site banner

Culture War Roundup for the week of July 27, 2026

This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.

Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.

We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:

  • Shaming.

  • Attempting to 'build consensus' or enforce ideological conformity.

  • Making sweeping generalizations to vilify a group you dislike.

  • Recruiting for a cause.

  • Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.

In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:

  • Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.

  • Be as precise and charitable as you can. Don't paraphrase unflatteringly.

  • Don't imply that someone said something they did not say, even if you think it follows from what they said.

  • Write like everyone is reading and you want them to be included in the discussion.

On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.

2
Jump in the discussion.

No email address required.

Anthropic is destroying old books to build AI.

This is an old story that has been picking up on twitter this week. Here's the Washington Post from January:

Inside an AI start-up’s plan to scan and dispose of millions of books

In early 2024, executives at artificial intelligence start-up Anthropic ramped up an ambitious project they sought to keep quiet. “Project Panama is our effort to destructively scan all the books in the world,” an internal planning document unsealed in legal filings last week said. “We don’t want it to be known that we are working on this.”

Within about a year, according to the filings, the company had spent tens of millions of dollars to acquire and slice the spines off millions of books, before scanning their pages to feed more knowledge into the AI models behind products such as its popular chatbot, Claude.

It turns out that destroying old books is one of the cheapest ways to build an AI dataset. Old books are well-structured texts and (generally) contain quality information; digital archives incur tremendous legal fees and liabilities; copyright risks, in fact, are unsettled and could threaten the entire AI buildout. There's no such problem with maintaining a private library of books: the problem is how to scan them. Physical books have an awkward, blocky shape if your goal is to lay them flat on a reader and extract all the text. The simplest solution is to rip the spine out, lay the pages flat, and run them through a scanner:

The document describes how the scanning company’s “hydraulic powered cutting machine” would “neatly cut” books, whose pages would later be “scanned on high speed, high quality, production level scanners.” Finally, it notes, the scanning company will “schedule with the recycling company to pick up the completed books.”

"Completed books" is such an interesting euphemism, isn't it? You cut the books "neatly," let's not elaborate on that too much, scan them on "high speed, high quality" scanners, narrative detail, narrative detail, then you "schedule with the recycling company". For "the completed books". It's kind of gorgeous.

Ars Technica goes into other details in their article from June 2025, here:

Anthropic destroyed millions of print books to build its AI models

The 32-page legal decision tells the story of how, in February 2024, the company hired Tom Turvey, the former head of partnerships for the Google Books book-scanning project, and tasked him with obtaining “all the books in the world.” The strategic hire appears to have been designed to replicate Google’s legally successful book digitization approach—the same scanning operation that survived copyright challenges and established key fair use precedents.

So Anthropic is really just copying what Google did before, except that Google didn't have to "schedule" the "completed books":

While destructive scanning is a common practice among some book digitizing operations, Anthropic’s approach was somewhat unusual due to its documented massive scale. By contrast, the Google Books project largely used a patented non-destructive camera process to scan millions of books borrowed from libraries and later returned. For Anthropic, the faster speed and lower cost of the destructive process appears to have trumped any need for preserving the physical books themselves, hinting at the need for a cheap and easy solution in a highly competitive industry.

But it was a good deal for Anthropic nonetheless. Destroying books saved them from billions in copyright disputes:

Ultimately, Judge William Alsup ruled that this destructive scanning operation qualified as fair use—but only because Anthropic had legally purchased the books first, destroyed each print copy after scanning, and kept the digital files internally rather than distributing them. The judge compared the process to “conserv[ing] space” through format conversion and found it transformative. Had Anthropic stuck to this approach from the beginning, it might have achieved the first legally sanctioned case of AI fair use. Instead, the company’s earlier piracy undermined its position.

Before continuing, I would like to introduce here one other critical detail I will come back to later:

The court documents don’t indicate that any rare books were destroyed in this process—Anthropic purchased its books in bulk from major retailers—but archivists long ago established other ways to extract information from paper. For example, The Internet Archive pioneered non-destructive book scanning methods that preserve physical volumes while creating digital copies. And earlier this month, OpenAI and Microsoft announced they’re working with Harvard’s libraries to train AI models on nearly 1 million public domain books dating back to the 15th century—fully digitized but preserved to live another day.

