The copyright fight over authors and OpenAI's ChatGPT | Hacker News Reader

The copyright fight over authors and OpenAI's ChatGPT

vox.com

41 pointswebmaven55 comments

"The difference, when it comes to AI, is one of scale. ChatGPT can “read” more published words in a few seconds than I could in several lifetimes and, unlike me, that data isn’t immediately replaced in my human-limited short-term memory by whatever I’m thinking of next."

I think this misses the point. The issue of scale isn't on the ingest side, it's on the output side. Once you train an LLM on a book (however long that takes), then the LLM can be the interface to that book for an unlimited number of users. That scales very differently to, say, a person reading a book and writing something influenced by it.

In the case of the LLM, it's a complete interface to the contents of the book. It lets you "talk to the book". If that exists, why would anyone buy the book? If I could ask ChatGPT to "summarize the new book by XYZ", then spend an hour or two asking the questions _I_ have about the book from it, then buying the book would be a net negative.

If we don't solve attribution (like BMI solved for music), then the financial upside of publishing might be majority-captured by whoever trains LLMs on the copyrighted material.

By your argument, writing summaries of books, "explain <book> in 3 minutes" youtube videos, and commentaries on books should be made illegal too.

Or more precisely, they should be made illegal if and only if they achieve "scale" of maybe at least a couple million viewers.

The fundamental premise of copyright is flawed. Taking medieval concepts involving censorship of the printing press and extending them to the 21st century is bound to produce awkward results. I'm not hopeful that copyrights will be reconsidered from the ground up during this AI shock, but at least we shouldn't pretend that any arguments about copyrights should be reasonable and make sense. I honestly believe a "realpolitik" approach is more helpful, at least we know that those with more political influence and spend more effort lobbying will probably "win" in the end...

I appreciate the example, but here's where I think it differs as an analogy of what LLMs do:

A summary doesn't have infinite or variable depth. If you read the summary of a non-fiction (I'll limit my argument to that, as another poster pointed out) book, and either aren't convinced, or want to learn more about the matter, you'd have to purchase the book.

An LLM that has ben trained on the book, if somehow designed not to hallucinate, would be able to answer any question you have about the book at any depth, seamlessly blending in material from other books to answer a question or explain a concept. That seems like an entirely better experience than reading the book from start-to-finish. I don't see how the original can compete.

Assuming you know the right questions to ask. Most people don't know what they don't know. I've tried this. I'd prefer to pay a small amount to read the book.

Is that really that much of a barrier? Off the top of my head: you could start with a prompt like "write a summary of book XYZ, followed by a summary of each chapter". Then dive deeper into each one from there using the same prompt recursively, etc.

Yes, it’s a tremendous barrier, which is why academic fields tend to have introductions or surveys of the domain, stepwise instruction, with more in-depth or specialist knowledge being premised on having this understanding. I feel like this should be obvious to everyone? Did your education not follow such a progression?

Summaries are fair use.

> if somehow designed not to hallucinate

We know this isn't possible at the moment. Are we going to legislate for something that is not yet technologically possible? Should judges decide cases because maybe ML researchers will figure out how to reliably stop models from hallucinating?

This is a very basic and naive, poor scenario. Words are in public domain. But somehow their arrangements makes all the difference. Can AI just solve this "arrangement" problem better than humans do. Arrangements can be liked to series of moves in chess and AlphaGo solves for this through selfplay given only the rules.

LLMs will never be able to not hallucinate. Also, it's insane to me that people like you would prefer to ask a chatbot about a book rather than read the book itself. Part of the value of books is the voice of an author.

Under U.S. copyright law you can't copyright facts, only the artistic expression.

Unless we're talking about source code of course.

Even for source code, the US does not offer copyright to programs that are simple enough to have effectively one way to accomplish the desired function rather than requiring creative (aka artistic) choices by the programmer.

Example program specification for which the straightforward implementation in any common programming language would not be copyrightable by itself without adding additional scope: “When executed, output ‘Hello, world!’ plus a new line character to standard output, and then exit returning exit code 0.”

