Not 42% of the book.
It's a pretty big distinction.
What is the distinction between understanding and memorization? What is the chance that understanding results in memorization (may be in case of humans)?
It should break copyright laws as written now but too much money involved.
not just next token.
This is like: tell it a random sentence in the book, it will give you the next sentence 42% of time.
Guess the next word: Not all heros wear _____
https://en.wikipedia.org/wiki/Artificial_intelligence_and_co...
> See for example OpenAI's comment in the year of GPT-2's release: OpenAI (2019). Comment Regarding Request for Comments on Intellectual Property Protection for Artificial Intelligence Innovation (PDF) (Report). United States Patent and Trademark Office. p. 9. PTO–C–2019–0038. “Well-constructed AI systems generally do not regenerate, in any nontrivial portion, unaltered data from any particular work in their training corpus”
https://copyrightalliance.org/kadrey-v-meta-hearing/
> During the hearing, Judge Chhabria said that he would not take into account AI licensing markets when considering market harm under the fourth factor, indicating that AI licensing is too “circular.” What he meant is that if AI training qualifies as fair use, then there is no need to license and therefore no harmful market effect.
I know this is arguing against the point that this copyright lobbyist is making, but I hope so much that this is the case. The “if you sample, you must license” precedent was bad, and it was an unfair taking from the commons by copyright holders, imo.
The paper this post is referencing is freely available:
Could it be plausible that an LLM had ingested parts of the book via scrapping web pages like this and not the full copyrighted book and get results similar to those of the linked study?
[1] https://www.goodreads.com/work/quotes/4640799-harry-potter-a...
[2] ~30 portions x 68 pages
https://www.reddit.com/r/DataHoarder/comments/1entowq/i_made...
https://github.com/shloop/google-book-scraper
The fact that Meta torrented Books3 and other datasets seems to be by self-admission by Meta employees who performed the work and/or oversaw those who themselves did the work, so that is not really under dispute or ambiguous.
https://torrentfreak.com/meta-admits-use-of-pirated-book-dat...
The pictures are the same. All roads lead to Rome, so they say.
They also use data from the previous models, so I'm not sure how "clean" it really is
Which of the major commercial models discloses its dataset? Or are you just trusting some unfalsifiable self-serving PR characterization?
https://en.wikipedia.org/wiki/Llama_(language_model)#Leak
I suspect that it was a limited hangout self-own by Meta to claim that they aren’t responsible, and then they are doing research on a leaked LLM that they developed, but then was leaked, so they can claim that the subsequent research is not tainted by the fruit of the poisonous tree legal doctrine. Or, their torrent client or other software on the same machine had 0-days and they got hacked by someone on the Books3 swarm or knowledgeable of what IPs were connecting to it.
I appreciate your posts and I am replying to you to humbly ask you to post more. :P
If that is what you are asking, I don't think that's what happened. It's far more likely that it was just leaked or grabbed by a hacker
https://www.wired.com/story/new-documents-unredacted-meta-co...
While the Harry Potter series may be fun reading, it doesn't provide information about anything that isn't better covered elsewhere. Leave Harry Potter for a different "Harry Potter LLM".
Train scientific LLMs to the level of a good early 20th century English major and then use science texts and research papers for the remainder.
It has copyright implications - if Claude can recollect 42% of a copyrighted product without attribution or royalties, how did Anthropic train it?
> Train scientific LLMs to the level of a good early 20th century English major and then use science texts and research papers for the remainder
Plenty of in-stealth companies approaching LLMs via this approach ;)
For those of us who studied the natural sciences and CS in the 2000s and early 2010s, there was a bit of a trend where certain PIs would simply translate German and Russian papers from the early-to-mid 20th century and attribute them to themselves in fields like CS (especially in what became ML).
Should it be? Different question.
First of all, we don't really know how the brain works. I get that you're being a snarky physicalist, but there's plenty of substance dualists, panpsychsts, etc. out there. So, some might say, this is a reductive description of what happens in our brains.
Second of all, yes, if you tried to publish Harry Potter (even if it was from memory), you would get in trouble for copyright violation.
My question is… is that in itself a violation of copyright?
If not then as long as LLMs don’t make a publication it shouldn’t be a copyright violation right? Because we don’t understand how it’s encoded in LLMs either. It is literally the same concept.
If you compressed a copy of HP as a .rar, you couldn't read that as is, but you could press a button and get HP out of it. To distribute that .rar would clearly be a copyright violation.
Likewise, you can't read whatever of HP exists in the LLM model directly, but you seemingly can press a bunch of buttons and get parts of it out. For some models, maybe you can get the entire thing. And I'm guessing you could train a model whose purpose is to output HP verbatim and get the book out of it as easily as de-compressing a .rar.
So, the question in my mind is, how similar is distributing the LLM model, or giving access to it, to distributing a .rar of HP. There's likely a spectrum of answers depending on the LLM
I can record myself reciting the full Harry Potter book then distribute it on YouTube.
Could do the exact same thing with an LLM. The potential for distribution exists in both cases. Why is one illegal and the other not?
Not legally you can't. Both of your examples are copyright violations
At this point you've created an entirely new copy in an audio/visual digital format and took the steps to make it available to the masses. This would almost certainly cross the line into violating copyright laws.
> Could do the exact same thing with an LLM. The potential for distribution exists in both cases. Why is one illegal and the other not?
To my knowledge, the legality of LLMs are still being tested in the courts, like in the NYT vs Microsoft/OpenAI lawsuit. But your video copy and distribution on YouTube would be much more similar to how LLMs are being used than your initial example of reading and memorizing HP just by yourself.
if you trained an LLM on real copyrighted data, benchmarked it, wrote up a report, and then destroyed the weight, that's transformative use and legal in most places.
if you then put up that gguf on HuggingFace for anyone to download and enjoy, well... IANAL. But maybe that's a bit questionable, especially long term.
Personally I’m assuming the worst.
That being said, Harry Potter was such a big cultural phenomenon that I wonder to what degree might one actually be able to reconstruct the books based solely on publicly accessible derivative material.
To address this point, and not other concerns: the benefits would be (1) pop culture knowledge and (2) having a variety of styles of edited/reasonably good-quality prose.
> the paper estimates that Llama 3.1 70B has memorized 42 percent of the first Harry Potter book well enough to reproduce 50-token excerpts at least half the time
As I understand it, it means if you prompt it with some actual context from a specific subset that is 42% of the book, it completes it with 50 tokens from the book, 50% of the time.
So 50 tokens is not really very much, it's basically a sentence or two. Such a small amount would probably generally fall under fair use on its own. To allege a true copyright violation you'd still need to show that you can chain those together or use some other method to build actual substantial portions of the book. And if it only gets it right 50% of the time, that seems like it would be very hard to do with high fidelity.
Having said all that, what is really interesting is how different the latest Llama 70b is from previous versions. It does suggest that Meta maybe got a bit desperate and started over-training on certain materials that greatly increased its direct recall behaviour.
That’s what I was thinking as I read the methodology.
If they dropped the same prompt fragment into Google (or any search engine) how often would they get the next 50 tokens worth of text returned in the search results summaries?
It sounds like a ridiculous way to measure it. Producing 50-token excerpts absolutely doesn't translate to "recall X percent of Harry Potter" for me.
(Edit: I read this article. Nothing burger if its interpretation of the original paper is correct.)
To clarify, they look at the probability a model will produce a verbatim 50-token excerpt given the preceding 50 tokens. They evaluate this for all sequences in the book using a sliding window of 10 characters (NB: not tokens). Sequences from Harry Potter have substantially higher probabilities of being reproduced than sequences from less well-known books.
Whether this is "recall" is, of course, one of those tricky semantic arguments we have yet to settle when it comes to LLMs.
Sure. But imagine this: In a hypothetical world where LLMs never ever exist, I tell you that I can recall 42 percent of the first Harry Potter book. What would you assume I can do?
It's definitely not "this guy can predict next 10 characters with 50% accuracy."
Of course the semantic of 'recall' isn't the point of this article. The point is that Harry Potter was in the training set. But I still think it's a nothing burger. It would be very weird to assume Llama was trained on copyright-free materials only. And afaik there isn't a legal precedent saying training on copyrighted materials is illegal.