So even though Anthropic didn't have to destroy their books, at least they apparently haven't destroyed anything truly rare.

This old story has been all over my twitter this week with a spate of discussion. Maybe this is a clever marketing ploy by Big Elon, because when I went searching for a recent article to corroborate this post I found this PR puff piece from today:

Elon Musk says SpaceXAI will preserve rare books instead of destroying them for AI training

Elon Musk has taken a pointed swipe at Anthropic’s controversial book-scanning methods, saying SpaceXAI will preserve rare volumes rather than damage them while preparing training material for artificial intelligence systems. His comments arrive as questions deepen over how AI companies acquire, digitise and use books to build increasingly powerful models.

There you have it: Anthropic destroyed millions of books, one of the greatest evils we can imagine, to save a few bucks. And now that the public has been informed the great AI companies are going to behave much more responsibly and treasure our collective heritage. Happy, happy!


I would now like to argue that Anthropic was right.

I know that this strikes many as a priori evil. Destroying books for many evokes (or is even the same as) book-burning. And many consider destroying books to be one of the greatest evils imaginable.

So I would like to start by asserting that, before Anthropic destroyed these books, they were basically worthless.

I think people are living in a fantasy. These unread books are not sitting in a warehouse vault somewhere, pristine, untouched, waiting for their future destiny forever home. A wise scholar is not going to stumble across these books in a stony brook and find exactly the wisdom he was seeking. Your future wife is not a perfect passive virgin waiting for you to find her, however long that takes. No, it doesn't work like that. These books are constantly decaying. Everything is constantly decaying. Everything is constantly being destroyed.

By destroying books Anthropic put them to a better use. Destruction is, in fact, the only way anything is ever created.

The Odyssey is current now because of Nolan's latest film. The Odyssey is a beautiful poem that reached us across thousands of years of preservation. One imagines Orthodox monks painstakingly copying everything by hand, vellum and inkwells, Gregorian chants and prayers. It's beautiful. But the less-remembered half is that there were other poems in the Epic Cycle besides The Odyssey (and The Iliad) that did not make it to us. Another six poems are lost. Why? Because readers found them boring. They were not, apparently, beautiful. At least, nobody found them important enough to dedicate their life to saving them. They were never burned or actively destroyed. They just decayed and, though disuse, were lost.

That is the only way anything is ever saved. Art is either alive and part of a vital culture, or it is dead and decays. "Anything not saved will be lost."

And it is necessary to destroy in order to create. The world is choked up with lots of preserved junk. Museums are full of it. Forests burn when you let it pile up too high. Cultures stagnate when too many idols become sacred cows. And it is very easy to make it difficult to build. Every "historic neighborhood" designation and preserved building makes it more difficult to build new homes. Every regulation is an assertion that we would rather curtail some forms of life, that not everything is equal.

I love going to local rinky-dink museums, it's a guilty pleasure of mine, and you know what I invariably find when I go? Old books, lots of them, sitting behind glass shelves where no one is allowed to read them. It's a minor industry. You can go to Mount Vernon or Monticello or the Henry Ford house and everywhere you'll find old dirty books just sitting there. There are companies you can place orders to if you want to open your own old house museum. Books sit there rotting, decor, not used by anyone, nobody minds.

The balance, I think, is that the new has to justify itself as better than the old. It was wrong to destroy old Penn Station in New York, because what replaced it was far less beautiful. It was wrong to let Cleveland and Detroit rot and some of the most beautiful house neighborhoods in America go to grass. I think it was wrong to destroy all the Confederate Statues, because nobody erected a better set of heroes. I remember when ISIS rampaged through Syria and they destroyed the Assyrian Lions. For what, nothing beautiful, and for what.

But I don't think destruction is inherently bad. I think New York is beautiful, London is beautiful, the built environments surpass the natural landscape that preceded them. And I consider what Anthropic is doing to be inherently more worthwhile than a few million burned books, especially if they didn't even destroy anything rare.