> Even for source code, the US does not offer copyright to programs that are simple enough to have effectively one way to accomplish the desired function

I think the fact you use the word "function" here is extremely telling. Writing code is obviously in a closer intellectual domain to designing a car engine, than it is to drawing a picture.

Maybe you have ground to stand on when talking about things like code golf which could be analogous to poetry. But no, the vast majority of code is not the product of artistic expression. It is the product of functional desires.

Not sure why you're trying to make an argument about trivial software. The same is true about trivial art: draw a black square on a white canvas. Good luck claiming copyright for that.

I disagree that the vast majority of code lacks artistic expression, especially when using the inclusive sense of the word “artistic” (or often “creative”) that the law uses to determine copyrightability.

There are so many different styles and designs when implementing any nontrivial underlying functional specification, and the preferences, choices, skill, and aesthetic of individual programmers definitely shine through. The ways you and I would find it straightforward to write a given program and the way I would write the same program are very probably recognizably different, beyond purely functional programs like the one I gave.

The existence of an underlying functional desire does not change the necessary artistic element in how to achieve that desire. Even in the traditional art world, an underlying functional desire is often more present than you think. Many artworks throughout history and even today are in fact commissioned, whether explicitly per-piece or through a patronage or employment relationship. A commissioned artwork is trying to satisfy either the specifications or the desires of the client. And among those which aren’t commissioned, like personal photographs, the underlying desire is often a functional one of remembering an occasion, despite the many clearly copyrightable artistic choices and skill required to create the work.

The black square on a white canvas example could very well be copyrightable, and I’d even guess that it usually is. Your functional specification still leaves the artist much freedom to choose the dimensions, relative positions and angles, exact shades of color, and materials of both the black square and the white canvas, as well as the shape of the canvas. Many ways to do it - and, importantly, no obvious one straightforward way to do it as there is in my trivial programming example.

> I disagree that the vast majority of code lacks artistic expression

I disagree that I made a claim that you're disagreeing with here. I said the vast majority of code is the product of functional desire. Building an engine is the product of functional desire. Building a birdhouse is the product of functional desire. Building a bridge is the product of a functional desire. Drawing a portrait is the product of artistic expression. All of these require creative thinking. One of them is copyrightable. Software is definitely closer in intellectual domain to what is not copyrightable than to what is.

I don't think copyright should exist at all, but I also 100% think you're kidding yourself if you think most software is an artistic endeavor.

> If we don't solve attribution

It can't be solved, by design. We want LLMs to behave naturally. Humans, naturally, don't provide any attribution, unless it really matters for the conversation.

No one (except for the copyright holders) wants LLMs to be a marketing department's dream, something straight out of cyberpunk novels, spewing brand names(tm) non-stop.

> then buying the book would be a net negative

Surely this is not true. At least for the fiction, people read books instead of their short summaries, because they want to spend time enjoying the story. That's why people are so against any spoilers.

> It lets you "talk to the book". If that exists, why would anyone buy the book?

Interactive and non-interactive experiences are two different things. Although, for sure, after a good book, I'd surely enjoy a "what-if" or "explain that" chat with an LLM (here, a possible business model for rightholders). But a chat cannot replace a story.

For a non-fiction, I probably might enjoy a brief summary first. That's why science papers start with an abstract, anticipating the reader's needs. But even then, if I'm interested, I will probably need full unabridged text to get into the exact details (without LLMs hallucinating me anything).

You're right. I think it's fair to carve out fiction from my argument. For that, I would surely go to the source material until the point where the LLM was coming up with better long-form fiction de-novo. But for non-fiction, which I would guess is the economically and intellectually more important category to protect, the effects may be devastating.

I also agree that attribution can't be solved easily in the current paradigm. Perhaps, during training, one could deduce how much of the net gradient on a particular weight was derived from the batches covering some book, and then during inference, assign attribution based on the effect of that weight on the output. All of this is very expensive to do, and I don't have strong intuitions for whether the resulting attributions would be in any way meaningful.

To your point about hallucinations, if there's not a solution to that, then perhaps the whole point is moot when, after a while, the hype dies down. But if somehow hallucinations are solved (I don't see a technical way this can happen now, but who knows?), then I think we'll need to address attribution for non-technical material.

