AI and the Quantity and Quality of Creative Products: Have Llms Boosted Creation of Valuable Books?

Quick Overview

The AI revolution in publishing has significantly increased the sheer volume of creative products, specifically books, with AI-assisted authors producing 172 new titles per month in 2025 alone, yet this massive increase in quantity has not translated to higher quality, as evidenced by the fact that 51% of the sample books had zero ratings, suggesting the market is flooded with low-value content, even though established authors are using AI to boost their output.

Key Points: The volume of AI-assisted book creation reached 172 new titles per month by mid-2025, significantly tripling compared to the pre-AI era. Despite the volume increase, the average quality of books, measured by ratings, dropped significantly, with 22% of the sample having zero ratings. Established authors are leveraging AI tools to boost their productivity, increasing output but maintaining high quality in their own mid-tier books. The NBER paper suggests that the flood of low-quality AI-generated content negatively impacts consumer surplus by creating noise that obscures valuable human-created works. The US Copyright Office ruled that AI outputs without significant human expressive elements are not copyrightable, creating a clear distinction between AI tools and human authors. The study analyzed 8 specific subcategories, finding that genres like romance, economics, and women's sleuths showed massive AI production, while established authors maintained quality. The success of AI-assisted content relies on the human operator's skill to edit and direct the AI, otherwise, the output is low-value.

Context: The discussion centers around a working paper from the National Bureau of Economic Research (NBER), authored by K. Rymer and J. Waldvogel, which examines the impact of Large Language Models (LLMs) on the publishing industry, specifically focusing on the quantity versus quality of creative products like books. The paper compares book metrics before and after the widespread adoption of generative AI tools, noting the massive increase in supply and questioning whether this abundance correlates with increased consumer value or simply an increase in low-quality, AI-generated content.