Hacker News Reader

Comment by reinitctxoffset | Hacker News Reader

How is distillation an "attack" but gigascraping the Internet to the point of crashing servers and everyone needs Cloudflare and Anubis now not an "attack"?

I'm not aiming for a what about kickflip here: I'm saying we need to either agree on some rules or stop crying foul. Maybe the coherent legal theory is that neural networks and intellectual property don't interact. That would be weird but it would be consistent, a market could price it, I could do coding stuff and know if I was illegaling.

But this weird gerrymander that no judge will really rule on in an emphatic way is like, bad for the planet, bad for markets, bad business.

There are a lot of reasons to look forward to DeepSeek Huggingface drop kicking the unambiguous frontier weights in like, November, but I think my favorite one will be "who's distilling now bitch?"

View on HN

I think you've basically got the legal theory. Training a neural network isn't prohibited by copyright law so if you can legally get your hands on something (e.g. by sending a GET request to someone with rights to serve the contents of their web page, or by buying a book) without signing a contract to not train on it, you can train on it.

But the American AI companies only let you query their models if you first sign a contract to not train on the output.

It's hypocrisy and unfair, but I think there's a strong legal argument for it.

Of course China can simply decline to assist in enforcing that contract... But I would expect US courts to do their best to.

Contract law is never going to prevent this.

Why not? Seems like a perfectly normal contact term to me.

Or do you just mean that US courts don't have enough teeth to prevent Chinese companies from violating contracts? On that I agree.

I mean the latter, but more narrowly: China would never allow the United States to have a monopoly on machine intelligence if the only thing standing in the way of a domestic alternative was the Anthropic ToS. In general, I think that China is willing to agree on certain things relating to intellectual property. But not on this, it’s too big.

The US is already publicizing the way they are using Claude with Palantir for war gaming purposes. It’s a matter of national defense. Contract law has no meaning here.

Even if enforced do you know what the penalty would be? Not much!

> but I think there's a strong legal argument for it.

Maybe today. I doubt it tomorrow. Legal and not legal, largely, has to answer to the population sooner or later. Ultimately, humanity decides legality. And I don't think the frontier labs will get a pass from humanity in the midterm, let alone the long term. I think you'll see the rules change towards something more "intent" driven. And there's absolutely no difference in intent between Frontier labs and everyone chasing them.

Frontier labs just want the door closed behind them, as do their investors, because they know the money will never be recouped if others can do the same magic tricks.

Eh, I think you've done a pretty good job summarizing a collection of settlements with a few narrow bench rulings for seasoning. I'm not sure I follow you to it being a coherent legal theory. Buying a book in a bookstore is sure legal, and excerpting from it for e.g. literary criticism is pretty settled. Downloading every torrent of all e-books ever is pretty clearly illegal (or at least it fuckin would be if I did it). Pretty sure like, multiple labs have been popped for that though.

Situation right now seems more like a fragile detente: if you got a Hill staffer drunk and hounded him long enough he'd probably be like "God damnit the market will fucking tank if we don't get these two IPOs out north of a trillion. And don't even get me started on how I'm going to sell Chinese AI to a Senate that still calls people Nipponesians when no one is looking. We're doing the best we can alright, get off my back man."

We have a situation, but it's not exactly A&M Records, Inc. v. Napster.

> Downloading every torrent of all e-books ever is pretty clearly illegal (or at least it fuckin would be if I did it). Pretty sure like, multiple labs have been popped for that though.

Oh it is, and at least anthropic has paid $1.5 billion and deleted there torrented copies and not released any models derived from them as a consequence.

The thing is it turns out to be not that expensive to just buy a copy of every book legally and scan them. And there's even precedent that this is legal predating LLMs (Google books)

> and deleted there torrented copies and not released any models derived from them as a consequence.

I have a bridge to sell you

Great, let's go down to the courthouse and get some sworn testimony as to the ownership, value, condition, and so on and so forth of the bridge. And some document review and discovery run through professional legal firms under the same conditions. And perfectly reasonable and verifiable explanations as to why you own the bridge and are selling it (namely that you bought a copy of literally every book in existence in the meantime).

Facts are in fact knowable, and the US legal system is in fact not terrible at getting to them.

I think you're right to point out that historically the rule of law in the United States has been very robust by the standards of whatever era, it's been a tremendous advantage in attracting business and capital and talent, it's good stuff.