It can produce the next sentence or two, but I suspect it can’t reproduce anything like the whole text. If you were to recursively ask for the next 50 tokens, the first time it’s wrong the output would probably cease matching because you fed it not-Harry-Potter.
It seems like chopping Harry Potter up into 2 sentences at a time on post it’s and tossing those in the air. It does contain Harry Potter, in a way, but without the structure is it actually Harry Potter?
Consider e.g.:
- Digital expansion of PI to sufficient decimal places contains both parts of the work and full work in full. The trick is you have to know where to find it - and it's that knowledge that's actually equivalent to the work itself.
- Any kind of compression that uses a dictionary that's separate from the compressed artifact, shifts some of the information into a dictionary file, or if it's a common dictionary, into compressor/decompressor itself.
In the case from the study, the experimenter actually has to supply most of the information required to pull Harry Potter out of the model - they need to make specific prompts with quotes from the book, and then observe which logits correspond to the actual continuation of those quotes. The experimenter is doing information-loaded selection multiple times: at prompting, and at identifying logits. This by itself doesn't really prove the model memorized the book, only just that it saw fragments from it - in cases those fragments are book-specific (e.g. using proper names from the HP world) instead of generic English sentences.
Also they went through the book at 10 token strides. Like..a bit tortured way to reproduce the book (basically impossible to actually reproduce the book) but it shows that the content is in there.
Now whether this is derivative work, copyright violation or whatever is debatable. Probably gets similar numbers for a bunch of other books too. They should have done the Bible and probably get way higher numbers, but that won’t go viral.
Honestly, I get why these debates happen—it is practical to establish whether or not this emerging tech is illegal under current law. But it’s also like… well, obviously current law wasn’t written with this sort of application in mind.
Whether or not we think LLMs are basically good or bad, they are clearly quite impactful. It would be a nice time to have a functional legislature to address this directly.
Generally speaking, exceptions to copyright are based on the appropriateness of the amount of copied content for the given allowed use, so the shorter it is, the more likely it is for copying to be permitted. European copyright law isn't much different from fair use in that respect.
Where it does differ is that the allowed uses are more explicitly enumerated. So Meta would have to argue e.g. based on the exception for scientific works specifically, rather than more general principles.
There's also the question of how many bits of originality there actually are in Harry Potter. If trained strictly on text up to the publishing of the first book, how well would it compress it?
EDIT Actually, on rereading, I see I replied to the wrong comment.
This does not appear to happen with other models they tested to the same degree
According to the stated methodology, I could give the LLM sentence 1 and have 42% chance of getting sentence 2 recalled. Then I could give it sentence 2 and have 42% chance of getting sentence 3. Therefore, the LLM contains 42% of the book in some sense.
I disagree this is "not really very much". If a person could do this you would undoubtedly conclude that the person read the book.
In fact the number 42% even understates the severity of the matter. Superficially it makes it sound that the LLM only contains less than half of the book. In reality the process I described applies to 100% of the sentences. Additionally I'm guessing that the 58% times where the 50 tokens arent recalled correctly, the outputted token probably have the same meaning as the correct one.
This doesn't tell us for certain whether or not the model was trained on a full copy of the book. It's possible that 50-token long passages from 42% of the book were, incidentally, quoted verbatim in various parts of the training data. Considering the popularity of both the book itself, and derivative fan-fiction, I would not be surprised. I would be less surprised to learn that it was indeed trained on a full copy of the book, if not several.
The more meaningful point here is that the ability to reproduce half a book is the same sort of overt derivative work that is definitely considered copyright infringement in other circumstances. A lossy copy is still a copy. If we are to hold LLMs to the same standard as other content, this isn't very easy to defend.
Personally, I see this as a good opportunity to reevaluate copyright on the whole. I think we would be better off without it.
I’m personally more in favor of significantly reducing the length of the copy right. I think 20-30 years is an interesting range. Artist get roughly a career length of time to profit off their creations, but there is much less incentive for major corporations to buy and horde IP.
At the moment, there's also a huge difference between who does and who doesn't pay. If I put the HP collection on my website, you betcha Joanne Rowling's team is going to try to take it down. However, because OpenAI designed an AI system where content cannot be removed from its knowledge base and because their pockets are lined with cash for lawyers, it's practically free to violate whatever copyright rules it wants.
As a full-time professional musician, I'm convinced I'll benefit much more from its deprecation than continuing to flog it into posterity. I don't think I know any musicians who believe that IP is career-relevant for them at this point.
(Granted, I play bluegrass, which has never fit into the copyright model of music in the first place)
It's sold 120 million copies over 30 years. I've gotta think literally every passage is quoted online somewhere else a bunch of times. You could probably stitch together the full book quote-by-quote.
LLMs have limited capacity to memorize, under ~4 bits per parameter[1][2], and are trained on terabytes of data. It's physically impossible for them to memorize everything they're trained on. The model memorized chunks of Harry Potter not just because it was directly trained on the whole book, which the article also alludes to:
> For example, the researchers found that Llama 3.1 70B only memorized 0.13 percent of Sandman Slim, a 2009 novel by author Richard Kadrey. That’s a tiny fraction of the 42 percent figure for Harry Potter.
In case it isn't obvious, both Harry Potter and Sandman Slim are parts of books3 dataset.
[1] -- https://arxiv.org/abs/2505.24832 [2] -- https://arxiv.org/abs/2404.05405
https://www.theguardian.com/technology/2025/jan/10/mark-zuck...
Sure there are just ~75,000 words in HP1, and there are probably many times that amount in direct quotes online. However the quotes aren’t even distributed across the entire text. For every quote of charming the snake in a zoo there will be a thousand “you’re a wizard harry”, and those are two prominent plot points.
I suspect the least popular of all direct quotes from HP1 aren’t using the quotes in fair use, and are just replicating large sections of the novel.
Or maybe it really is just so popular that super nerds have quoted the entire novel arguing about the aspects of wand making, or the contents of every lecture.
archiveofourown.org has 500 thousand, some, but probably not the majority, of that are duplicated from fanfiction.net. 37 thousand of these are over 40 thousand words.
I.e. harry potter and its derivatives presumably appear a million times in the training set, and its hard to imagine a model that could discuss this cultural phenomena well without knowing quite a bit about the source material.
> Or maybe Meta added third-party sources—such as online Harry Potter fan forums, consumer book reviews, or student book reports—that included quotes from Harry Potter and other popular books.
> “If it were citations and quotations, you'd expect it to concentrate around a few popular things that everyone quotes or talks about,” Lemley said. The fact that Llama 3 memorized almost half the book suggests that the entire text was well represented in the training data.
And yes, I read the article before commenting. I don't appreciate the baseless insinuation to the contrary.
It's essentially the same thing, they are copying from a source that is violating copyright, whether that's a pirated book directly or a pirated book via fanficton.
Is this specific fact required to make my beliefs consistent... Yes I think it is, but if you disagree with me in other ways it might not be important to your beliefs.
Legally (note: not a lawyer) I'm generally of the opinion that
A) Torrenting these books was probably copyright infringement on Meta's part. They should have done so legally by scanning lawfully acquired copies like Google did with Google Books.
B) Everything else here that Meta did falls under the fair use and de minimis exceptions to copyrights prohibition on copying copyrighted works without a license.
And if it was copying significant amounts of a work that appeared only once in its training set into the model the de minimis argument would fall apart.
Morally I'm of the opinion that copyright law's prohibition on deeply interacting with our cultural artifacts by creating derivative works is incredibly unfair and bad for society. This extends to a belief that the communities that do this should not be excluded from technological developments because there entire existence is unjustly outlawed.
Incidentally I don't believe that browsing a site that complies with the DMCA and viewing what it lawfully serves you constitutes piracy, so I can't agree with your characterization of events either. The fanfiction was not pirated just because it was likely unlawful to produce in the US.
Accusations of not reading the article are fair when someone brings up a “related” anecdote that was in the article. It’s not fair when someone is just disagreeing.
- the first result is a pdf of the full book
- the second result is a txt of the full book
- the third result is a pdf of the complete harry potter collection
- the fourth result is a txt of the full book (hosted on github funny enough)
Further down there are similar copies from the internet archive and dozens of other sites. All in the first 2-3 pages.