My impression is that attribution on limited datasets isn't terribly hard. If you can prompt the LLM to say a sentence that is approximately in the source material, then the nearest sentence vector in the source material can be looked up in a vector DB, which can attribute it in context.

I think this might be one of the few places where LLMs can provide straightforward value, since it can work as a search engine that can accept vague queries, create approximate answers, fetch the real answers, translate the source material into layman's terms with citations, and allow the newly informed user to refine or dig deeper with that context. The most dangerous part is translation, and the data I've seen show that transformers almost never hallucinate on tasks where no external knowledge is needed.

I’m confused why you claim attribution is somehow “unnatural”? Every actually useful lecture, essay, report, etc. I’ve encountered included things like footnotes, references, or a bibliography. So much so, in fact, that I tend to disregard things that don’t include them. So-and-so claims X. What are their sources? There are none? Who cares. Life is too short to engage with arguments that lack rigor or support, even though these things themselves require verification!

Life is too short for me to engage with your argument, because you've failed to attribute the first writers of sentences / ideas semantically similar to each of the lines in your comment.

Ah, you’re right. My mistake. I should’ve simply claimed it’s natural to cite sources instead. After all, there is no debating what is natural or those who are simple.

My apologies, my perception of LLMs is somewhat skewed, because I primarily think of conversation agents.

It's unnatural in a conversation. When we're talking about, say, Superman, we don't ever say that it's "a registered trademark of DC Comics, Inc." With obligatory exceptions for comical or satirical effects, or if we're specifically talking about trademarks or copyrights, etc. And of course when we're talking about robots we don't normally give any nods to Karel Čapek.

I believe that, same as humans, LLMs already try to provide references when requested, or if the style/format (such as lecture) prompts for having them. Just remember that famous anecdote where a lawyer used ChatGPT and it wrote a speech and provided believable references (then judge threw this out of court because quality/reliability is another problem - which is out of scope, though).

> without LLMs hallucinating me anything

That's the trouble with LLMs. You cannot rely on what it is regurgitating.

> If that exists, why would anyone buy the book?

Because (if the book affords it), reading can be a form of psychic traveling. A reader enters an altered state of consciousness, lives in the world of the book, and comes back changed.

A summary of the information and 'plot points' would seem like a replacement only for those who have never really been absorbed in reading a book.

> why would anyone buy the book? If I could ask ChatGPT to "summarize

Summary is not the same as reading the book. Anyone can read reviews, human written summaries, on internet, or even some analyses, instead of asking AI model or buying a book. AI model usually cannot reproduce even small fragments. But it can indefinitely 'creatively' fantasize in books universe. Usually messing facts and mixing it all with other books.

By the way model doesn't have to be as big as ChatGPT. Anyone with good gaming GPU can get an open source free model and train it for academic research. Mixing fantasy with something else can produce interesting results.

> If I could ask ChatGPT to "summarize the new book by XYZ", then spend an hour or two asking the questions _I_ have about the book from it, then buying the book would be a net negative.

I believe you lose some data when doing so, summarizing is good when you want to get the gist of it, but not good when you want the actual details.

I know, this sounds very obvious but some people seriously jump to a summary directly and believe that is enough when they research.

This is personal for me, as I have a large dataset (near books3 sized) in my specialized domain that I tried to publish an NLP paper for. Excellent scores but I was too honest about the fact that some stuff in my evidence was gathered from paywalled sources. Despite excellent conference scores, the senior area chair intervened and they rejected the paper.

Seems that the current copyright fights are having a chilling effect in the context of research on language resources and datasets. This is extremely sad. I expected the academic AI world to have a bit more institutional compassion for copyright abolitionism…

Modern copyright law is so far behind the times and broken in more ways one can count. The only sensible way one can deal with it is to ignore it and keep their mouth shut.

Say for example, the Mistral AI approach. The released Mistral 7B and destroyed every other model in its class in performance while giving it a much less restrictive license. They gave zero info about what they trained on... and nobody is asking. A very smart move from both sides that benefits just about everyone.

