AI Book Explosion on Amazon: Quantity Up, Quality Polarized

Since ChatGPT launched in late 2022, the Amazon Kindle ecosystem has experienced an unprecedented surge in new titles. According to a major new NBER working paper by economists Imke Reimers and Joel Waldfogel, the monthly rate of new e-book releases nearly tripled between 2022 and late 2025. Researchers estimate that roughly two-thirds of these new books are now generated or heavily assisted by large language models.

This is the clearest picture yet of how generative AI is reshaping one of the world’s largest creative markets.

---

The Quantity Boom vs. Quality Divide

The flood of new books is not evenly distributed in quality.

The study reveals a striking polarization:

• New post-LLM authors (those who entered the market after ChatGPT) dominate the surge in volume. However, their books overwhelmingly cluster at the bottom: poor sales ranks, low reader ratings, and minimal engagement. Much of this output can fairly be described as “AI book slop” — generic, low-effort content churned out for quick monetization.
• Pre-LLM (established) authors have not only maintained or increased their output, but have also significantly improved the quality of their new releases. These experienced writers are leveraging AI tools as powerful assistants — for ideation, drafting, editing, and research — while applying their own taste, structure, and storytelling skill. As a result, they continue to dominate bestseller lists and high-rating tiers.

In short: AI is a massive amplifier, but it amplifies talent more than it creates it.

What This Means

The pattern mirrors what we’ve seen in other creative domains (images, music, code). In unskilled hands, generative AI lowers the barrier to entry so dramatically that it floods the market with low-value noise.

In skilled hands, it becomes a superpower that boosts productivity and creative output without sacrificing (and often enhancing) quality.

Amazon’s Kindle Direct Publishing (KDP) platform, designed for easy self-publishing, has become ground zero for this experiment.

While the platform has introduced some limits (e.g., caps on daily uploads and AI disclosure requirements), enforcement remains challenging against the sheer scale of production.

Also read:

• Savor’s Butter from Air: The Next Big Thing in Food Tech or Just Another Eco-Hype?
• Li Auto and the Pioneer Trap: When Being First Becomes a Curse
• Mad Bulls: The Ferocious Activists Who Tear Companies Apart to Rebuild Them
• Ideal Real Estate, Chinese Style — Or Another Startup Idea (Spoiler: Probably Not)
• Amazon Caught in Secret Price Manipulation Scheme with Levi’s, Walmart, Target, Home Depot and More

---

The Bigger Lesson

This NBER study provides rigorous evidence for a nuanced but important truth about generative AI:

Tools don’t replace judgment. They multiply it.

The people who already know how to write good books are now writing more and better ones. The people who don’t are producing mountains of mediocre content that mostly fails to find an audience.

For readers, the signal-to-noise ratio on Amazon has gotten worse in the aggregate, but the best books at the top remain as strong (or stronger) than before. For aspiring writers, the message is clear: AI won’t turn bad writing into good writing — but it can help good writers become exceptional.

The age of abundant content is here. The age of abundant valuable content still belongs to those who know how to wield the tools rather than be replaced by them.