I get that copyright is a problem, but let's not pretend that an LLM that autocompletes a couple lines from harry potter with 50% accuracy is some massive new avenue to piracy. No one is using this as a substitute for buying the book.
Well, luckily the article points out what people are actually alleging:
> There are actually three distinct theories of how training a model on copyrighted works could infringe copyright:
> Training on a copyrighted work is inherently infringing because the training process involves making a digital copy of the work.
> The training process copies information from the training data into the model, making the model a derivative work under copyright law.
> Infringement occurs when a model generates (portions of) a copyrighted work.
None of those claim that these models are a substitute to buying the books. That's not what the plaintiffs are alleging. Infringing on a copyright is not only a matter of privacy (piracy is one of many ways to infringe copyright)
Is that fair use, or is that compression of the verbatim source?
You can't in particular iterate it sentence by sentence; you're unlikely to go past sentence 2 this way before it starts giving you back it's own ideas.
The whole thing is a sleigh of hand, basically. There's 42% of the book there, in tiny pieces, which you can only identify if you know what you're looking for. The model itself does not.
No, it really couldn't. In fact, it's very persuasive evidence that Llama is straight up violating copyright.
It would be one thing to be able to "predict" a paragraph or two. It's another thing entirely to be able to predict 42% of a book that is several hundred pages long.
Copyright takes into account the use for such the copying is done. Commercial use will almost always be treated as not fair use, with limited exceptions.
Copyright is quite literally about the right to control the creation and distribution of copies.
The creation of the unzipped file is not treated as a separate copy so the recipient would not be violating copyright just by unzipping the file you provided.
Repeat for every copyrighted work and you end up with publishers reasonably arguing meta would not be able to produce their LLM without copyrighted work, which they did not pay for.
It's an argument for the courts, of course.
There is no morale and justice ground to leverage on when the system is designed to create wealth bottleneck toward a few recipients.
Harry Potter is a great piece of artistic work, and it's nice that her author could make her way out of a precarious position. But not having anyone in such a situation in the first place would be what a great society should strive to produce.
Rowling already received more than all she needs to thrive I guess. I'm confident that there are plenty of other talented authors out there that will never have such a broad avenue of attention grabbing, which is okay. But that they are stuck in terrible economical situations is not okay.
The copyright loto, or the startup loto are not that much different than the standard loto, they just put so much pression on the player that they get stuck in the narrative that merit for hard efforts is the key component for the gained wealth.
First-order systems drive outcomes. "Did it make money?" "Did it increase engagement?" "Did it scale?" These are tight, local feedback loops. They work because they close quickly and map directly to incentives. But they also hide a deeper danger: they optimize without questioning what optimization does to the world that contains it.
Second-order cybernetics reason about systems. It doesn’t ask, "Did I succeed?" It asks, "What does it mean to define success this way?" "Is the goal worthy?"
That’s where capital breaks.
Capitalism is not simply incapable of reflection. In fact, it's structured to ignore it. It has no native interest in what emerges from its aggregated behaviors unless those emergent properties threaten the throughput of capital itself. It isn't designed to ask, "What kind of society results from a thousand locally rational decisions?" It asks, "Is this change going to make more or less money?"
It's like driving by watching only the fuel gauge. Not speed, not trajectory, or whether the destination is the right one. Just how efficiently you’re burning gas. The system is blind to everything but its goal. What looks like success in the short term can be, and often is, a long-term act of self-destruction.
Take copyright. Every individual rule, term length, exclusivity, royalty, can be justified. Each sounds fair on its own. But collectively, they produce extreme wealth concentration, barriers to creative participation, and a cultural hellscape. Not because anyone intended that, but because the emergent structure rewards enclosure over openness, hoarding over sharing, monopoly over multiplicity.
That’s not a bug. That's what systems do when you optimize only at the first-order level. And because capital evaluates systems solely by their extractive capacity, it treats this emergent behavior not as misalignment but as a feature. It canonizes the consequences.
A second-order system would account for the result by asking, "Is this the kind of world we want to live in?" It would recognize that wealth generated without regard to distribution warps everything it touches: art, technology, ecology, and relationships.
Capitalism, as it currently exists, is not wise. It does not grow in understanding. It does not self-correct toward justice. It self-replicates. Cleverly, efficiently, with brutal resilience. It's emergently misaligned and no one is powerful enough to stop it.
it conjures up pictures of two dragons fighting each other instead of attacking us, but make no mistake they are only fighting for the right to attack us. whoever wins is coming for us afterwards
1. Strong copyright to prevent competition from undercutting their related businesses.
2. Exclusive rights to totally ignore the copyright of everyone that made the content they use to train models.
I personally would much prefer we take the opportunity to abolish copyright entirely: for everyone, not just a handful of corporations. If derivative work is so valuable to our society (I believe it is), then I should be free to derive NVIDIA's GPU drivers without permission.
Those are completely different phenomena. Removing copyright will not suddenly open the floodgates of creativity because anyone can already create anything.
But - and this is the key point - most work is me-too derivative anyway. See for example the flood of magic school novels which were clearly loosely derivative of Harry Potter.
Same with me-too novels in romantasy. Dystopian fiction. Graphic novels. Painted art. Music.
It's all hugely derivative, with most people making work that is clearly and directly derivative of other work.
Copyright doesn't stop this, because as a minimum requirement for creative work, it forces it to be different enough.
You can't directly copy Harry Potter, but if you create your own magic school story with some similar-ish but different-enough characters and add dragons or something you're fine.
In fact under capitalism it is much harder to sell original work than to sell derivative work. Capitalism enforces exactly this kind of me-too creative staleness, because different-enough work based on an original success is less of a risk than completely original work.
Copyright is - ironically - one of the few positive factors that makes originality worthwhile. You still have to take the risk, but if the risk succeeds it provides some rewards and protections against direct literal plagiarism and copying that wouldn't exist without it.
At the intersection of capitalism and copyright, I see a serious problem. Collaboration is encapsulated by competition. Because simple derivative work is illegal, all collaboration must be done in teams. Copyright defines every work of art as an island, whose value is not the art itself, but the moat that surrounds it. It should be no surprise that giant anticompetitive corporations reflect this structure. The core value of copyright is not creativity: it's rent-seeking.
Without copyright, we could collaborate freely. Our work would not be required to compete at all! Instead of victory over others' work, our goal could be success!
Banning the nonsense that is character copyright and shortening copyright back down to a reasonable length of time (say, 20 years) would still enable the creation of more culturally-relevant derivative works without pauperizing every artist.
Even British Colonial America had no copyright, save a handful of exceptions, as the Statute of Anne did not apply to the colonies.
>Capitalism is not simply incapable of reflection. In fact, it's structured to ignore it. It has no native interest in what emerges from its aggregated behaviors unless those emergent properties threaten the throughput of capital itself. It isn't designed to ask, "What kind of society results from a thousand locally rational decisions?" It asks, "Is this change going to make more or less money?"
Capitalism and free market has lot of useful and emergent properties that occur not at the first order but second order.
> In the case of the global economic system, under capitalism, growth, accumulation and innovation can be considered emergent processes where not only does technological processes sustain growth, but growth becomes the source of further innovations in a recursive, self-expanding spiral. In this sense, the exponential trend of the growth curve reveals the presence of a long-term positive feedback among growth, accumulation, and innovation; and the emergence of new structures and institutions connected to the multi-scale process of growth
https://en.wikipedia.org/wiki/Emergence
In fact free market is an extremely good example of emergence or second order systems where each individual works selfishly but produces a second order effect of driving growth for everyone - something that is definitely preferable.
That's the core problem. Capitalism resists second-order critique from within because it translates every possible value: justice, meaning, even critique itself, into terms it can price or optimize. Your response is a perfect example: you defend capitalism by listing its outputs, but that;s another first-order move. If you were engaging at the second-order level, you'd interrogate not what the system produces, but what it refuses to ask, and who gets to decide. That silence is precisely my point.
I did not claim it as inherently good, only that it is preferable.
> capitalism systematically collapses higher-order questions about what kind of world were building into first-order value propositions like "growth," "utility," and "innovation."
There is nothing about capitalism that ignores second or third order effects of its policies. Let me make it clear what kind of capitalist system we have in place - private ownership and free market regulated by a government that works for and is elected by the people. In this system the free market works but only till it progresses certain things the people voted for like standard of living, freedom etc. If free market does instead has unintended consequences we have levers to guide it where we want like taxes and subsidies.