There is a load-bearing thought I have been trying to get at that came to me from Loren Eiseley through John Donne -- I believe, forgive me, the reference is not in front of me as I write. It goes like this: "Nature is not all that natural." The world is not in a wild state of nature. I believe in Genesis, I think everything is like a garden, there is no default state of nature that precedes our intervention. Yellowstone is beautiful, it's "natural," but it's made more beautiful because we maintain it. We assert that Yellowstone is more beautiful than the oil pipelines we could build there instead. Tomatoes are beautiful to me, I assert the right to pluck weeds out of the garden, not everything is equal. I eat meat, I think my life is worth more than a cow's, however beautiful a heifer may be. I don't take this destruction casually either, I don't mean this to be glib, I think I have a responsibility to life up to the value of what has been destroyed to give me life. How could it be otherwise? I think any other answer is self-abnegation, the first step toward a kind of selfish selflessness that leads to a suicidal bend.

My argument is that it's right for Anthropic to burn books because they are creating something higher than what they've destroyed. This argument is not reciprocal. It would be wrong if what Anthropic was creating was worse. I think this is just what it means to be alive. This is Genesis, this is the Parable of the Talents, this is the basis of civilization.

I think people think that there are more good books than there actually are.

Looking at e.g., the latest book releases on Standard Ebooks, I recognize none of the new titles, so in some sense that site is already scraping the bottom of the barrel, despite having a catalog of a mere ~1500 books. Gutenberg's at like 78k books (though many are duplicates), and both all the titles and all the authors of the new releases are completely unrecognizable.

At the scale of millions of "books" (many of which aren't really books as the term's commonly used), well... you can randomly sample a url between https://catalog.hathitrust.org/Record/000000001 and https://catalog.hathitrust.org/Record/009999999 to get some sense how nothing most texts really are. It's great that they're digitized, and hopefully no non-digitized books are lost fully, but it would hardly matter if they were. No human will ever read the vast majority of these.

I think people think that there are more good books than there actually are.

Looking at e.g., the latest book releases on Standard Ebooks, I recognize none of the new titles, so in some sense that site is already scraping the bottom of the barrel, despite having a catalog of a mere ~1500 books. Gutenberg's at like 78k books (though many are duplicates), and both all the titles and all the authors of the new releases are completely unrecognizable.

Books can be "good" in a conditional sense.

When I was researching the real world history of elementals for my D&D game, I found Parcelsus' "A Book on Nymphs, Sylphs, Pygmies, and Salamanders, and on the Other Spirits" and "The Comte De Gabelis" invaluable. And while those books are hardly obscure in the right circles, I also don't think most people's lives would be at all impoverished if they never read those books. And yet, I am happy I was able to read them, and would consider it a shame if they hadn't survived to the modern day.

There are plenty of other books that are only conditionally useful for someone researching a very particular thing, and unlikely to be read by the vast majority of the reading public.

Most written texts have been lost. Scholars generally estimate that well over 90% of classical Greek and Latin literature is gone.

I don't know about you but I would certainly like to read the remaining hundred of works by Aristotle that didn't make it, like his treatise on comedy, the 158 constitutions or the Protrepticus.

Sure there's a lot of slop and we're overflowing with slop since printing became ubiquitous, but the reflex to keep everything is understandable given how much actually important mega famous history defining classics we have lost to neglect over the years.

A slightly more solid number (Ctrl-F "commercial value"): As of year 1998, only two percent of copyrights with age of 55–75 years retained any commercial value. That is, after 55 years, 98 percent of copyrighted works are forgotten and worthless.

I think this is one of those moments where you'd want to step in in favor of the scoundrels of books, if you want a principled defense of the books you want saved. It seems like it was only yesterday folks were lamenting the original Dr Seuss books. Do you want to entrust anyone but yourself with the right to decide what is archived, and what is fed into the slopmachine and then wiped?

Do you want to entrust anyone but yourself with the right to decide what is archived, and what is fed into the slopmachine and then wiped?

Unless Anthropic (and other LLM companies) are hunting down every copy of a particular edition of a particular book when they scan it in, I don't see how their behavior would have particular impact on this. Anthropic has negative incentive to do such a thing, because each copy of a book they find costs money to purchase and destroy, but the value they get out of using the book for training only exists for the first copy of the book.

