Hacker News new | ask | show | jobs
by CamperBob2 952 days ago
They aren't worried about people sharing books at this point, or at least they shouldn't be. They're worried about people building their own models.

Here's something I'd pay for (probably with some sort of sketchy crypto, sadly): send me a huge-ass hard drive (or array of drives) with a complete mirror of Z-Library, Library Genesis, and sci-hub on it.

4 comments

That will be a very reasonable thing to expect in a few years. Based on the rate of advancement of memory density and LLM tech I wrote this [0] a while back.

Containment of "the human corpus" (all significant written works since antiquity) is a lost cause for the authorities. Stunts like this Australian thing are symbolic rituals by the dying publishing industry against a downhill battle.

For a brief window you'll have Fahrenheit 451 style police trying to enforce physical warrants and seizure of "illegal reading materials", but proliferation, diminishing size, sheer utility and cost of enforcement will soon make the whole misadventure water under the bridge.

Yeah, we will look back at early 21st century as a sorry bloody episode to be embarrassed by.

[0] https://www.timeshighereducation.com/opinion/2048-informatio...

> Stunts like this Australian thing

You mean OP's Austrian thing? Or something else? Australia's the one with the kangaroos and Austria the one with famous painters.

> Austria the one with famous painters

As an Austrian, I hope you are talking about Klimt and Schiele.

:) I'm afraid that very funny comment might be lost on a few people. Forgive me for being so crude [0].

[0] https://en.wikipedia.org/wiki/Paintings_by_Adolf_Hitler

Just go on Anna's Archive, happy downloading
Not practical, and not safe (as this very article illustrates).
Isn't anyone capable of building their own models technically sophisticated enough to get around basically any normal anti-piracy measure? Or for that matter has already downloaded all the books?
Eventually, someone is going to figure out how to train these things in a distributed fashion. When that happens, the more people who own as much data as possible, the better.
problem is keeping up with the volume of new books/articles coming out daily.