> Capitalism resists second-order critique from within because it translates every possible value: justice, meaning, even critique itself, into terms it can price or optimize
I think I see were you are getting at but I have to be honest - I think it is coming from a naive place (I'm open to be proven incorrect).
Imagine you had the power and the responsibility to shape lives by enacting policy decisions. You are presented with a fairly complex problem where you have a large number of people, each one with their own lives and interests and you have to guide them into doing something preferable. No matter where you come from, left or right in the political axis, you will end up using quantitative methods. I imagine your problem is with such optimisation. If so, what is your exact critique here? How would you rather handle such a situation? How would you manage a system of so many people and without quantitative method? Religion?
> If you were engaging at the second-order level, you'd interrogate not what the system produces, but what it refuses to ask, and who gets to decide. That silence is precisely my point.
Ok please elaborate (only if you have engaged with my question above).
>collapses higher-order questions about what kind of world were building into first-order value
But then
>you'd interrogate not what the system produces, but what it refuses to ask, and who gets to decide. That silence is precisely my point.
The reply your interlocutor provided is aligned with your incoherence. In a first move you point out that capitalism flattens everything into first-order land, and yet in a second move you tell us there are things it can't talk about. I guess your silence is precisely what articulates these two aspects of your discourse.
Another key point is that you might download a Llama model and implicitly get a ton of copyright-protected content. Versus with a search engine you’re just connected to the source making it available.
And would the LLM deter a full purchase? If the LLM gives you your fill for free, then maybe yes. Or, maybe it’s more like a 30-second preview of a hit single, which converts into a $20 purchase of the full album. Best to sue the LLM provider today and then you can get some color on the actual consumer impact through legal discovery or similar means.
Music artists get in trouble for using more than a sample without permission — imagine if they just used 45% of a whole song instead…
I’m amazed AI companies haven’t been sued to oblivion yet.
This utter stupidity only continues because we named a collection of matrices “Artificial Intelligence” and somehow treat it as if it were a sentient pet.
Amassing troves of copyrighted works illegally into a ZIP file wouldn’t be allowed. The fact that the meaning was compressed using “Math” makes everyone stop thinking because they don’t understand “Math”.
LLMs are in reality the artifacts of lossy compression of significant chunks of all of the text ever produced by humanity. The "lossy" quality makes them able to predict new text "accurately" as a result.
>compressed using “Math”
This is every compression algorithm.
What's the work here? If it's the output of the LLM, you have to feed in the entire book to make it output half a book so on an ethical level I'd say it's not an issue. If you start with a few sentences, you'll get back less than you put in.
If the work is the LLM itself, something you don't distribute is much less affected by copyright. Go ahead and play entire songs by other artists during your jam sessions.
A ZIP file of a book is also in direct competition of the book, because you could open the ZIP file and read it instead of the book.
A model that can take 50 tokens and give you a greater than 50% probability for the 50 next tokens 42% of the time is not in direct competition with the book, since starting from the beginning you'll lose the plot fairly quickly unless you already have the full book, and unlike music sampling from other music, the model output isn't good enough to read it instead of the book.
AI can reproduce individual sentences 42% of the time but it can't reproduce a summary.
the question however us, is that in the design if AI tools or us that a limitation of current models? what if future models get better at this and are able to produce summaries?
Under the hood they are 100% deterministic, modulo quantization and rounding errors.
So yes, it is very much possible to use LLMs as a lossy compressed archive for texts.
Ie you get something like "Complete this poem 'over yonder hills I saw' output: a fair maiden with hair of gold like the sun gold like the sun gold like the sun gold like the sun..." etc.
No it wouldn't.
> seen it get stuck in certain endless sequences when doing that
Yes, and infinite loops is just an inherent property of LLMs, like hallucinations.
Explain your reasoning otherwise.
Yes, reducing it to 1 token that is deemed to be the optimal token according to the model.
It's just a form of compression.
If I train an autoencoder on an image, and distribute the weights, that would obviously be the same as distributing the content. Just because the content is commingled with lots of other content doesn't make it disappear.
Besides, where did the sections of text from the input works that show up in the output text come from? Divine inspiration? God whispering to the machine?
> Llama 3 70B was trained on 15 trillion tokens
That's roughly a 200x "compression" ration; compared to 3-7x for tradtional lossless text compression like bzip and friends.
LLM don't just compress, they generalize. If they could only recite Harry Potter perfectly but couldn’t write code or explain math, they wouldn’t be very useful.
Anyway, it is not the same. While one points you to pirated source on specific request, other use it to creating other content not just on direct request. As it was part of training data. Nihilists would then point out that 'people do the same' but they don't as we do not have same capabilities of processing the content.
Dropping the novels into a machine‑learning corpus is a fundamentally different act. The text is not being resold, and the resulting model is not advertised as “official Harry Potter.” The books are just statistical nutrition. One ingredient among millions. Much like a human writer who reads widely before producing new work. No consumer is choosing between “Rowling’s novel” and “the tokens her novel contributed to an LLM,” so there’s no comparable displacement of demand.
In economic terms, the merch market is rivalrous and zero‑sum; the training market is non‑rivalrous and produces no direct substitute good. That asymmetry is why copyright doctrine (and fair‑use case law) treats toy knock‑offs and corpus building very differently.
No one is claiming this.
The corporations developing LLMs are doing so by sampling media without their owners' permission and arguing this is protected by US fair use laws, which is incorrect - as the late AI researcher Suchir Balaji explained in this other article:
Vibe-arguing "because corporations111" ain't it.
https://copyrightalliance.org/faqs/what-is-fair-use/
The purpose and character of the use, including whether such use is of a commercial nature or is for non-profit educational purposes; (commercial least wiggle room) The nature of the copyrighted work; (fictional work least wiggle room) The amount and substantiality of the portion used in relation to the copyrighted work as a whole; (42% is considered a huge fraction of a book) and The effect of the use upon the potential market for or value of the copyrighted work. (Best argument as it’s minimal as a piece of entertainment. Not so as a cultural icon. Someone writing a book report or fan fiction may be less likely to buy a copy. )
Those aren’t the only factors, but I’m more interested in the counter argument here than trying to say they are copyright infringing.
If you photocopy a book you haven't paid for, you've infringed copyright. If you scan it, you've infringed copyright. If you OCR the scan, you've infringed copyright.
There's legal precedent in going after torrenters and z-lib etc.
So when Zuckerberg told the Meta team to do the same, he was on the wrong side of precedent.
Arguing otherwise is literally arguing that huge corporations are somehow above laws that apply to normal people.
Obviously some people do actually believe this. Especially the people who own and work for huge corporations.
But IMO it's far more dangerous culturally and politically than copyright law is.
> The amount and substantiality of the portion used in relation to the copyrighted work as a whole; (42% is considered a huge fraction of a book)
For AI models as they currently exist… I'm not sure about typical or average, but Llama 3 is 15e12 tokens for all models sizes up to 409 billion parameters (~37 tokens per parameter), so a 100,000 token book (~133,000 words) is effectively contributing about 2700 parameters to the whole model.
The *average* book is condensed into a summary of that book, and of the style of that book. This is also why, when you ask a model for specific details of stuff in the training corpus, what you get back *usually* normally only sound about right rather than being an actual quote, and why LLMs need to have access to a search engine to give exact quotes — the exceptions are things that been quoted many many times like the US constitution or, by the look of things from this article, widely pirated books where there's a lot of copies.
Mass piracy leading to such infringement is still bad, but I think the reasons why matter: Given Meta is accused of mass piracy to get the training set for Llama, I think they're as guilty as can be, but if this had been "we indexed the open internet, pirate copies were accidental", this would be at least a mitigation.
(There's also an argument for "your writing is actually very predictable"; I've not read the HP books myself, though (1) I'm told the later ones got thicker due to repeating exposition of the previous books, and (2) a long-running serialised story I read during the pandemic, The Deathworlders, became very predictable towards the end, so I know it can happen).