The only winning move is not to play.

I suspect they’re covered under fair use (for now) but also nobody cares because the larger models are the ones that get used by people who would care enough to be upset about it.

Everyone claims fair use, but the terms for that are pretty narrow and wouldn't hold up anywhere afaik. Unless you're doing a news report or a review with actual criticism it probably won't pass the smell test. An LLM definitely doesn't fit that description in the slightest.

I maintain my position that despite whatever moral or ethical quandaries you feel are being violated, “no one cares”.

"Hey chief, I stole all this data but the scores are great!"

I think the entitlement to the work of others I've seen in software engineering generally has been so gross with regards to this technology. Copyright isn't just some antiquated idea from the middle ages, it's how artists and writers eat. They rightly spiked your paper.

Just a brief reminder of the last hundred years of copyright: large lobbies like UMG, Disney, or Elsevier have hugely circumscribed the public domain for all but historical purposes (copyright lengths increased from 28 years to *128*). Open access in academic publishing (for public funded research, even!) has been a decades-long fight that is not fully won. The DMCA has created myriad abuses and denied the freedom to control your own hardware, like allowing companies like John Deere to prevent right to repair. Six studios own most movies, a few labels control most music. The cultural heritage of the last hundred years, which should be the collective property of humanity, is locked up in the coffers of a few megacorps: who, by the way, don't much care how well their artists are eating. Yes, there are edge cases, but spare me the sympathy for the poor artist on Fiverr with five images in LAION: this is a fight between tech titans and copyright monopolists, where the little people are always going to be an afterthought. Come out and say the original copyright term of 28 years (well long enough to eat!) should be brought back, condemn the DMCA and the naked abuse of democracy in the Sonny Bono act, and THEN maybe we can get to chiding the lapses of a few researchers just wanting to understand the world better.

I don't support tech titans or copyright monopolists and mostly just think stuff should enter the public domain earlier as that is what both groups don't want to have happen. And in the case of software I think it ought to be illegal to not release source code into the public domain after a reasonable amount of time, say 25 years. In fact I think it's absolutely crazy we allow large companies to sell infrastructure with software in it and then basically light the source code on fire so that people are forced to buy new versions of their product rather than maintain the source for the old version themselves.

Artists who perform earn money by performing not by restricting others from re-viewing a previous performance.

Artist who create unique pieces are not worried about re-prints.

Writers who connect get followers who will pay for there next word.

Artists in the past needed a patron and we have providers today who do that

Most of the actual artists I've spoken to really hate this technology.

Just for the record, the domain I'm working in is educational (like literally afterschool teacher chaperone kind of thing). They have had a long tradition of open sourcing their data going back literally decades in word documents, and before computers, ripping stuff out of books. The dataset I contributed was doing nothing more than taking these word documents and parsing the data out and making it easy for NLP systems to use it. The actual dataset creation/curation is a crowd sourced effort from high school and college students, alongside their coaches that has been ongoing for decades.

Besides the clear fair use exemption due to the educational nature of the activity it targets, it's also an activity that a significant amount of lawyers/judges in the legal system participated in. For that reason, I think that there's very little risk of the dataset being considered "actually infringing". If it were litigated, and this dataset were deemed illegal, than the entire activity itself it targets would be shattered. Killing that activity would meaningfully and significantly harm the lawyer pipeline in the USA.

But, I can also tell you that I think the majority of people should share their creative things for access to all. I have very little sympathy for artists or writers who are scared of LLMs because LLMs are coming for coders at literally the exact same time, but coders have been open source fanatics ideologically for awhile. It's time for writers and artists to embrace open source. Particularly since so many academic intellectuals have left-wing/socialist sensibilities anyway - it's time for them to practice what they preach.

If I read a book, adding its contents to my brain, and then having that content influence what I write on HN, is that a copyright violation?

If you say you are retelling the story and then are writing down the story chapter for chapter, yes. That is what LLMs can do, they don't write something inspired by it they write down a copy of it with switched out words and grammar structure.

It doesn't matter if you personally aren't using them like that, the fact that they can be used like that is the problem.

