4 ms·

Why don't Google, OpenAI, Anthropic, Facebook & Co defend Anna's Archive publicly? Coming out would be a bold move for them.

toomuchtodo 1mo ago
No gain, all liability. Easier to cut them a check for access to training data and say nothing. Unless legal discovery was performed, the outside world would never know, and the payment records would roll off corporate records through a record retention schedule eventually. Could obfuscate it as a contractor consulting fee ("knowledge management subject matter expert") if you wanted to get tricky, depending on the risk appetite of whomever would receive the funds. (not legal advice!)
- p-e-w 1mo ago
  There’s zero liability in a company stating publicly that they support Anna’s Archive. Zero. Free speech protections cover much more egregious statements than that.
  - toomuchtodo 1mo ago
    I disagree. Anyone with even a hint of standing will sue, and keep suing. As someone who has to work with corporate counsel often, do not say anything you don't have to say. Only say what is absolutely necessary. Free speech protects you from your government. It does not shield you from civil suits, and the US is extremely litigious.
  - criddell 1mo ago
    Maybe they are worried about claims of contributory infringement?
  - pibaker 1mo ago
    Your freedom of speech is your opponents' lawyers' wet dream. Your publicized support for a known piracy operation will not look very good in the court when you get sued by copyright holders.
peri-cl 1mo ago
Anthropic paying $1.5 billion in fines for downloading Anna's Archive established a moat. They want it to be illegal to pirate books: they can afford the penalties and continue doing it. Just like they want it to be illegal to run local ML inference.
- spwa4 1mo ago
  In theory none of them actually got the right to train on illegally downloaded books. Anthropic was simply punished for doing it once. One wonders if they're still doing it.
  - outside1234 1mo ago
    Of course they are. They have just put on their Swiss Banker suit now and have all sorts of deflection techniques in place such that, of course, "the money has the stamps that says its clean" (when it it really blood money hidden behind a pretty wall).
  - ungut 1mo ago
    OpenAI plainly admitted that it is impossible not to do so in a House of Lords inquiry. So, presumably there is no way around it to train models. There is just not enough non-copyrighted data out there.
    - yorwba 1mo ago
      You mean this one https://committees.parliament.uk/writtenevidence/126981/pdf/ https://committees.parliament.uk/writtenevidence/126981/pdf/ where they write "it would be impossible to train today’s leading AI models without using copyrighted materials"? That doesn't mean they have to download those materials illegally. For a billion dollars, you can easily buy one legal copy of each book in Anna's Archive and still have some cash left over to run a whole-of-internet scraping operation.
      - spwa4 1mo ago
        I'm pretty sure we would know if they did that. And we don't. Plus this is not legal in the EU (and Canada, and ... let's just say the entire rest of the world, and accept that I'll be wrong for one or two smaller countries). Doesn't that matter? Or is only Mistral disallowed from training on copyrighted materials? Je veux ma chaton fat, goddamit!
kmeisthax 1mo ago
I'm pretty sure[0] they're all using shadow libraries, and saying things in favor of them would increase their liability. Furthermore, every pirate wants to be an admiral. None of the big tech companies are actually in favor of any amount of copyright reform. They never have been. There is a huge gulf between "personally benefitting from copyright theft" and "actually wants to legalize the theft". Anthropic still believes they deserve to be paid for their models, they just have this delusion in their head that doing a bunch of computation on stolen data is equivalent to actual human creativity. [0] OpenAI, Anthropic and Facebook have been shown in court to be using shadow libraries, I don't know about Google.
TZubiri 1mo ago
Those are all law-abiding organizations, which AA is not. INB4: "Here is one time one of those organizations broke the law". Don't go there, absolute lowest level of conversation.

toomuchtodo 1mo ago

No gain, all liability. Easier to cut them a check for access to training data and say nothing. Unless legal discovery was performed, the outside world would never know, and the payment records would roll off corporate records through a record retention schedule eventually. Could obfuscate it as a contractor consulting fee ("knowledge management subject matter expert") if you wanted to get tricky, depending on the risk appetite of whomever would receive the funds. (not legal advice!)

p-e-w 1mo ago
There’s zero liability in a company stating publicly that they support Anna’s Archive. Zero. Free speech protections cover much more egregious statements than that.
- toomuchtodo 1mo ago
  I disagree. Anyone with even a hint of standing will sue, and keep suing. As someone who has to work with corporate counsel often, do not say anything you don't have to say. Only say what is absolutely necessary. Free speech protects you from your government. It does not shield you from civil suits, and the US is extremely litigious.
- criddell 1mo ago
  Maybe they are worried about claims of contributory infringement?
- pibaker 1mo ago
  Your freedom of speech is your opponents' lawyers' wet dream. Your publicized support for a known piracy operation will not look very good in the court when you get sued by copyright holders.

peri-cl 1mo ago

Anthropic paying $1.5 billion in fines for downloading Anna's Archive established a moat. They want it to be illegal to pirate books: they can afford the penalties and continue doing it. Just like they want it to be illegal to run local ML inference.

spwa4 1mo ago
In theory none of them actually got the right to train on illegally downloaded books. Anthropic was simply punished for doing it once. One wonders if they're still doing it.
- outside1234 1mo ago
  Of course they are. They have just put on their Swiss Banker suit now and have all sorts of deflection techniques in place such that, of course, "the money has the stamps that says its clean" (when it it really blood money hidden behind a pretty wall).
- ungut 1mo ago
  OpenAI plainly admitted that it is impossible not to do so in a House of Lords inquiry. So, presumably there is no way around it to train models. There is just not enough non-copyrighted data out there.
  - yorwba 1mo ago
    You mean this one https://committees.parliament.uk/writtenevidence/126981/pdf/ https://committees.parliament.uk/writtenevidence/126981/pdf/ where they write "it would be impossible to train today’s leading AI models without using copyrighted materials"? That doesn't mean they have to download those materials illegally. For a billion dollars, you can easily buy one legal copy of each book in Anna's Archive and still have some cash left over to run a whole-of-internet scraping operation.
    - spwa4 1mo ago
      I'm pretty sure we would know if they did that. And we don't. Plus this is not legal in the EU (and Canada, and ... let's just say the entire rest of the world, and accept that I'll be wrong for one or two smaller countries). Doesn't that matter? Or is only Mistral disallowed from training on copyrighted materials? Je veux ma chaton fat, goddamit!

kmeisthax 1mo ago

I'm pretty sure[0] they're all using shadow libraries, and saying things in favor of them would increase their liability. Furthermore, every pirate wants to be an admiral. None of the big tech companies are actually in favor of any amount of copyright reform. They never have been. There is a huge gulf between "personally benefitting from copyright theft" and "actually wants to legalize the theft". Anthropic still believes they deserve to be paid for their models, they just have this delusion in their head that doing a bunch of computation on stolen data is equivalent to actual human creativity. [0] OpenAI, Anthropic and Facebook have been shown in court to be using shadow libraries, I don't know about Google.

TZubiri 1mo ago

Those are all law-abiding organizations, which AA is not. INB4: "Here is one time one of those organizations broke the law". Don't go there, absolute lowest level of conversation.