381–390 of 961 posts
The main question is why aren't they leaking it to AA themselves? Trying to keep an edge with their training sets? Isn't it ridiculous, considering the sheer size of them and statistical insignificance of the set differences?
Because that’s illegal.
I was surprised to read that Anthropic (and probably other data / model companies) are doing this and it's extremely disappointing, as working towards the benefit of humanity is not an exclusive right / domain of theirs but rather, is a shared responsibility and mission carried by all of humanity itself as a collective responsibility. Thus, the preservation of this knowledge, its availability, and accessibility are t…
If a company is scanning material protected by copyright, it has to send a digital copy of the scanned material to Library of Congress within 5 working days.
It is not a big deal. Since the invention of the printing press any important book has been duplicated by thousands, tens of thousands or even million of units. Just taking one of those and "destroying them"(it is not destroyed, a digital copy with the ability of doing millions of copies is stored somewhere) is not problematic for Humanity. By the way, I always search for second hand books. Most of the books there ar…
somewhere were we can't access it. as the article states: “permanently locking human knowledge inside private corporate servers”
the issue isn't that the physical copy is gone, it's that they are preventing people from making digital copies that are actually accessible by destroying the physical copies.
> If you go to a recycling centre, the garbage to quality ratio is over 100 or more.
archivists keep everything, because we don't know right now what will be important 100 years from now.
It is not a big deal. Since the invention of the printing press any important book has been duplicated by thousands, tens of thousands or even million of units. Just taking one of those and "destroying them"(it is not destroyed, a digital copy with the ability of doing millions of copies is stored somewhere) is not problematic for Humanity. By the way, I always search for second hand books. Most of the books there ar…
Read more about this. Depends on the definition of the "big deal" but from what I can understand the problem is that they buy rare things - which exist in just several copies - and they tend to buy _all_ copies.
We've been seeing that headline for a few weeks now and I really don't understand the problem. Companies are paying for the books now, great! And destroying a book is really the best way to scan it. I could destroy the books I buy myself if I wanted, they're my books and I can do whatever I want with them. A book is really just a stack of paper, I don't get the sacred feeling attached to it. Especially since books ar…
> A book is really just a stack of paper, I don't get the sacred feeling attached to it. Especially since books are printed in thousands to millions of identical copies. If you've read the articles covering this issue, you'll be aware the concern is over the fate of rare and out of print books, rather than your straw man (ie those available in 'thousands to millions' of copies). https://www.bbc.com/news/articles/cp3r…
> "It would be a much more significant problem IF one like that were to be bought for destruction, having survived this long."
I agree with the sentiment of course but it is really a huge IF they are doing that.
IF they wanted to train on, say, Leviathan by Thomas Hobbes, why buy an expensive edition from the 1600s when they would get the same text from a Penguin edition for a fraction of the price? It gets much cheaper secondhand too of course.
I'll go further, why would they want to train on expensive rare and out of print books? Are they, perhaps, competing on an AI benchmark based on extensive medieval knowledge of the cosmos? There's been a lot of pearl-clutching about lost obscure knowledge but y'all really reckon that kind of knowledge is valuable to LLMs?
Earlier quoted context omitted.
Rare and out of print does not mean important or valuable.
Usually it means the opposite. Books that are old and valuable tend to be out of copyright, so they do see new printing runs.
This idea that there can only be merit in a work if it's commercially viable is incredibly ignorant, and if that becomes the standard for whether a work is preserved or not, we stand to lose a great deal of our cultural heritage.
I was surprised to read that Anthropic (and probably other data / model companies) are doing this and it's extremely disappointing, as working towards the benefit of humanity is not an exclusive right / domain of theirs but rather, is a shared responsibility and mission carried by all of humanity itself as a collective responsibility. Thus, the preservation of this knowledge, its availability, and accessibility are t…
A small change in the copyright law would fix this problem. Something like: If a company is scanning material protected by copyright, it has to send a digital copy of the scanned material to Library of Congress within 5 working days.
Earlier quoted context omitted.
Read more about this. Depends on the definition of the "big deal" but from what I can understand the problem is that they buy rare things - which exist in just several copies - and they tend to buy _all_ copies.
Where did you see they tend to buy all the copies? This comment is the first time I've heard of this.
It is not a big deal. Since the invention of the printing press any important book has been duplicated by thousands, tens of thousands or even million of units. Just taking one of those and "destroying them"(it is not destroyed, a digital copy with the ability of doing millions of copies is stored somewhere) is not problematic for Humanity. By the way, I always search for second hand books. Most of the books there ar…
Read more about this. Depends on the definition of the "big deal" but from what I can understand the problem is that they buy rare things - which exist in just several copies - and they tend to buy _all_ copies.
But where did you hear that they’re buying “all copies”? And to what end?