken copyright law was. AI companies also broke (and continue to break) that law
I think the problem is that “transformation” has never been clearly defined in copyright law, and that ambiguity is exactly what AI companies lean on in their defense.
At a human scale, those boundaries get clarified through litigation on a case by case basis once an infringement becomes large or obvious enough. But there has always been a gray area around when a derivative work crosses the line into infringement.
And AI didn’t create that ambiguity, it just pushed it to a more extreme scale
Copyright law has always been excessively restrictive, and is long overdue for reform. The informal practices that people have been following (i.e. free creation and distribution but no monetization and no confusion with official releases) are pretty close to what a reasonable law would state.
Agreed on all of these points. I think that this will happen (...eventually)
> If you paint a picture of Sonic the Hedgehog in your living room, you are technically creating an unauthorized derivative work
Is this even true? It might violate a trademark, but I don't think it would violate copyright law unless it was a copy of an existing picture.
It is true but meaningless. Derivative works are not illegal, and you don’t need authorisation to create one.
So if you draw sonic in your living room you are indeed creating an unauthorised derivative work. And someone can call it an unauthorised derivative work. And the only reaction that should induce is raising an eyebrow and replying “ok?”
LLMs are often framed as lossy compression, and surely converting some copyrighted Sonic the Hedgehog image from PNG to JPEG is considered copyright infringement, no?!
Relevant: What Colour are your bits? https://ansuz.sooke.bc.ca/entry/23
Yeah, in some countries private stuff isn't illegal at all.
I don’t believe there’s any country where derivative work is intrinsically illegal. The Berne convention considers them protected same as originals.
It's not true. You don't need "authorization" to create a derivative work. You do need a license to distribute copyrighted works, including derivatives. And this only matters when you are distributing it to a sizable audience.
For example, if you rent a movie, you can watch it with your family. Nobody is going to sue you for distributing the movie with 5 people in your room. That's pure nonsense. Same with music, books, etc.
If you try to play the movie in an establishment with dozens of people, then it can become a problem, because you're essentially a theater now.
I'm not a lawyer so I don't know what the law is on selling fan art on a convention or even privately commissioned fan artwork. But things aren't as draconian as people assume it is.
UK law, and I assume other countries, specifically says that playing the movie to the public is a breach of copyright, not playing it to your family.
https://www.legislation.gov.uk/ukpga/1988/48/section/19