Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Meta VR Glasses(meta.com)
    35comments
  2. Linux support is coming to Snapdragon X2 Series(qualcomm.com)
    39comments
  3. Claude discovers a novel enzyme system with CRISPR-like repeats(anthropic.com)
    503comments
  4. VSCode's SSH Agent Is Bananas (2025)(fly.io)
    71comments
  5. The "Windows XP Box" (2003)(mini-itx.com)
    2comments
  6. We just shipped support for the ugliest part of HTTP: Vary(cloudflare.com)
    4comments
  7. Fixing the Portobello Police Station Clock(pointinthecloud.com)
    84comments
  8. ArXiv receives multiyear commitments to support it as an independent nonprofit(arxiv.org)
    discuss
  9. Mercury 2.5 LLM hits 770 tokens per second(artificialanalysis.ai)
    9comments
  10. LensVLM: Compressing long context as images, expanding only relevant pages(huggingface.co)
    5comments
  11. Italian parliament votes for return to nuclear energy(apnews.com)
    354comments
  12. The mystery animal on an ancient god's head(signoregalilei.com)
    14comments
  13. A brief history of Windows scroll bar shortcuts(devblogs.microsoft.com/oldnewthing)
    45comments
  14. Gemini 3.8 text-to-speech(blog.google)
    118comments
  15. Radicle: Disclosure of Vulnerability in the Network Protocol(radicle.dev)
    45comments
  16. Making Tailscale Faster(tailscale.com)
    16comments
  17. The Curious Power of Punctuation(newyorker.com)
    2comments
  18. Tokens too cheap to meter(jyn.dev)
    178comments
  19. I don't want the details(michaelheap.com)
    193comments
  20. Swap, ZRAM, Zswap and Hibernate on NixOS(matthewbrunelle.com)
    6comments
  21. Z80 REPL (2018)(abagames.github.io)
    18comments
  22. Show HN: I built a post-mortem debugger for native Windows x64/x86 crashes(forensicdbg.com)
    4comments
  23. Claude Code reads AGENTS.md only when telemetry is on [fixed](szypowi.cz)
    250comments
  24. 28% of job postings on company career sites have been open over 90 days(unlisted.careers)
    286comments
  25. Once Claude can measure something, it can make it faster(claude.dev)
    92comments
  26. A refined phylochronology of the second plague pandemic in Western Eurasia(pnas.org)
    discuss
  27. Stripe's Knowledge AI Platform(stripe.dev)
    108comments
  28. Seattle City Council votes to ban surveillance pricing in sale of groceries(consumerreports.org)
    187comments
  29. UK military jamming other nations' satellites to defend itself, BBC told(bbc.com)
    177comments
  30. GPT-6 Sol and Luna(openai.com)
    822comments

Bulk Orders Surge at Used Bookstores in Japan; 50 Tons Sent to U.S.

19 pointsby 2h agonews.ntv.co.jp
16 comments
1h agoHN ↗

So they ran out of information on the internet...

1h agoHN ↗

or more like they can't use the internet anymore because it's just slop now

1h agoHN ↗

They ran out of free information on the internet. They can't steal it just like that anymore because the providers are now aware of the value they could provide and could sue the now trillion dollars companies for stealing the data.

They didn't bother when it was a non-profit sponsored by Elon Musk.

1h agoHN ↗

Via google translate:

It has been discovered that online used bookstores across Japan have been receiving a surge of large orders for books since around August of this year. Interviews with these bookstores reveal reports of "100 books sold per day" and "days where sales have increased fivefold," leading to widespread speculation within the industry that the orders are intended to collect training data for generative AI (artificial intelligence). Further investigation revealed records of over 50 tons of books being exported from Japan to the United States. Is it acceptable for books to be consumed and discarded for AI training?

[...]

Nippon Television investigated using "Sayari," a tool that analyzes import and export data, and confirmed records that a group company of this major Japanese book distributor exported more than 50 tons of "JAPANESE BOOKS" to the US since last year. Assuming that all the books were heavy hardcovers (calculated at 500 grams), this would amount to the equivalent of 100,000 books.

1h agoHN ↗

Buy as many books as you want but please don't burn after scanning.

1h agoHN ↗

Unless somebody steps in and prohibits it they will do it, because somehow that's most cost efficient.

1h agoHN ↗

Yeah. It somehow keeps the competitors at bay, apparently, from previous discussions about AI companies who burn books.

Can we really trust these AI companies when they're basically assimilating human art and culture and everything, only to go on and destroy the evidence?

1h agoHN ↗

I don't understand why can't they donate the books somewhere. Even if they don't want to share the scans, why not give up the books?

1h agoHN ↗

They can, but it would cost more and take longer, so they don't.

It's not as though they particularly care, and it's not as though bad press matters with a captured regulatory environment.

1h agoHN ↗

Largely, because books are heavy, and there's limited demand for most used books. If you've ever been at a book sale on the last day, there are always lots of leftover ones that no one wants.

There's a general sense that librarians are preservationists. This is far from the truth. Most books get pulped within a few years of printing. Librarians are constantly discarding books that don't get checked out to make space for new books.

1h agoHN ↗

In the name of efficiency they cut the spines, rip the pages apart for easier scanning, books are literal dead wood which would cost a lot in shipping so it's easier and cheaper to just discard or burn the remnants.

The book ceases to be a book right at the beginning of the extractive process.

1h agoHN ↗

What do you want them to do with the books after they’ve digitized them?

1h agoHN ↗

Give them away. Put them on the street for people to pick up. Donate them to anyone. I can't tell if you're being serious or if your comment is sarcasm. I'm leaning toward the latter.

54m agoHN ↗

You want them to put them on the street in front of the warehouse? Do they post them on free stuff Facebook group after or just wait for the garbage truck?

31m agoHN ↗

Isn't the "burn after scanning" due to copyright? It seems like destructively scanning was an explicitly allowed method in Bartz v Anthropic as transformative fair use.

34m agoHN ↗

I'm strongly against destroying books to train AI. That said, I'd be extremely interested in exploring the contents of Japanese books, which probably have a lot of stuff not available outside Japan, through a LLM.