301–310 of 761 posts
Earlier quoted context omitted.
Right, the settlement doesn't. However, the judge already ruled on the only important piece of this legal proceeding: > Alsup ruled in June that Anthropic made fair use of the authors' work to train Claude...
Which is very important for e.g. the NYT lawsuit against OpenAI. Basically there’s now precedent that training AI models on text and them producing output is not copyright infringement.
Earlier quoted context omitted.
Afaik to scan a book you need to destroy it by cutting the spine so it can feed cleanly into the scanner. Would incur a lot of fines.
Nah, that's just if you want archival-quality scans. "Good enough for OCR" is a much lower bar.
Earlier quoted context omitted.
> But nobody was ever going to that If this is a choice between risking to pay 1.5 billion or just paying 15 mil safely, they might.
Option 1: $183B valuation, $1.5B settlement. Option 2: near-$0 valuation, $15M purchasing cost. To an investor, that just looks like a pretty good deal, I reckon. It's just the cost of doing business - which in my opionion is exactly what is wrong with practices like these.
What's actually wrong with this?
They paid $1.5B for a bunch of pirated books. Seems like a fair price to me, but what do I know.
The settlement should reflect society's belief of the cost or deterrent, I'm not sure which (maybe both).
This might be controversial, but I think a free society needs to let people break the rules if they are willing to pay the cost. Imagine if you couldn't speed in a car. Imagine if you couldn't choose to be jailed for nonviolent protest.
This isn't some case where they destroyed a billion dollars worth of pristine wilderness and got off with a slap on the wrist.
Earlier quoted context omitted.
This implies training models is some sort of right.
No, it implies that having the power to train AI models exclusively consolidated into a handful of extremely powerful companies is bad.
Earlier quoted context omitted.
> handsomely profiting Well actively generating revenue at least. Profits are still hard to come by.
Operating profits certainly but if you include investments the big players are raking it in aren't they?
Earlier quoted context omitted.
Paying $3,000 for pirating a ~$30 book seems disproportionate.
I feel like proportionality is related also to the scale. If a student pirates a textbook, I’d agree that 100x is excessive, but this is a corporation handsomely profiting off of mass piracy. It’s crazy to imagine, but there was surely a document or slack message thread discussing where to get thousands of books, and they just decided to pirate them and that was OK. This was entirely a decision based on ease or cost,…
Earlier quoted context omitted.
Thank you very much. There seems to be a lot of friction in this seemingly simple process…
For what it's worth the friction exists for a reason, conflicts of interest. The lawyers suing Anthropic here will probably walk away with several hundred million dollars - they have won the lottery. If they managed to extract twice as much money from Anthropic for the class, they'd walk away with probably twice as much... but winning the lottery twice isn't actually much better than winning the lottery once. Meanwhi…
Earlier quoted context omitted.
> Buying used copies of books, scanning them, and training on it is fine. But nobody was ever going to that, not when there are billions in VC dollars at stake for whoever moves fastest. Everybody will simply risk the fine, which tends to not be anywhere close to enough to have a deterrent effect in the future. That is like saying Uber would have not had any problems if they just entered into a licensing contract wit…
Anthropic literally did exactly this to train its models according to the lawsuit. The lawsuit found that Anthropic didn't even use the pirated books to train its model. So there is that
Earlier quoted context omitted.
Anthropic literally did exactly this to train its models according to the lawsuit. The lawsuit found that Anthropic didn't even use the pirated books to train its model. So there is that
The lawsuit didn't find anything, Anthropic claimed this as part of the settlement. Companies settle without admission of wrongdoing all the time, to the extent that it can be bargained for.
Obviously, that's not part of the current settlement. I'm no expert on this, so I don't know the extent to which the earlier ruling applies.