Judge Alsup issued the original order that determined they were liable for piracy but that training LLMs on books was fair use. It's worth reading if you're interested in the topic. https://www.courtlistener.com/docket/69058235/231/bartz-v-an...
He's also a longtime hobbyist programmer working in BASIC, much of it in support of his ham radio hobby. Screenshots of his shortwave propagation prediction program here [1].
[1] https://www.theverge.com/2017/10/19/16503076/oracle-vs-googl...
His middle name is Haskell.
It's funny that crimes can be settled in cash. IOW, everything has a price; and the price is always right. Settlement ought to be the euphemism for blood money.
In addition to the settlement, what I'd consider fair is to have these companies pay royalties in perpetuity. Of course, that's not tractable.
How does intellectual work get funded in this insane world if yours, pray tell?
So payments for the right to do so aren't payments required to bring anything new into existence at that point, save for the legal fiction.
Now you might argue that the future copy-licencing rents are necessary to bring the _original_ creation into being. But that doesn't make them _not rents_.
But I would say that's the second assumption you're baking in here.
As in, we live in a world where e.g. the movie Toy Story exists. Now, certainly Toy Story does provide some good or value to the world. But I don't think you can assume such things provide more value than e.g. open science, free transformation of works, etc.
I get that people enjoy our current IP culture but saying certain things wouldn't exist in an IP-free world is just an argument from consequences that doesn't even really compare consequences between the two.
For works published before copyright laws existed or after property protections expire, anyone should be able to use it for anything forever without consent.
Intellectual work still manages to get funded in this 'insane world' - although given the classical artist/patron system has given way to state-based grants and a select capitalisation of Art post-Warhol, the concept of Universal Basic Income tends to be promoted the desired successor.
Speaking of insane worlds, how does the concept of the Public Domain work in yours?
> Speaking of insane worlds, how does the concept of the Public Domain work in yours?
Can you elaborate on what you're asking? I don't understand your question.
1. Copyright protections as a concept are an incredibly modern phenomenon, mostly limited in practice to Western Capitalist Democracies. 2. Outside of a short monetisable window (albeit one extended and irrevocably marred by Disney/Sonny Bono) your 'insane' hypothesis is in fact the status quo 3. Much intellectual work is published into the Public Domain, and all copyrighted work eventually ends up in the Public Domain. Your position appears to presuppose a world without such an entity.
As to what copyright actually achieves? It's mostly a mechanism by which the media gatekeepers and owners of capital use legislative and social imbalance of power to deny artist the rights and royalties for mechanical reproduction and otherwise impose financial serfdom.
This is achieved mainly by Copyright Enclosure, whereby musicians are typically pressured or contractually obligated to surrender their master recordings and intellectual property, and by contractual clauses like Controlled Composition Clauses, whereby Labels reduce the mechanical royalties they pay to artists who write their own songs, often paying below the standard statutory rate.
Maybe not all creative works are deserving of monopoly profits just by sitting on the ass in any case, and should stand on their own merits by producing downstream value that can be sold for whatever they can be sold for, by whoever puts in the work to deliver the value to the end user in a competitive manner. You know, open markets.
Attribution I can see. Consent or payment beyond market value, why? Just because you put in a billion hours to make a shitty $1 value output I should pay you a billion hours worth of labor?
Most of these turn intellectual work into that of indie musicians, or outright beggars. You are stepping dangerously close to stripping people rights in favour of giant AI companies.
> How do newfangled "content creators" fund their work? We already live in this world.
"Content creators" heavily rely on IP protections. You could always try taking some youtube videos with 100M views, altering them a bit an using them as your own and seeing how that goes down. Do let me know!
> Maybe not all creative works are deserving of monopoly profits just by sitting on the ass in any case, and should stand on their own merits by producing downstream value that can be sold for whatever they can be sold for, by whoever puts in the work to deliver the value to the end user in a competitive manner. You know, open markets.
And open market is not one where I can say that you are just sitting on your ass, so I'll take your stuff and sell it.
> Attribution I can see. Consent or payment beyond market value, why?
Because it's my stuff of course! And why should one even pay market value in your world? Why not always 0?
> Just because you put in a billion hours to make a shitty $1 value output I should pay you a billion hours worth of labor?
What in the world are you on about? If the price someone puts on their IP seems too high to you, you should not pay that price. We wholeheartedly agree. Where we disagree is where you from this conclude that you can just choose the price yourself and take it anyway!
