Rewritten Title: The AI Paradox: How the Technology That Feeds on Books Is Reshaping Global Reading Culture
Article:
A Global Reading Recession
A stark warning is emerging from academic circles in the US and Europe: the very technology that relied on books for its growth is now contributing to a decline in reading. This paradox presents a significant challenge to cultural and intellectual development worldwide. The trend is not isolated; even in nations with rich literary histories, the habit of dedicated book reading is under pressure.
The Challenge of Measuring a Cultural Shift
The situation in Iran highlights the complexity of measuring this shift. During a recent press conference for the 33rd National Book Week, officials highlighted a lack of current, transparent data. Ebrahim Heidari, the secretary of the event, cited the most recent official study rate from 2016, which was 12.2 minutes per day. He noted that a subsequent study, completed in 2023, has yet to be made public, pointing to challenges in accurately assessing this key cultural indicator.
Azadeh Nazarboland, Secretary General of the Public Libraries Institution, further clarified that study rates are a multi-dimensional metric that cannot be solely derived from library lending statistics. This nuanced understanding underscores the difficulty in capturing the full picture of a nation’s reading habits.
The Perfect Storm: Technology and Shifting Habits
The decline in reading is a global phenomenon. Professor Naomi Baron, a linguistics expert at American University who has long researched technology’s impact on reading, warns of a “perfect storm” for book reading. “AI arrived precisely when children and adults were already spending less time reading books,” she states.
Supporting data is alarming. In the United States, the National Assessment of Educational Progress shows the percentage of fourth graders who read for fun almost daily has plummeted from 53% in 1984 to 39% in 2022. The figures for eighth graders are even more dramatic, dropping from 35% to a mere 14%. Similarly, a 2024 survey in the UK by the National Literacy Trust found that only one-third of children and young people aged 8 to 18 enjoy reading in their free time.
The Core Paradox: AI’s Appetite for Books
The emergence of advanced artificial intelligence adds a complex layer to this issue. Ironically, the large language models (LLMs) that power today’s AI were built by ingesting massive volumes of textual data, including millions of books.
- GPT-4 (OpenAI): This model was trained on 13 trillion tokens of text, with a significant portion sourced from datasets containing over 170,000 books.
- Claude (Anthropic): Court documents revealed this company purchased and physically scanned millions of books to create its training data, manually categorizing each one.
- Other Models: Companies like Meta and Bloomberg have also utilized large book datasets to train their own AI models.
This has led to legal challenges from authors in the US and Europe who argue that their copyrighted works were used without permission to train these commercial systems. However, some courts have ruled in favor of AI companies, characterizing AI training as a “highly transformative” use of information.
A Critical Crossroads for Culture and Technology
The central paradox is now clear: the tool that was built by consuming books is making the act of reading them seem less necessary. Professor Baron’s research indicates that students can now use AI-generated summaries to pass literature classes, a convenience that may undermine the motivation to read the original works.
This trend raises a profound question for the future: If the next generation reads fewer books, where will future AI models find high-quality data for training? Researchers warn of “Model Collapse,” a scenario where AI trained on its own generated outputs leads to a progressive degradation in quality.
The path forward requires finding a balance that encourages the production of high-quality human content while fostering technological innovation. The ongoing global dialogue and legal proceedings will shape how intellectual property is respected in the AI age. The core challenge is to ensure that the digital advancement of AI does not come at the cost of the rich, human-centric culture of reading that has been essential to societal progress. The future of both may depend on achieving this equilibrium.