Oh look, another tech giant treating open knowledge initiatives like their personal data buffet. Let me translate this corporate nonsense for you:
Meta: “We need training data for our AI!” Also Meta: Let’s leech 81.7TB from a community project without contributing anything back.
The absolute audacity of downloading terabytes through torrents while their employees were internally admitting it was “legally problematic”. And the best part? They couldn’t even be bothered to seed properly - just grab and go, classic corporate behavior.
Remember when companies actually contributed to open source instead of just parasitically consuming it? But no, they’d rather burden volunteer-run projects with massive bandwidth costs while their lawyers probably bill more per hour than these projects’ entire monthly budget.
Pro tip Meta: If you’re going to pilfer knowledge from the commons, at least seed back properly. Your “move fast and break things” motto isn’t supposed to apply to community archives.
Not seeding is crazy …
My seedbox is locked and load, please point me to the. Torrent in need. Archive team assemble!
Yes please support annas-archive!! It is a wonderful project. I can essentially get an epub file for any book (including banned books) I want. They have so much more than that too.
This is the website listed in the article
An alternate domain https://annas-archive.li/
When you’re shilling for copyright, at least pick a lane. Are they bad for “pirating” or bad for not supporting “piracy”?
I guess it doesn’t matter as long as the owners collect their rent.
They are pirating, while also DOSing the providers.
Just gotta love these big tech companies and their bullshit double standards.
Do it, Judge. Protect the wealthy and say it’s not piracy. Do it.
It’s not piracy. For corporations. For you and me believe it or not, straight to jail!
Just make an llc, now its legal again.
I’d almost like to think an LLC would be enough, but I suspect that only works if you also have a billion in VC funding and political connections.
Oh for sure, since the law is basically toilet paper for billionaires at this point.
They’ll be fined 100k
And they’ll ham up how punished and sorry they are, and how thankful they are for the judge handing down “fair and impartial” justice.
If buying ain’t owning, than downloading…
oh wait, that’s our slogan
https://phys.org/news/2010-11-million-dollar-verdict-music-piracy-case.html
In all fairness, meta should be assessed a fee of 250k per EACH pirated work.
This would amount to forfeiting all assets to doge.
Assuming 2.6 MB per book.
81 TB would be 32,667,175 books.
At $250k per book that would come out to:
$8.17 trillion.
And I’d guess all that money would then go to military funding, with Anna’s Archive, again getting nothing out of it?
It would go to… Uh…
HEY SOMEONE PUT A DEAD CAT ON THE TABLE!
They might end up having to pay more money than exists on the planet at that rate.
Good
Edit - See Gary Bowser
Yes, that one.
Anna’s Archive: Mirror our database, help us preserve Humanity’s knowledge
Facebook: I’ll just torrent what I need, see yaa
These big tech monopolies are a curse to humanity…
Facebook: I’ll just
torrent what I needburden your underfunded project and volunteers with over 81 TB of bandwidth costs without contributing anything in return, see yaaFTFY
Rules for thee, not for me
But did they keep a good ratio though?
1000% guarantee those mf’s had their upload choked to 20kbps
Nah they used a leeching client. No upload at all.
In copyright protection terms the ratio shouldn’t matter. They should pay for all the lost profits from pirating everything they’ve downloaded. Every time someone pirated it should be counted. And every time someone uses the AI trained on the data.
They can become the corporate Jesus of the interwebs, having paid for our sins.
Meta has open sourced every single one of their llms. They essentially gave birth to the whole open llm scene.
If they start losing all these lawsuits, the whole scene dies and all those nifty models and their fine-tunes get removed from huggingface, to be repackaged and sold to us with a subscription fee. All the other domestic open source players will close down.
The copyright crew aren’t the good guys here, even if it’s spearheaded by Sarah Silverman and Meta has traditionally played the part of the villain.
If the existence of open source LLMs hinges on the benevolence of one of the few most cancerous tech companies in the world, maybe they’re not really worth it?