But we've gone through some pretty weird times too. Turn of the last century was pretty tech billionaire edits, reconstruction was uh, not smooth, it's a mixed bag.

And most takes I hear seem to acknowledge that this is one of those weirder times: serious election fraud rhetoric from most everybody from 2016 to the present, very politicized courts (on both sides to be clear), very soft on anti-trust, very soft on adventurous accounting. The Epstein files and like, no consequences (pretty much uniquely for a developed nation with Epstein people). It's weird right now.

And I think I would be hard pressed to think of a weirder part of this weird time than the rule of law meets AI. We can haggle on where laws end and norms begin (stare decis being maybe the midpoint), but in the 90s, the Justice Department got their brass knuckles on for a lot less.

I don't think it's a simple "the law works nothing to see here" story.

I broadly agree with your take on the state of the US - but this is a case where given the specific facts at hand I'm confident it still got to the truth.

I can understand why as someone who didn't follow it and the more corrupt legal developments closely you wouldn't be confident in that.

> It's hypocrisy and unfair, but I think there's a strong legal argument for it.

That right there is the problem.

> to someone with rights to serve the contents

Now THAT'S doing some heavy lifting lmao. The vast, vast, VAST majority of the original datasets were from pirated books and the like. Also, arguably a robots.txt is the exact mechanism to follow to do the mass GET-ing, yet the AI cos choose time and time and time again to simply ignore it and be as abusive as they possibly fucking can

> The vast, vast, VAST majority of the original datasets were from pirated books and the like

And there's been significant legal consequences as a result

> Also, arguably a robots.txt is the exact mechanism to follow to do the mass GET-ing

You're free to argue this of course, but the courts have largely rejected it already pre LLMs. See for example hiQ Labs v. LinkedIn

Yes, anthropic and openai have really been brought to their knees and ipos cancelled because of the legal consequences of obtaining their training data.

This would have been a problem but it turns out that Anthropic is actually valued multiple orders of magnitude more than a copy of all the books in the world. So they survived the significant legal consequences.

In a just world, the punishment can be more than just the sum of the direct damages, otherwise there's no incentive to stop reoffending.

Anthropic (and friends) proved they're willing to do obviously illegal things, and it didn't end them. Why do we think they stopped after doing it once?

In this world too, the punishment was, ballpark, 100x direct damages.

> In this world too, the punishment was, ballpark, 100x direct damages.

Which ballpark are you playing in to come to a number like that? I have friends whom are authors, and they've certainly not seen a penny from Anthropic. I somehow doubt they're the only ones out there that haven't been compensated for having their work blatantly stolen. In a just world we'd be ensuring Anthropic was destroyed as a result of their actions, since their entire existence hinged on en-masse piracy.

Anthropic paid (or is paying, not sure the schedule for class members receiving payouts) 3000 per work they were found to have infringed - 30 is a fair approximation of the cost to buy a book.

And the typical book only sees 1 copy sold?

No, the damages however are the cost of a single book sale (actually just the profit from one) because the only thing that anthropic did that broke the law was pirate it instead of buying a copy (as found by the court...)

In a just world, copyright should be abolished with extreme prejudice should it's continued existence prevent humanity from developing aligned artificial super intelligence.

I disagree on so many levels that it would be impossible to have a reasonable debate. To give only my most charitable counterpoint:

(If, big if*) successful AGI will be such an accelerant to human thought, it should be able to reproduce every human work, every scientific theory, etc. without any training data, in a decade.

All we have ever done as humans is look at the world and ourselves, and then create new data out of what we've seen[0]. Now imagine how much parallel compute we could throw at doing the same thing.

[0] https://m.youtube.com/watch?v=f-XyJ3GlrWw

* Big, enormous, trillion dollar if.

The penalties for breach of contract are much less than for copyright infringement. Breaching contracts and paying the penalties is often a viable business strategy. You're supposed to make the other side whole for any losses they incurred due to you not following it.

Knowlege should not have ownership. Training and distillation should be allowed

Granting people some form of control over knowledge only serves the public interest inasmuch it provides incentive to create more of it. Mass media, effortless duplication, and copyright extensions had already broken this to the point where control of knowledge was suppressing creation of new knowledge more than it facilitated.

The world has changed, we need a mechanism that works for the public interest that applies to the facts as they now are.

Reply on news.ycombinator.com