Notes from the Shelf

The Current AI Book Burning Crisis

Right now, there are AI companies that are buying books and then destroying them once AI scans them, to prevent “AI slop.” There are even some companies who are acting as intermediaries between the booksellers and the AI companies. This way, the AI companies get their hands on the books while protecting their privacy, so some of this is being done in secret.

I heard this at first on TikTok, and I wanted to do some research into it to verify, and it’s definitely happening. I found articles for The Netherlands, Australia, Silicon Valley, so this seems to be a phenomenon across the world.

According to Fortune, a Dutch bookseller received an email requesting for 3,000 books to be sent to China. He dismissed the email as spam but found out later it was a legitimate order once he talked to other Dutch antiquarian booksellers and realized they had received the same requests.

According to The Guardian, some Australian booksellers, some dealing in rare or antique books, received large orders for a Canadian company called Zoom Books, which describes itself as a book recycler. The orders were odd because they asked for newer books alongside very old rare books, and they didn’t seem concerned on the costs of the books or of shipping overseas.

This is straight out of a dystopian novel. Now, people say they’re not destroying every copy of that book, but if multiple companies are doing this, what’s to stop a book from being snapped up and totally destroyed?

There are a lot of books that are now out of print, so there’s only so many copies in the world, and some of these are antique or rare copies of books. Also, it’s not like the people ruining these books care to ensure that the book isn’t lost to history. They just want to get whatever they can to feed their AI machines, so they’re not doing due diligence to make sure they don’t wipe out the only copies of a certain book.

Plus, these articles do state that some of the books they’re buying up are rare copies, so they could be destroying one of a handful of copies in existence. No one is tracking across companies on which books are being destroyed and how many are left in the world.

Once that book is gone, if it’s only digitized, then the words in it could be changed, or it’s just totally lost to history and people forget about the knowledge.

I’m viewing this as the burning of the library of Alexandria. Think of all the knowledge there where it was the only copy and it was just lost, never to be recovered again. What did that do to our societies? What technology did we lose in that? And what is the outcome going to be from this AI book purge?

I don’t think it’s far-fetched to assume we’re going to lose a lot of knowledge to this.

What Can You Do?

My first piece of advice is stop using generative AI. I do think AI has its uses in certain industries, like for instance it’s being used in the packaging industry to flag misprinted boxes in real time on the line, but generating a paragraph for people is not one of the good uses. It’s totally unnecessary.

Also, they don’t need to be buying up millions of copies of books to scan and destroy to train their generative AIs. They could train them organically with people’s use of it, like how translation softwares work.

I used Trados in college, and it’s an LLM. You first put a dictionary file into it, so it has all the base words for whatever language you’re translating, in my case Japanese to English. Then, as you use it to translate, it saves the word pairings you did, so later it can start flagging phrases you’ve translated in the past to make it easier in the future. That’s an LLM, and it doesn’t require millions of books to be destroyed to use it. It requires the user to correct as they use it, and it saves it for later. The buying of millions of books to scan and destroy just to feed an AI machine is insanity.

Second, spread the word. The AI companies were trying to do this in secret, but the news broke a week or two ago, so keep talking about it. Also, they can scan the books without totally destroying them, they just choose not to, so put pressure on them to preserve the books they scan.

Third, start building your own physical library. Hoard the knowledge that you want to keep. I know books can be expensive, but there’s also cheaper ways to get books, especially if they’re older copies.

Check out your library book sales. I know my library has them every quarter, and they also have a book nook to buy books on the shelves open all year. These books can be as little as a dollar, and usually they’re not more than $5 from what I’ve seen.

Check out your local booksellers, see if they sell used copies. You can also check out sites like thriftbooks.com and bookoutlet.com. They usually have good deals.

Also, sites like bookfinder.com allow you to see all the listings where that book is currently for sale, like on eBay. Slugbooks.com is also a great site, but it’s more for textbooks. Some nonfiction does show up when you search as well, just not really any fiction books.

So, create a list of books you don’t want to see destroyed and buy them when you can.

Banned Book Recs

Here’s my list I would recommend. Most of them I already have but I’m still on the hunt for a few of them. I’ll mostly stick with nonfiction, but there’s some fiction banned books as well.

Nonfiction:

Your country’s constitution or other governing documents. One of the first things I bought when Trump became president again. They’re not going to change the wording on my physical copy. It doesn’t matter what country you’re in, you should have a copy of your government’s rules so they can’t change it without your knowledge. Or if you can find it online, that’s fine, just make sure to print it off so you have a physical edition.

• How to Hide an Empire: A History of the Greater United States by Daniel Immerwahr
• On Tyranny by Timothy Snyder (Everyone should read this and have a copy of it.)
• Invisible Women: Data Bias in a World Designed for Men by Caroline Criado Perez (Another one everyone should read. It’s eye-opening and will ignite a rage inside of you.)
• The New Jim Crow: Mass Incarceration in the Age of Color Blindness by Michelle Alexander
• The Color of Law: A Forgotten History of How Our Government Segregated America by Richard Rothstein
• A Queer History of the United States by Michael Bronski
• A People’s History of the United States: 1492-Present by Howard Zinn
• Good People Break Bad Laws: Civil Disobedience in the Modern Age by Topher Field
• Rules for Radicals: A Pragmatic Primer for Realistic Radicals by Saul D. Alinsky
• Erasing History: How Fascists Rewrite the Past to Control the Future by Jason Stanley

Fiction:

• 1984 by George Orwell
• Brave New World by Aldous Huxley
• Fahrenheit 451 by Ray Bradbury
• The Handmaid’s Tale by Margaret Atwood
• All Boys Aren’t Blue by George M Johnson
• People Kill People by Ellen Hopkins (a story about gun violence and white supremacy)