This isn’t about “heroes” and “villains”. Facebook has been and has stayed the “villain”, they’ve done something colossally illegal that any mere mortal would be sued to death for (by an another “villainous” instance, the media system that has made piracy a necessity in the first place), and they’re hoping to get away with it simply on technicalities and by having more money for better lawyers. Rules are rules, if you don’t like them maybe Facebook should try to change them (and not just for themselves, but for the rest of us too)?
The existence hinges on the rewriting and strengthening of copyright laws by data brokers and other cancerous tech companies. It’s not Meta vs us, but opensource vs Google and Openai.
They are being sued for copyright infringement when it’s clearly highly transformative. The rules are fine as is, Meta isn’t the one trying to change them. I shouldn’t go against my own interests and support frivolous lawsuits that will negatively impact me just because Meta is a boogeyman.
It’s not Meta vs us, but opensource vs Google and Openai.
I never said it’s Meta vs us. It’s Meta vs (in this particular case) the book publishing industry. You can’t reduce the whole situation to open source vs closed source, there’s other “axes” at play here as well.
They are being sued for copyright infringement when it’s clearly highly transformative
They downloaded the entire Libgen and more. Going by the traditional explanations of piracy, that’s like stealing several hundred bookstores worth of books all at once, and then claiming it’s alright because your own writing is not plagiarised from any of the books you’ve stolen. (Piracy is not the same as actual stealing of course, but countless people have been being legally bullied and ruined with that logic.) Meta also got its data from Internet Archive; unless they only obtained their materials that are public domain or under a similar license, they’ve obtained a lot of material that IA has been sentenced for allowing unlimited access to back in 2020 (if you’ve followed the Hachette v. Internet Archive case). The brainfucking conclusion of your and Facebook’s case is that using illegal services is perfectly legal as long as you sufficiently transform the results of the illegal activity.
The rules are fine as is
Actually they’re not. Copyright law is insanely restrictive, and I don’t think you’ve dealt much with media if you think it’s fine (but I don’t wish to delve into this further as it’s beyond the scope of discussion).
Meta isn’t the one trying to change them
Of course they’re not trying to change them, that’s the point, they will get away with breaking them while being perfectly fine with other actors not being able to do so.
All LLMs and Gen AI use data they don’t own. The Pile is all scraped or pirated info, which served as a starting point for most LLMs. Image gen is all scraped from the web. Speech to text and video gen mainly uses YouTube data.
So either you put a price tag on that data, which means only a handful of companies can afford to build these tools (including Meta), or you understand that piracy is the only way for most to aquire this data but since it’s highly transformative, it isn’t breaching copyrights or directly stealing from them as piracy “normally” is.
I’m being pragmatic.
Meta stole from everyone, including those that struggle to make ends meet, so it doesn’t matter that they gave you back some of it. Any moral qualms should evaporate when you consider that they did it to create shareholder value and the rest is philanthropy (aka pretend tax). As a socialist I believe that man is owed for his work and you can’t take from him even though technology makes it so easy.
Calling property labor, doesn’t make you a socialist.
You’re confusing libleft with left.
No. Seriously, why do you want to call yourself a socialist?
The world is in a mess is that we were told to choose between fascists and pro-market technocrat libertarians pretending to be leftists. This is a worldwide issue that’s doubly important because those liberals guilt trip us for not supporting them and that’s why I’m just laying little bricks here and there. At the end of the tunnel we either rework our society into a socialist one or we succumb to feudal lords again. Years of neoliberal hegemony needs to be undone so I try to go against the grain like that sometimes, hoping I made someone think.
When you call yourself a socialist, what do you mean by that term?
I assume you probably want to know how this kind of leftism is different from others or other ideologies calling themself leftist, rather than for me to write an essay on myself.
I believe in equal opportunity but reject that you should be able to „win” in any system. I believe in empathy over soulless meritocracy. I believe in collective ownership but don’t reject that one is owed for his work. You could say it all stems from egalitarianism but this term has been caricatured by liberals too. For a long time I thought social democracy as an ideology gives you enough levers in the system to steer it toward that goal but time and time again it turned out that in most places SocDem parties are no different from liberal ones and so I learned from past mistakes.
As a socialist I believe intellectual property is a falsehood and technological advancement should be for the public good. Open source LLMs are for the public good.