No, that's what they got in trouble for - a lack of consent.
If the author consents, it would have been fine. If they bought the books, then it is fine. Digitisation through destruction, like most book scanning systems. As long as the original work is destroyed during the process, and you actually paid for it, then it is fair use.
If it regurgitates, then the author can sue you again. So you are incentivised to make damn sure it doesn't. That's not covered by fair use.
Its only if the original cannot be accessed anymore, and you paid to get the original. Both must be true, for fair use to hold.
> Literary works, including computer programs and databases, protected by access control mechanisms that fail to permit access because of malfunction, damage, or obsoleteness.
DRM being covered under other laws, and being gross, still applies. And still applies to industry giants, too. Which is why most who do this, like Google, actually buy physical copies and scan it destructively, so they don't have to deal with it.
If you asked the right politician when all these rules were being written, the intent was that each person who needs to format-shift their media would independently write their own decryption tools, use them for lawful purposes only, and then dutifully delete them the moment they were no longer needed. This is, of course, laughable.
Of course, if Anthropic was, say, buying and decrypting Kindle books TODAY; they probably could get Claude to vibe-code a DRM decryption tool[0]. That would actually be within the bounds of this asinine law. If Anthropic started off by doing this, however, they probably would have just used a decryption tool found on the Internet, and that would have invited different legal challenges. Like, is it legal to use an unlawful tool to accomplish something otherwise legally protected? The courts so far have been very hostile to ANY attempt to tie the anticircumvention provisions of the DMCA to fair use. They could easily say "No, you only get to format shift with your own tools".
[0] Related note: I really wish I had Mythos access, just so I could jailbreak my iPad on modern iPadOS. No other reason.
Learning from and building on previous work is civilization. Copyright maximalism is a plague.
We may need some new legislation. An LLM is not a person, but its also not just a storage solution.
Those regulations and principles are for humans.
Either the major LLMs are software tools deployed by ostensibly-profit-seeking companies, and regulations based on the notion that "making humans pay to make use of the things they've learned is profoundly antisocial" don't apply, or the LLM companies have a bigass swarm of unpaid -er- "servants", and labor laws and other human rights regulations do apply.
They are, presumably, human. We can perfectly well say that humans have certain rights without needing to give machines those same rights.
For example, we've more or less all agreed that it's fine for a human to watch a movie and enjoy the memories forever, and be inspired by it forever. But we've also more or less all agreed that that doesn't mean that a human can use a machine to record that movie and keep it forever.
> Learning from and building on previous work is civilization. Copyright maximalism is a plague.
The debate has existed for several generations at this point. You may disagree with the mainstream opinion, but it's disingenuous to frame it as "copyright maximalism".
Like most things said on social media, not being copyrightable.
Yeah, you are right. Have you been paying your dues to the authors of your math books in first 4 grades? I think 15% of your wages as an engineer would suffice. These kids continue to profit for years, and they are so many. Gotta pot a stop to that IP theft.
In my mistake I thought copyright was about copying rights, not paying for using ideas themselves. If just being downstream from a copyrighted work is infringement even without substantial similarity, then it's more like patents that expire in lifetime + a million years.
IMO the spirit of the law would prohibit LLMs from training, and the letter of the law leaves room for that only because nobody thought to write down "books are for people to read".
> therefore laws target them too, and now it looks like humans using LLMs to train via ingestion of books is deemed fair use
Which is orthogonal to my point.
Yes, that is the entire history of humanity. People steal the last generations works and build something from it to make it their own.
If he was not aware, I wonder if he still would have described the process as "exceedingly transformative" had he been aware.
Note also that Sonnet 3.7 had to be jailbroken.
Note also that they got high memorization for a few books that were widely quoted. The books in question can probably also be "retrieved" by putting phrase prefixes into Google, which is probably why Sonnet 3.7 knows them with the precision of a fanboy. Material being widely repeated in the training set is a well-known cause of memorization.
Edit: Apologies, I misread it as "100 pages". My point about copyright still stands, though.
Personally I think our legal system is just not set up for a world where we can download mindstates in numeric form. Would a sufficiently detailed recording of my brain violate copyright? If simulated, it could certainly be elicited to commit violations.
edit: At any rate, Anthropic are not publishing the Sonnet 3.7 weights.
Some of us have a good enough memory.