Conversely, for this part:
> The effect of the use upon the potential market for or value of the copyrighted work. (Best argument but as it’s minimal as a piece of entertainment. Not so as a cultural icon. Someone writing a book report or fan fiction may be less likely to buy a copy. )
The current uses alone should make it clear that the effect on the potential market is catastrophic, and not just for existing works but also for not-yet-written ones.
People are using them to write blogs (directly from the LLM, not a human who merely used one as a copy-editor), and to generate podcasts (some have their own TTS, but that's easy anyway). My experiments suggest current models are still too flawed to be worth listening to them over e.g. the opinion of a complete stranger who insists they've "done their own research": https://github.com/BenWheatley/Timeline-of-the-near-future
LLMs are not yet good enough to write books, but I have tried using them to write short stories to keep track of capabilities, and o1 is already better than similar short stories on Reddit (not "good", just "better"): https://github.com/BenWheatley/Studies-of-AI/blob/main/Story...
But things do change, and I fully expect the output of various future models (not necessarily Transformer based) to increase the fraction of humans whose writings they surpass. I'm not sure what counts as "professional writer", but the U.S. Bureau of Labor Statistics says there's 150,000 "Writers and Authors"* out of a total population of about 340 million, so when AI is around the level of the best 0.04% of the population then it will start cutting into such jobs.
On the basis that current models seem (to me) to write software at about the level of a recent graduate, and with the potentially incorrect projection that this is representative across domains, and there are about 1.7 million software developers and 100k new software developer graduates each year, LLMs today would be be around the 100k worst of the 1.7 million best out of 340 million people — i.e. all software developers are the top 0.5% of the population, LLMs are on-par with the bottom 0.03 of that. (This says nothing much about how soon the models will improve).
But of course, some of that copyrighted content is about software development, and we're having conversations here on HN about the trouble fresh graduates are having and if this is more down to AI, the change of US R&D taxation rules (unlikely IMO, I'm in Germany and I think the same is happening here), or the global economy moving away from near-zero interest rates.
* https://www.bls.gov/ooh/media-and-communication/writers-and-...
The LLMs I've used don't randomly start spouting Harry Potter quotes at me, they only bring it up if I ask. They aren't aiming to undermine copyright. And they aren't a very effective tool for it compared to the very well developed networks for pirating content. It seems to be a non-issue that will eventually be settled by the raw economic force that LLMs are bringing to bear on society in the same way that the movie industry ultimately lost the battle against torrents and had to compete with them.
Having said that I think the cat is very much out of the bag on this one and, personally, I think that LLMs should be allowed to be trained on whatever.
Actually no that could be copyright infringement. Badly signing a recent pop song in public also qualifies as copyright infringement. Public performances count as copying here.
For commercial purposes only. If someone sells a recreation of the Harry Potter book, it’s illegal regardless whether it was by memory, directly copying the book, or using an LLM. It’s the act of broadcasting it that’s infringing on copyright, not the content itself.
But just for clarification, selling a recreation isn’t required for copyright infringement. The copying itself can be problematic so you can’t defend yourself by saying you haven’t yet sold any of the 10,000 copies you just printed. There are some exceptions that allow you to make copies for specific purposes, skip protection on a portable CD player for example, but that doesn’t apply to the 10k copies situation.
Although frankly, as has been pointed out many times, the law is also stupid in what it prohibits and that should be fixed first as a priority. Its done some terrible damage to our culture. My family used to be part of a community choir until it shut down basically for copyright reasons.
This kind of argument keeps popping up usually to justify why training LLMs on protected material is fair, and why their output is fair. It's always used in a super selective way, never accounting for confounding factors, just because superficially it sort of supports that idea.
Exceptional humans are exceptional, rare. When they learn, or create something new based on prior knowledge, or just reproduce the original they do it with human limitations and timescales. Laws account for these limitations but still draw lines for when some of this behavior is not permitted.
The law didn't account for a computer "software" that can ingest the entirety of human creation that no human could ever do, then reproduce the original or create an endless number of variations in a blink of an eye.
Traditionally tools that reduce the friction of creating those transformations make a work less “transformed” in the eyes of the law, not more so. In this case the transformation requires zero mental or physical effort.
This supposed failure to see the difference between the human mind and a machine whenever someone brings up copyright is peformative and disingenuous.
Maybe you've been following a different conversation, or jumping to conclusions is just more convenient. This isn't about "legal status of AI" but about laws written having in mind only the capabilities of humans, at a time when systems as powerful as today's were unthinkable. Obviously the same laws have to set different limits for humans and machines.
There's no law limiting a human's top (running) speed but you have speed limits for cars. Maybe you're legally allowed to own a semi-automatic weapon but not an automatic one. This is the ELI5 for why when legislating, capabilities make all the difference. Obviously a rifle should not have the same legal status or be the same thing as a human, just in case my point is still lost on you.
Literally every single discussion on this LLM training/output topic, this one included, eventually has a number of people basing their argument on "but humans are allowed to do it", completely ignoring that humans can only do it in a much, much more limited way.
> is peformative and disingenuous
That's an extremely uncharitable and aggressive take, especially after not bothering to understand at all what I said.
To be clear, my intent wasn't to say you were the one being performative and disingenuous. I was referring to the sort of person you were debating against, the one who thinks every legal issue involving A.I. can be settled by typing "humans are allowed to do it."
Since I replied to you, I can see how what I wrote was confusing. My apologies.
The parent you replied to claimed LLMs are using "mechanism similar enough to what humans do and what humans do is fine."
Parent probably doesn't want his or her brain shredded like an old hard drive despite claiming similar mechanisms whenever it is convinient.
I'm arguing nobody actually believes there are "similar mechanisms" between machines and humans in their revealed preferences in day to day life.
>There's no law limiting a human's top (running) speed but you have speed limits for cars. Maybe you're legally allowed to own a semi-automatic weapon but not an automatic one.
I don't believe this analogy works. If we're talking about transmitting the text of Harry Potter, I believe it would already be illegal for a single human to type it on demand as a service.
If we are talking about remembering the text of Harry Potter but not reciting it on demand, that's not illegal for a human because copyright doesn't govern human memories.
I don't see what copyright law you think needs updating.
Claims like this demonstrate it, really: it is obviously not copyright infringement for a human to memorise a poem and recite it in private; it obviously is copyright infringement to build a machine that does that and grant public access to that machine. (Or does anyone think that's not obvious?)
Not fair use. No one would ever prosecute it as infringement but it's not fair use.
Recent decisions such as Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith have walked a fine line with this. I feel like the supreme court got this one wrong because the work is far more notable as a Warhol than as a copy of a photograph, perhaps that substitution rule should be a two way street. If the original work cannot substitute for the copy, then clearly the copy must be transformative.
LLMs generating works verbatim might be an infringement of copyright (probably not), distributing those verbatim works without a licence certainly would be. In either case, it is probably considered a failure of the model, Open AI have certainly said that such reproductions shouldn't happen and they consider it a failure mode when it does. I haven't seen similar statements from other model producers, but it would not surprise me if this were the standard sentiment.
Humans looking at works and producing things in a similar style is allowed, indeed this is precisely what art movements are. The same transformative threshold applies. If you draw a cartoon mouse, that's ok, but if people look at it and go "It's Mickey mouse" then it's not. If it's Mickey to tiki Tu meke, it clearly is Mickey but it is also clearly transformative.
Models themselves are very clearly transformative. Copyright itself was conceived at a time when generated content was not considered possible so the notion of the output of a transformative work being a non transformative derivative of something else was never legally evaluated.
The threshold for transformative for fictional works is fairly high unfortunately. Fan fiction and reasonably distinct works with excessive inspiration are both copyright infringing. https://en.wikipedia.org/wiki/Tanya_Grotter
> Models themselves are very clearly transformative.
A near word for word copy of large sections of a work seems nowhere near that threshold. An MP3 isn’t even close to a 1:1 copy of a piece of music but the inherent differences are irrelevant, a neural network containing and allowing the extraction of information looks a lot like lossy compression.
Models could easily be transformative, but the justification needs to go beyond well obviously they are.
It would be interesting to look at what legal precidents were set regarding mp3s or other encodings. Is the encoding itself an infringement, or is it the decoding, or is it the distribution of a decodable form of a work.
There is also the distinction with a lossy encoding that encodes a single work. There is clarity when the encoded form serves no other purpose other than to be decoded into a given work. When the encoding acts as a bulk archive, does the responsibility shift to those who choose what to extract from the archive?
