When Books Become Data: What Are We Losing To AI

Welcome to The Reading Room! I’m so happy you’re here!

There is something unsettling about watching a company that began by selling books now become part of a conversation about destroying them.

Amazon was founded in 1994 and opened its virtual doors in 1995 as an online bookseller. In its early years, the company proudly described itself as the world’s largest online bookstore, eventually embracing the nickname “Earth’s Biggest Bookstore.”

So when reports surfaced that Amazon is purchasing large quantities of books, sending them to a Las Vegas facility, cutting off their bindings, scanning their pages and destroying the physical copies, the irony is hard to miss.

According to recent investigations by 404 Media, a shipment of rare books was tracked to an Amazon facility where employees reportedly described a process of removing book bindings so the pages could be scanned more quickly. The physical books are destroyed in the process. The investigation connected the operation to the collection of AI training data.

And that raises a question that goes far beyond Amazon.

What happens when books stop being treated as books and start being treated as data?

A Book Is More Than The Words Inside It

For a machine, a book can be reduced to text.

For a reader, it can be so much more.

A physical book can tell us when something was published, how it was published, what the cover looked like, who owned it, whether it was signed, how it was marketed and even how the physical object. changed over time.

A first edition isn’t necessarily identical to a later edition. Publishers make corrections. Authors revise texts. Covers change. Introductions are added. Passages can be edited. Formatting can be changed.

When our only surviving record of a work becomes a digital file, those changes can become much harder for ordinary readers to notice.

That’s one reason physical books matter.

They give us an artifact that can be held up against another artifact.

We can compare editions. We can see what changed. We can preserve evidence of what existed before.

Destroying physical copies doesn’t automatically mean that information disappears. A scanned copy may preserve the words, but preservation isn’t only about preserving words.

It’s about preserving history.

And These Aren’t Necessarily Replaceable Books

This is where the situation becomes especially uncomfortable for book lovers.

The books being purchased aren’t necessarily recent bestsellers sitting on the shelves of every. bookstore. Reports have described bulk purchases involving older, obscure, out-of-print, and rare titles. Some books may exist in very limited physical quantities.

Once the last — or one of the last — physical copies is destroyed, we can’t simply print another original.

We might be able to reproduce the text. We might be able to reproduce the cover. We might be able to create a digital version, but a reproduction isn’t the same thing as the original artifact,

That’s an important distinction when we’re talking about books that may have a cultural, historical, or scholarly significance.

The Amazon of Today vs. The Amazon That Started With Books

There’s an almost poetic contradiction here.

Amazon built its identity on books.

It began as a place where readers could find books they couldn’t easily find anywhere else. Its enormous catalog was one of the things that made the company revolutionary.

Amazon’s own early announcements emphasized its massive selection and its ability to connect readers with books on incredibly specific subjects.

Today, the company is one of the world’s largest technology companies — and AI is a major part of its future.

But somewhere along the way, the relationship between Amazon and. books changed.

The question is whether books are still being valued as cultural objects or whether, in some contexts,, they are simply being valued for the information they contain.

Because once a book becomes nothing more than a collection of words to be extracted, scanned, processed, and discarded, we’ve lost something.

What Happens When the Original Disappears?

Digital books are convenient. They're accessible. They're searchable. They're incredibly useful.

But digital isn't automatically permanent.

Files can be updated. Editions can be replaced. Platforms can change. Access can disappear.

And sometimes, changes to digital texts can happen without the average reader ever realizing that the version they're reading isn't exactly the same as the one that existed before.

Physical books give us something digital files can't quite replicate: a fixed historical snapshot.

A physical copy sitting on a library shelf or in someone's personal collection can show researchers exactly what that edition looked like and what it contained.

That's why libraries, archives, museums, and collectors preserve physical books even when the same text is available digitally.

The physical object is part of the record.

Progress Doesn’t Have to Mean Destruction

None of this means AI research shouldn't happen.

It doesn't mean books shouldn't be digitized.

In fact, digitizing books can be an incredible tool for preservation. It can make rare texts accessible to people who could never physically visit a particular archive or library.

The problem isn't necessarily the scanning.

It's what happens afterward.

There is a huge difference between:

• “Let's digitize this book so more people can access and preserve it.”

and

• “Let's buy this book, remove its contents, extract the text, and destroy the physical object.”

Those are two very different approaches to knowledge. And perhaps that's the conversation we should be having.

Not simply whether AI can learn from books. But whether we are comfortable allowing the physical record of our literary history to become disposable in the process.

So, What Do We Do With This?

As readers, we probably can't stop technological progress.

But we can remember why physical books matter.

We can support libraries. We can support independent bookstores and booksellers. We can donate books to collections when appropriate.

We can preserve older editions instead of automatically assuming that everything worth keeping can be stored in a cloud.

And most importantly, we can keep asking questions.

Because a book isn't just data.

It is someone's words. Someone's research. Someone's imagination. Someone's history. And sometimes, the physical copy sitting in front of us is itself a piece of history.

Amazon started with books.

Maybe that's exactly why this story feels so significant.

A company that once built its identity around helping people find books is now being accused of participating in a process where some books are being found, scanned, and destroyed.

And perhaps the question we should all be asking is:

In our rush to teach machines everything we've written, are we remembering to preserve the things that taught us in the first place?

Until next time, keep turning the page 🤍