Given the options between having open source LLMs and the US Govt banning non-corpo non-proprietary LLMs and giving a free pass to people like Musk and Altman and Zucc to monopolize, I happily pick the former.
You’re delusional if you think they will pay anyone, the only way zucc will pay is with a guillotine.
Corpos will make inter-platform deals that’ll simply make all online data licensable for the right price and enrich each other so you can’t avoid it while still actually being a career creative, but price out academic researchers and the public sector so that all fruits of it stay behind closed R&D doors and be free of ethics etc.
Continuing in your role as a useful idiot, you’ll also most likely also foot the bill for it via subsidies from your taxes to “develop the AI sector” in some anti-China dick measuring contest by the US.
You will then be sold this data back via proprietary chat bots via a monthly subscription and you better pay up because once it gets really good, it will become mandatory to use for just about any job, leaving you with no choice.
Or you can support FOSS LLMs.
I support FOSS LLMs, but which actually exist? Which LLMs have open-sourced all their training data?
Mistral? Deepseek?
Not LLM but also SD which uses a very popular free dataset.
Can I freely download all the training data for any of those? I was under the impression they were all trained on non-licensed and copyrighted data.
It’s complicated.
I know Stable Diffusion best so I’ll speak to that, they used to the LAION-5B dataset, which is, in practice freely available to download and use:
https://www.kaggle.com/code/vitaliykinakh/guie-laion-5b-collect-and-download
https://github.com/opendatalab/laion5b-downloader
It’s also on HuggingFace but it’s unavailable.
https://huggingface.co/datasets/danielz01/laion-5b
But you can use this smaller newer version:
https://huggingface.co/datasets/laion/relaion2B-en-research
Whether it’s appropriately licensed is an unsolved question though.
The dataset itself and the text portion of the text-imags pairs needed for training is CC-BY-SA, the newer versions linked above are CC-BY-4.0. https://creativecommons.org/licenses/by/4.0/deed.en
The images however are technically under their own copyright, which in practice means each of the billions of images could or could not have a licence that implicitly or explicitly forbids AI training use or forbids it only for commercial use.
Whether such a license is legally binding is at present unknown though, since licenses primarily deal with reproductions, which the pro-AI folks argue isn’t the case, and that training of NNs is more akin to viewing an image and memorising the patterns and relationships within, like a person viewing it.
That would make it non-infringing and therefore the model itself libre. In that case Mistral and LLaMa are also libre as long as the model itself is open source, which in this case really means “open weights”, so not like GPT and anything by “”“OpenAI”“”.
Weights are the result of a model being trained essentially. They’re they key bit that makes it or breaks it and how it works. Given that and knowing the structure of the model and framework used you can refine, modify and distribute it.
Those against AI will say that it’s more akin to file compression and that in one form or another it’s misuse. That would make the model an infringing derivative work and as such nor libre even if the model weights are open source.
In a way though you could argue that me vaguely memorising the imagery of a dude dressed in white holding a laser sword is just a lossy compressed copy of the copyrighted work of Star wars, and it’d be absurd to think that’s a violation and that infringement only occurs if I reproduce a work of substantial similarity commercially from that memory.
If I use Krita and draw a beautiful landscape which has been informed and inspired by at least in part by a movie I saw, is that copyright infringement or not? What if I use AI?
Well, current laws don’t say. We measure infringement in substantial similarity, provenance of information only comes in later (e.g. to prove against accidental similarity).
That’s also my own personal stance on the legal side of things, so up to you how you see it.
“Meta downloaded millions of pirated books from LibGen through the bit torrent protocol using a platform called LibTorrent. Internally, Meta acknowledged that using this protocol was legally problematic,” the third amended complaint noted.
Just want to make clear that Libtorrent is just the torrent application they were using, while the Libgen torrents are easily accessible on the libgen site, not through a separate “platform” called Libtorrent.
I wish people like us could help with these complaints, because then they might actually get the details more accurate to reality.
https://libgen.is/repository_torrent/
The amended complaint makes it sound like Libtorrent is a private tracker website when its just the application they were using on the publicly available torrents.