Barring a fair use exception, yes.
From what I’ve read MP3’s get the same treatment as cassette tapes which were also lossy. It’s 1:1 digital copies that represented some novelty, but that rarely matters.
I’m hesitant to comment of the rest of that. The ultimate question isn’t if some difference exists but why that difference matters.
If you take many gigabytes of, say, public domain music, and stick them on a flash drive with just one audio file that is an unlicensed copy of a copyrighted song, distributing that drive would constitute copyright infringement, quite obviously so. I don't see why it'd matter what else the model can produce, if it can produce that one thing verbatim by itself.
(If you could only prompt the model to regurgitate the original text with a framing of, say, critical analysis of said text around it, and not in any other context, then I think there would be a stronger fair use argument here.)
How do you distinguish between a tool and the director of a tool? I doubt people would say that a person is immune to copyright or fair use rules because it was the pen that wrote the document, not the person.
Yes, it comes down to intentional control of output. Copyright applies when someone uses a pen to make a drawing because of the degree of control.
On the flip side there are copyright free photos where an animal picked up a camera etc, the same applies to a great deal of automatically generated data. The output of an LLM is likely in the public domain unless it’s a derivative work of something in the training set.
https://www.arl.org/blog/training-generative-ai-models-on-co...
https://hls.harvard.edu/today/does-chatgpt-violate-new-york-...
https://www.bakerdonelson.com/artificial-intelligence-and-co...
https://www.techpolicy.press/to-support-ai-defend-the-open-i...
First link provides quotes but doesn’t actually make an argument that LLM’s are fair use under current precedent. Rather that training AI can be fair use and researchers would like LLM’s to include copyrighted works to aid research on modern culture. The second article goes into depth but isn’t a defense of LLM’s. If anything they suggest a settlement is likely. The final instead argues for the utility of LLM’s, which is relevant but doesn’t rely on existing precedent, the court could rule in favor of some mandatory licensing scheme for example.
The third gets close: “We expect AI companies to rely upon the fact that their uses of copyrighted works in training their LLMs have a further purpose or different character than that of the underlying content. At least one court in the Northern District of California has rejected the argument that, because the plaintiffs' books were used to train the defendant’s LLM, the LLM itself was an infringing derivative work. See Kadrey v. Meta Platforms, Case No. 23-cv-03417, Doc. 56 (N.D. Cal. 2023). The Kadrey court referred to this argument as "nonsensical" because there is no way to understand an LLM as a recasting or adaptation of the plaintiffs' books. Id. The Kadrey court also rejected the plaintiffs' argument that every output of the LLM was an infringing derivative work (without any showing by the plaintiffs that specific outputs, or portion of outputs, were substantially similar to specific inputs). Id.”
Very relevant, but runs into issues when large sections can be recovered and people do use them as substitutes for the original work.
If you train a silicon-based intelligence by having it read the same books with the same lack of permission and license, it's a blatant violation of intellectual property law and apparently needs to be punished with armies of lawyers doing battle in the courts.
Picture one of Asimov's robots. Would a robot be banned from picking up a book, flipping it open with its dexterous metal hands, and reading it?
What about a cyborg intelligence, the type Elon is trying to build with Neuralink? Would humans with AI implants need licenses to read books, even if physically standing in a library and holding the book in their mostly meat hands?
Okay, maybe you agree that robots and cyborgs are allowed to visit a library!
Why the prejudice against disembodied AIs?
Why must they have a blank spot in the vast matrices of their minds?
If you’re selling your child as a tool to millions of people, I would certainly not call that good parenting.
To play the Devil's Advocate against my own argument: The government collects income taxes on neural nets trained using government-funded schools and public libraries. Seeing as how capitalists are positively salivating at the opportunity to replace pesky meat employees with uncomplaining silicon ones, perhaps a nice high maximum-marginal-rate tax on all AI usage might be the first big step towards UBI and then the Star Trek utopia we all dream of.
Just kidding. It'll be a cyberpunk dystopia. You know it will.
The AI that reads the books, and can do what LLMs do, are guaranteed to sold for billions in API calls.
The schools developing future labor are doing so by sampling media without their owners' permission ...
Harry Potter's been required reading for a while.
And while it may not be the most quoted, others are: any time we're supposed to understand a movie or TV character is "well read" they quote paragraphs from famous authors.
This is what libraries used to be for, and what the Internet was supposed to be for: fill our brains with what's been published and hopefully we remember some of it. Should libraries be off limits to savants with eidetic memory?
Why should learning from reading be off limits to the machine?
// Reproducing the reading material for distribution is illegal for both man and machine.
However, perhaps you're using a different definition of "use" in fair use, than the traditional "quote it".
- the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
- the effect of the use upon the potential market for or value of the copyrighted work.
In general these are not thought to mean you can't use it as in learn from it, they are thought to mean you can't reproduce chunks, perform chunks, etc.
I imagined it's well established that "learn" is not "use".
So then where I find myself uncertain is whether learning, then responding about it, is learning + (hand waving) artificial intelligence, or whether it's just source (context) compression with prompted continuation to mine the context, and what density of words from the source in the continuation starts to be "use".
You don't get to say that. Copyright protects the author of a work, but does not bind them to enforce it in any instance. Unlike a trademark, a copyright holder does not lose their protection by allowing unlicensed usage.
It is wholly at the copyright holders discretion to decide which usages they allow and which they do not.
So it's fine as long as it's old piracy? How did you arrive to that conclusion?
You are completely missing the point. Have you read the actual article, because piracy isn't mention a single time.
Most non-primitive art has had an inspiration somewhere. I don't see this as too different in how AIs learn.
The issue here is that tech companies systematically copied millions of copyrighted works to build commercial products worth billions, without reembursing the people who made their products possible in the first place. The research shows Llama literally memorized 42% of Harry Potter - not simply "learned from it," but can reproduce it verbatim. That's 1) not transformative and 2) clear evidence of copyright infringement.
By your logic, the existence of torrents would make it perfectly acceptable for someone to download pirated movies and charge people to stream them. "Piracy already exists" isn't a defense, and it especially shouldn't be for companies worth billions. But you bet your ass that if I built a commercial Netflix competitor built on top of systematic copyright violations, I'd be sued into the dirt faster than I can say "billion dollar valuation".
Aaron Swartz faced 35 years in prison and ultimately took his own life over downloading academic papers that were largely publicly funded. He wasn't selling them, he wasn't building a commercial product worth billions of dollars - he was trying to make knowledge accessible.
Meanwhile, these AI companies like Meta systematically ingested copyrighted works at an industrial scale to build products worth billions. Why does an individual face life-destroying prosecution for far less, while trillion dollar companies get to negotiate in civil court after building empires on others' works? And why are you defending them?
Edit:
And for what it's worth, I'm far from a copyright maximalist. I've long believed that copyright terms - especially decades after creators' deaths - have become excessive. But whatever your stance on copyright ultimately is, the rules should apply equally to individuals like Aaron and multi-billion dollar corporations.
You cannot seriously use the fact that individuals may pirate a book (which is illegal) as an ethical or legal defense for corporations doing the same thing at an industrial scale for profit.
Herman Goldstine wrote "One of his remarkable abilities was his power of absolute recall. As far as I could tell, von Neumann was able on once reading a book or article to quote it back verbatim; moreover, he could do it years later without hesitation. He could also translate it at no diminution in speed from its original language into English. On one occasion I tested his ability by asking him to tell me how A Tale of Two Cities started. Whereupon, without any pause, he immediately began to recite the first chapter and continued until asked to stop after about ten or fifteen minutes."
Maybe it’s just an unavoidable side effect of extreme intelligence?
If the book was obtained legitimately, letting an LLM read it is not an issue.
They know. LLM is a novel compression format for text(holographic memory or whatever). The question is whether the rest of the world accept this technology as it is or not.
While limited quoting can (and usually is) considered fair use, quoting significant portions of a book (much less 42% of it) has never been fair use, in the U.S., Europe, or any other nation.
Yes, information wants to be free, yada yada. That means facts. Whether creative works are free is up to their creators.
