Hacker News Reader

Comment by infecto | Hacker News Reader

Once you start to realize just many books got thrown out annually before “AI” you have less sympathy. Libraries are one of the biggest contributors to this because there is simply too many books that nobody cares about.

I suspect folks are over weighting how much knowledge is being destroyed in these books. If someone actually quantified it that would be amazing but as someone who started going to used book sales at a very young age I just have no sympathy. Most books are worthless. I don’t mean that from a text perspective either.

View on HN

Libraries usually offer those books for sale to the general public. I just picked up two from my local library that way.

AI companies are not doing the same.

AI companies are building the knowledge from the books into their models, which they offer for sale to the general public.

Far more people will use the model than would ever have read the book.

They're not reproducing the information verbatim and with the ability to definitively reference its source. Moreover, not all AI companies are going to get the same books and therefore the same training set, and as companies go bankrupt, training sets are changed, and models abandoned, that information might disappear entirely.

It's not at all the same thing as still having the books themselves available somewhere.

On the other hand, the book can only be in one place and read by one person at a time.

People may not even know the book exists or that it contains the information they want. Obtaining access to a copy may be very difficult, even if it is not particularly rare. The AI is much more accessible.

Then find a way to make them digitally available, verbatim, like Google Books attempted to do. I'm sure with the power that AI companies have they can have legislation amended to make it happen.

You're defending the indefensible here.

Lol, as if they can just get copyright reform passed because they asked nicely for it.

They can't legally publish the digital copies, and you know it. Training is fair use but direct copies are not.

You are defending something that makes no sense. If you want to get mad at something get mad at copyright law.

That copyright law is a mess doesn't mean that companies that buy up rare books, destroy them, and keep the contents locked up are beyond criticism for their actions.

They don't have to scan and destroy these books, it's an entirely voluntary choice.

Name the rare book.

And libraries throw away a lot of books that go to the dump. So what?

You should be pushing for better copyright law not Amazon for buying up worthless books.

Reply on news.ycombinator.com