Surely there are some editions of some books out there for which only one or a handful of physical copies remains, which was then purchased by and destroyed by Anthropic (or Google Books or some other mass-scanner), leaving either none or very few which are extremely hard to access by most people (being in obscure archives in obscure libraries). It seems unlikely to me that this set of editions of books has much intersection with the set of editions of books for which keeping at least one copy around is of immense value to humanity; generally, such books also have sufficient cultural value and monetary value that people keep them protected and/or available for display.

If the belief is that humanity ought to indefinitely preserve at least one physical copy of each and every edition of every book that has ever been published, this doesn't seem like the worst goal, but also doesn't seem like a goal that Anthropic or other AI companies (or book scanners) have any responsibility of fulfilling. People who believe that humanity has a duty to preserve at least one physical copy of all these books ought to be the change they want to see in the world.

But that's the point, if you're hoovering up books by the bushel, why use destructive methods and not simply scan them and hand over the discards to people who care about conservation?

Besides the legal obligation to do the stupid thing, of course.

hand over the discards to people who care about conservation

Do these exist? Conservation as in, at a bare minimum, being willing to take the pages (we can leave off "and use appropriate conservation techniques," because that will make an already small group dwindle to literally nothing). And, to be explicit, "post an angry rant about evil AI companies on Reddit" does not count as caring about conservation for the point of this exercise.

Jason Scott is a real person, yes.

And yet, he also resorts to debinding for scanning (or "destructively" scans).

Understand I don't regard the artifact itself to have much value (when it comes to glued magazines anyways, old hardcovers deserve to live even if they're cookbooks). The difference here is the public gets the damn content for free, not a shadow filtered through a subscription based statistical model.

Besides the legal obligation to do the stupid thing, of course.

Unfortunately, I think we live in a world where some legal obligation is the reason for a lot of stupid, destructive things happening, and this is just the latest in a very long list.

I don't know the details, but I heard that destroying books while scanning is cheaper due to the logistics of opening the book and holding it open page-by-page, but also I heard that it was the opposite, so I don't know. If the former is the case, then that provides a good non-legally-mandated-stupidity reason for the destruction as well.

That said, I imagine we're not too far from having the image processing tech required to have something akin to Futurama where Bender (a 30th/31st century android) just flips through a book in 5 seconds like a flip book and then declares that he's read it. Taking a distorted, side-angled photo of a bent/curved page of a book and converting it to a perfect copy of the original page that is literally indistinguishable, even by an AI, from an actual proper scan of the page after it had been removed from the book seems like it will be within the realm of possibility pretty soon, if not already (depending on how literally you take the word "literally"), and once we have that, it's just a matter of having cameras with sufficiently high fidelity and framerates.

That said, I imagine we're not too far from having the image processing tech required to have something akin to Futurama where Bender (a 30th/31st century android) just flips through a book in 5 seconds like a flip book and then declares that he's read it. Taking a distorted, side-angled photo of a bent/curved page of a book and converting it to a perfect copy of the original page that is literally indistinguishable, even by an AI, from an actual proper scan of the page after it had been removed from the book seems like it will be within the realm of possibility pretty soon, if not already (depending on how literally you take the word "literally"), and once we have that, it's just a matter of having cameras with sufficiently high fidelity and framerates.

We're probably already there if we were to invest in expensive machines to do so.

I'm not sure it'd make a difference, the people who complain about the destructive scanning would pivot to complaining about so many ressources being spent that could be used for taking in more refugees being used just to feed the AI slopmachines, and what about all the water being wasted to cool all those scanners?

OP goes over this. Destroying doesn't mean they destroy the page, but scanning works best and fastest if you can get the pages very flat against a scanner. Book spines get in the way, sometimes the text get warped as the spine won't allow the page to lay fully flat against the scanner. Not to mention that you need someone (or a way more complicated machine) to manhandle all these books and spines to get the page to lay as flat as you can for scanning. Scanning a whole book of hundreds of pages that way is a job that would take someone many hours. Compare to cutting the spine, making all the pages loose and stacking them in a document feeder to feed them all one by one, perfectly flat against the scanner. It's not only the fastest, least annoying way to scan, it's the highest quality one. BUT it needs sacrificing the spine.