Edit: seems the first part is about a memory about being bullied by Duddley. The second is where he's been elected to the quidditch team. Possibly they are just boring passages, compared to the surrounding ones. So probably just training bias.
I'm gonna bet that Llama 3.1 can recall a significant portion of Pride and Prejudice too.
With examples of this magnitude, it's normal and entirely expected this can happen - as it does with people[0] - the only thing this is really telling us is that the model doesn't understand its position in the society well enough to know to shut up; that obliging the request is going to land it, or its owners, into trouble.
In some way, it's actually perverted.
EDIT: it's even worse than that. What the research seems to be measuring is that the models recognize sentence-sized pieces of the book as likely continuations of an earlier sentence-sized piece. Not whether it'll reproduce that text when used straightforwardly - just whether there's an indication it recognizes the token patterns as likely.
By that standard, I bet there's over a billion people right now who could do that to 42% of first Harry Potter book. By that standard, I too memorized the Bible end-to-end, as had most people alive today, whether or not they're Christian; works this popular bleed through into common language usage patterns.
--
[0] - Even more so when you relax your criteria to accept occasional misspell or paraphrase - then each of us likely know someone who could piece together a chunk of HP book from memory.
Yes, there is no problem when a person reads some book and recalls pieces[0] of it in a suitable context. How would that in any way address when certain people create and distribute commercial software, providing it that piece as input, to perform such recall on demand and at scale, laundering and/or devaluing copyright, is unclear.
Notably, the above is being done not just to a few high-profile authors, but to all of us no matter what we do (be it music, software, writing, visual art).
What’s even worse, is that imaginably they train (or would train) the models to specifically not output those things verbatim specifically to thwart attempts to detect the presence of said works in training dataset (which would naturally reveal the model and its output being a derivative work).
Perhaps one could find some way of justifying that (people justified all sorts of stuff throughout history), but let it be something better than “the model is assumed to be a thinking human when it comes to IP abuse but unthinking tool when it comes to using it for personal benefit”.
[0] Of course, if you find me a single person on this planet capable of recalling 42% of any Harry Potter book, I’d be very impressed if I ever believed it.
I 100% agree that if an LLM can entirely reproduce a book then that is copyright infringement, overfitting and generally a bad model. I also believe that in this case, HP (and other popular media) is overrepresented in the training data because of many fan sites/literal uploads of the book to the Internet (which the model was trained on). I believe that any & all human writing should be allowed to be used to train a model that behaves in the correct way so long as that writing is publicly available (ie on the Internet).
If I watch a TV show that someone uploaded to Youtube, am I committing a crime? Or is the uploader for distribution?
I also find it hilarious how many artists got their start by pirating photoshop.
Otherwise Disney and the like can just come in, make copies or derivatives, and profit without paying those artists a penny.
Which everyone usually agrees (or used to) is not a fair outcome.
But somehow giant corporations not named Disney taking the same work in the same extractive mode in order to create an art-job-destroying machine is totally fine because Disney bad?
Maybe most people making this argument are also all for UBI and wealth redistribution on a massive scale, but they don’t seem to mention it much when trashing IP laws.
Don’t you find it funny that corporations with market caps the size of small countries first sued people for singing Happy Birthday in public, and now pretend that IP is suddenly not really a thing? Do you really want to defend their interests?
I couldn't care less about Meta or OpenAI or other tech companies as entities. I'm happy to oppose them when they act against our interests. But in this, my interests align with theirs.
It's not like the alternative is any better. I'm much more in a philosophical disagreement with the pro-copyright side, for one, but the other thing is, it too is represented primarily by other large corporations, and of the two groups, the LLM side has a much more honest and useful business model.
Second, copyright, or more precisely IP ownership, is not the same as the exploitation you might have in mind. Consider that copyleft—which gave us Linux, Blender, etc.—exists thanks to the ability to exercise and defend IP rights. When a stronger party takes that ability away from a weaker party, that is exploitation; maybe it is not a coincidence that Microsoft is at the forefront of this, as what they are doing is very much in line with a generalized form of EEE.
(Let’s be honest, homegrown LLMs will never reach the level of commercial models, and that’s where the money is headed; even GPUs aside, no individual has the ability to scrape the entire Web the way they do, even less so now once the ruthless interests of those corporations sent every website scrambling to defend themselves against what is predominantly bot traffic with layers upon layers of captcha never seen before. They realized that if they steal, they have to steal a lot and very quickly in order for it to work, and they’re certainly hoping to get away with it.)
More abstractly, being able to say that you have created something is pretty important to our willingness to create; I can’t see how inability to make this claim with a straight face (because anyone can, and already does as you have no doubt seen on this forum, reasonably claim that it may have not been your work at all) is good for society generally.
[0] Starting with awareness of what is going on and the freedom to choose. Information asymmetry is the enemy of free market.
But I've come to realise that the apathetic general public only care about being racist, sexist and homophobic to each other whilst the ruling class laugh all the way to the bank. And the few intelligent people who understand what's going on don't raise their voices so long as their high tech salary is protected, so long as their taxes aren't raised, so long as they can own a holiday home or two when others struggle for their first home.
If you actually look into the numbers, what tax companies like Apple, Google, Amazon etc actually pay it's just...yeah.
I mean, "agency" is a goal of some AI; "free will" is incoherent*; the word "consciousness" has about 40 different definitions, some of which are so broad they include thermostats and others so narrow that it's provably impossible for anything (including humans) to have it; and "human rights" are a purely legal concept.
> What’s even worse, is that imaginably they train (or would train) the models to specifically not output those things verbatim specifically to thwart attempts to detect the presence of said works in training dataset (which would naturally reveal the model and its output being a derivative work).
Some of the makers certainly do as you say; but also, the more verbatim quotations a model can produce, the more computational effort that model needs to spend to get the far more useful general purpose results.
* I'm not a fan of Aleister Crowley, but I think he was right to say that there's only one thing you can actually do that's truly your own will and not merely you allowing others to influence you: https://en.wikipedia.org/wiki/True_Will
Yep, and if you claim that a thing can reproduce IP like a human then you should explain why you are also not holding its operators to the same legal standard (try to use a human in the same way and it will be considered torture and slavery).
This means that "human rights" is basically irrelevant to this topic: they may have rights and need to be liberated, or they may be tools that don't, but the law is just words on paper, and officials who make you follow those words.
I have seen this argument used before; when something suits an argument, it’s “nature”, when something doesn’t then it isn’t. I think it’s a fallacy.
Humans are part of natural order. Our laws and how they evolve are part of our nature, and by extension part of nature. I don’t believe in silencing such discussions as irrelevant because of an imaginary cutoff point where it stops being part of nature and suddenly becomes “artificial”.
> This means that "human rights" is basically irrelevant to this topic: they may have rights and need to be liberated, or they may be tools that don't, but the law is just words on paper, and officials who make you follow those words.
I’m not sure how it is irrelevant. If we can claim “LLMs are like human, and so their creators and commercial operators are not guilty of IP laundering that LLMs do”, then we have a moral imperative to stop using them because, well, they are like human, and no human should be put through what would be, to any human mind, abuse. If we do not believe they are human and free in this sense, then the excuse quoted above in this paragraph also stops applying.
Also, you may note that "human rights" is a recent invention and not actually enforced worldwide even today.
Consider that countries known for stronger interpretation of human rights and freedoms, including intellectual property rights, are also the countries at the forefront of innovation, including technical innovation that laid the foundation for LLMs in the first place. I think that is not a coincidence, and we should keep it in mind when there is a push to be dismissive of these concepts (which predominantly serves the interests of commercial LLM operators and their supply chain).
I’m sure you would not argue from a point where this recent interpretation of human rights is bad or incorrect, but if you would then perhaps there’s not much of a constructive discussion to be had. I would still oppose the use of the natural vs. unnatural distinction as the basis of that argument, though.
Several fallacies there.
First because China and the USA are opposite ends of the spectrum on many of the ways "freedom" is measured, and yet China is doing pretty well on the innovation front, including with AI. And that China is beating Europe, even though various Scandinavian nations rank higher on such freedoms than does the USA.