>That is what LLMs can do, they don't write something inspired by it they write down a copy of it with switched out words and grammar structure.

That's not at all what they do. This myth of "supercharged interpolation" really needs to die. https://arxiv.org/abs/2110.09485

>It doesn't matter if you personally aren't using them like that, the fact that they can be used like that is the problem.

Of course it matters. You penalize people when they have done something wrong not if they have the capacity to.

Let's not pretend copyright and intellectual property is somehow grounded in reality or logic. IP laws will have to get updated for generative AI - what that looks like will be determined by interest groups in play and legislators.

I’d say no because comparing a human brain to chatGPT isn’t really fair nor am I sure they work in identical way. chatGPT can and have spat out raw content it was trained on.

You understand the words you read in a book, you get the message out of it and are able to resonate and have thoughts. ChatGPT doesn’t, it just scrambles words based on statistics.

You, writing content based on your influence should not be counted as copyright violation in my opinion. But giving some credit isn’t a bad practice either.

If you did copy and paste the contents of the book and call it yours, then that should be counted as copyright violation (there are probably exceptions to this).

I think what some people are currently against is openai (closed ai) scraping all kinds of content on the web and build a fantastic product like chatgpt without compensation or giving credit. They have essentially built an amazing product on the shoulders of lots of creators without them earning anything from it.

Is that fair? I don’t think so. But that’s just me.

This might be a bit irrelevant, but Github training their ai on repositories without consent or consideration of their license wasn’t cool either.

You don't know enough about your own brain to say gpt is using different methods.

> You understand the words you read in a book,

Gpt understands better cause it can answer better when I ask about it.

> you get the message out of it and are able to resonate and

The bot can do this too. At least it can pretend to. We pretend too :)

> have thoughts

As in have hallucinations.

It is very cool. Don't put your code out in the open and then blame others (inc ai) for reading it and training themselves.

Reading is training. For human and ai.

> You don't know enough about your own brain to say gpt is using different methods.

I said I wasn’t sure they worked in the same way, I did not state it as a fact.

> Gpt understands better cause it can answer better when I ask about it.

Assuming Gpt understands because it is able to answer you doesn’t seem logical to me.

If I understand Gpt correctly, Gpt does not understand anything it says, it is just exceptionally good at convincing you it does.

I don’t think thoughts and hallucinations are identical things.

> Don't put your code out in the open and then blame others (inc ai) for reading it and training themselves.

So licenses doesn’t matter at all if its out in the open?

Surely it depends on how you obtained that book?

so humanity has to get access to less information because 1 guy's feelings got hurt for not being mentioned in a field where they were not the only contributor

AI will win. Just like Napster liberated music, AI will liberate knowledge. Only problem is those creating AI now. It's a silent arms race. Tempered by global interest rates. Who will win?

I'm not entirely sure if I would agree with what you say about Napster. While for a while what you said was true, these days we're back to most people listening to music in walled gardens like Spotify, Apple Music, etc.

The lesson of walled gardens is that people will pay for curation. That’s endlessly irritating to some people, but that doesn’t change the truth of it.

People will pay more for convenience. A music streaming service with access to millions of songs is the same price per month of what a single artist's CD was 20 years ago. Sure, you don't have a cool physical copy, but it's incredibly convenient to search for any song in seconds, play it instantly, and not worry about storage, etc.

And I don’t have to download it, copy it to my phone, make sure the mp3 metadata is correct for the song, and I can view the lyrics, and easily share the song with a friend, and make a playlist. I have a 50GB collection of MP3s on my desktop from my teens and 20s that have been collecting dust because I just use Spotify because it’s convenient.

Spotify was created in a market in which piracy was widespread. I'm not sure we can assume it would have ever been a thing otherwise.

Spotify/Steam/Netflix only exist in the form they do because piracy is such a compelling alternative.

Napster was an opinionated (political even) piece of tech built for a specific purpose.

AI is not like that. And even if ChatGPT currently gives you something nice, it’s not controlled by you and can be nerfed and extort you.

You’d need to make an argument that AI will be good for people. It’s not self evident.

Reply on news.ycombinator.com