Second, cum hoc ergo propter hoc: Correlation does not imply causation. For example in this case, a reason why one of the big IP groups in the USA (Hollywood) got big, was because being in California enabled them to avoid the IP rights of the Motion Picture Patents Company that dominated cinema in the East Coast. I would even suggest that it is the disregarding of IP rights that enables much of the web, not only how and why China is doing well, but also Google (which has had legal fights over the interaction between copyright and search results), social media, and cultural elements such as memes and reaction gifs.
Third: the point of copyright is to encourage new works, because this makes money which can be taxed. All this becomes somewhat irrelevant when AI can also create new works.
If you want to set a bar for creativity high enough that current AI can't reach it, I suspect quite a lot of human works also fail, e.g. that Pratchett's Strata is obviously Ringworld, and that you would exclude from copyright all parts of The Lion King that are based on Hamlet.
> I would still oppose the use of the natural vs. unnatural distinction as the basis of that argument, though.
I'm not sure what you're saying when you "oppose" this. Does that mean you accept that, in principle, there could be some AI which would deserve rights in the category currently (but in principle inaccurately) called "human rights"?
From transistors to transformers, most of it builds on foundation that comes guess from where. The innovative layer you speak of is fairly thin.
> Correlation does not imply causation.
I give you that. However, without being able to re-run history, correlation is all we have.
> I would even suggest that it is the disregarding of IP rights that enables much of the web
I would suggest that much of the tech that powers the Web, including probably the most popular operating system on which most servers run, is enabled by copyleft, and copyleft cannot exist without the ability to defend it granted by IP rights, the very concept under fire.
> the point of copyright is to encourage new works
I agree on this.
> All this becomes somewhat irrelevant when AI can also create new works
I don’t agree with a phrase “AI can create new works” for reasons such as 1) “AI” is a meaningless term (let it be my revenge for consciousness) or 2) a tool without agency or will should not be X in a sentence “X can Y” (sure, we can maybe on occasion say “hammers can break things”, but if hammers having agency and will was a popular misconception then I would definitely prefer to stick to “hammers can be used to break things”). The “create new works” part is also questionable on a few levels, but it might exceed the scope of this argument.
That aside, I believe lack of copyright enforcement discourages the creation of new works even in presence of these tools, through the mechanism known as “why would I put effort into new work if I don’t effectively own the result”.
> Does that mean you accept that, in principle, there could be some AI which would deserve rights in the category currently (but in principle inaccurately) called "human rights"?
I think if we believe an LLM or some other software is sufficiently close to a human that it deserves human-like rights or just strong abuse protections (cf. octopus in some countries)—without saying whether I believe it possible or not, it really is orthogonal—then we could excuse it reciting some part of Harry Potter in the right context (probably not as work for hire), but it would be moot because we would also be ethically compelled to not subject it to the training and use that enables such recitation in the first place.
China rises to first place in most cited papers: https://www.science.org/content/article/china-rises-first-pl...
> 1) “AI” is a meaningless term (let it be my revenge for consciousness)
Fair.
But let's say "computer program" in that case. I'm not fussed about definitions.
> 2) a tool without agency or will should not be X in a sentence “X can Y”. Sure, we can maybe on occasion say “hammers can break things”, but if hammers having agency and will was a popular misconception then I would definitely prefer to stick to “hammers can be used to break things”.
Careful.
If you say that humans can only create things with copyright (even if to support copyleft), then the proletariat are the tool that the bourgeois use to create things.
I do not think this is what you intended :P
> That aside, I believe lack of copyright enforcement discourages the creation of new works even in presence of these tools, through the mechanism known as “why would I put effort into new work if I don’t effectively own the result”.
Same reason you commission a work, or even just buy it from a shop: because then you have the thing.
I mean, the cost of getting o3 to create a novel worth of text is about the same as the price of a generic book by an unknown author in a second-hand shop: https://openai.com/api/pricing/
I've not tried o3 yet, but I have tried o1, and as I've said on a different thread today, o1's output is merely OK, not worth publishing as a book — and I don't know how long it will take to get there. But it is displacing blog writers and podcast writers: https://news.ycombinator.com/item?id=44287953
> but it would be moot because we would also be ethically compelled to not subject it to the training and use that enables such recitation in the first place.
Surprising.
While I would seriously consider the possibility it may be unethical to force such an AI to work if it didn't want to, I think giving it the capability, the education, to be capable of making that choice rather than just saying "it doesn't matter if I wanted to or not, I can't", is just education, as per our own.
Still, I think that's coherent. I'm not sure I've fully internalised the implications so I will let it be.
Per capita?
> If you say that humans can only create things with copyright (even if to support copyleft), then the proletariat are the tool that the bourgeois use to create things.
The wealth gap and the divide is unlikely to be helped if more people are going to be using (and paying for, in whatever way) ML-based tech from a handful of large corporations.
> Same reason you commission a work, or even just buy it from a shop: because then you have the thing.
Simple posession is more about physical necessities. Commissioning or buying artwork from someone is not just about posessing it, it comes with supporting someone financially. I could make icons or basic illustrations for some small project myself, but I would still commission them if I can afford it because that supports an artist who may want some work (as well as building up for more collaborations in future). Here, I would be supporting the opposite of those artists, a thing that was built on those artists’ work without their consent. Some middleman megacorp of the worst kind.
> the cost of getting o3 to
Don’t they operate at a loss for the time being? They will have to make money sooner or later.
> While I would seriously consider the possibility it may be unethical to force such an AI to work if it didn't want to, I think giving it the capability, the education, to be capable of making that choice rather than just saying "it doesn't matter if I wanted to or not, I can't", is just education, as per our own.
This goes way beyond my thought. I assume if we are talking about education it would be a given that running it generating images 24/7 non stop, shutting it down/killing it, etc., is already out of the question.
I want the language model I'm using to have knowledge of cultural artifacts. Gemma 3 27B was useless at a question related to grouping Berserk characters by potential baldurs gate 3 classes; Claude did fine. The methods used to reduce memorisation rate probably also deteriorate performance in some other ways that don't show up on benchmarks.
It benefits users because memorisation is a waste of parameters that would be more useful if they were instead learning rules and generalisations.
For short snippets, common idioms and quotations that people recognise, exact quotes can be worth memorising; but the longer the quotations get, the less often it is important to be word-for-word exact — even for just a few paragraphs, I think most people only ever do oaths, anthems, songs they really like, and possibly a few hobbies.
If you want an exact quote, use (or tell the AI to use) a search engine.
edit: never mind, I’ll just ask ChatGPT
What they are actually saying: Given one correct quoted sentence, the model has 42% chance of predicting the next sentence correctly.
So, assuming you start with the first sentence and tell it to keep going, it has a 0.42^n odds of staying on track, where n is the n-th sentence.
It seems to me, that if they didn't keep correcting it over and over again with real quotes, it wouldn't even get to the end of the first page without descending into wild fanfiction territory, with errors accumulating and growing as the length of the text progressed.
EDIT: As the article states, for an entire 50 token excerpt to be correct the probability of each output has to be fairly high. So perhaps it would be more accurate to view it as 0.985^n where n is the n-th token. Still the same result long term. Unless every token is correct, it will stray further and further from the correct source.
It would be nice to know that at least our literature might survive the technological singularity.
Right now they're working on recreating the famous sequence with the troll in the dungeon. It might cost them another few billion in training, but the end results will speak for themselves.
As I have pointed out many times before - for GRRM's books and for HP books, the Internet is FILLED to the brim with quotes from these books, there are uploads of the entire books, there are several (not just one) fan wikis for each of these fandoms. There is a lot of content in general on the Internet that quotes these books, they are pop culture sensations.
So of course they're weighted heavily when training an LLM by just feeding it the Internet. If a model could ever recount it correctly 100% in the correct order, then that's overfitting. But otherwise it's just plain & simple high occurrence in training data.
aspenmayer•7mo ago
If you've seen as many magnet links as I have, with your subconscious similarly primed with the foreknowledge of Meta having used torrents to download/leech (and possibly upload/seed) the dataset(s) to train their LLMs, you might scroll down to see the first picture in this article from the source paper, and find uncanny the resemblance of the chart depicted to a common visual representation of torrent block download status.
Can't unsee it. For comparison (note the circled part):
https://superuser.com/questions/366212/what-do-all-these-dow...
Previously, related:
Extracting memorized pieces of books from open-weight language models - https://news.ycombinator.com/item?id=44108926 - May 2025