Reading is facing a looming crisis. AI emerged at a time when both children and adults were already reading fewer books than in the recent past. As a linguist, I examine how technology shapes the way people read, write, and think.
That includes studying AI’s influence, which is rapidly transforming how people interact with books and other forms of writing—whether for assignments, research, or leisure. I’m concerned that AI is speeding up a broader decline in the value placed on reading as a uniquely human activity.
Anything Except The Book
AI’s writing abilities have drawn much attention, but only recently have researchers and educators begun discussing its capacity to “read” vast datasets and then produce summaries, analyses, or comparisons of books, essays, and articles.
Need to read a novel for class? Today, you could simply skim an AI-generated summary covering the plot and main themes. This shortcut, which can sap people’s motivation to read for themselves, inspired me to write a book on the pros and cons of letting AI do the reading.
Relying on summaries or analyses isn’t new. CliffsNotes has been around since the late 1950s. Centuries before that, the Royal Society of London produced digests of scientific papers in its extensive Philosophical Transactions. By the mid-20th century, abstracts had become standard in scholarly articles, allowing potential readers to review a brief overview before deciding whether to read the full work.
The internet introduced countless new shortcuts for reading. Take Blinkist, for example—an app-based subscription service that distills mostly nonfiction books into 15-minute text or audio summaries known as “Blinks.”
AI That Reads and Thinks for You
Generative AI takes these shortcuts even further. AI-powered tools like BooksAI now produce summaries and analyses once created by humans, while BookAI.chat lets users “chat” with books. In both cases, there’s no need to read the books firsthand.
If you’re a student tasked with comparing Mark Twain’s The Adventures of Huckleberry Finn and J.D. Salinger’s The Catcher in the Rye as coming-of-age stories, CliffsNotes can only take you so far. You might find summaries of each, but the actual comparison is up to you. With general large language models or specialized tools like Google NotebookLM, however, AI can handle both the “reading” and the comparison—and even generate insightful questions for class discussion.
The trade-off is missing a key benefit of reading a coming-of-age novel: the personal growth that comes from living through the protagonist’s struggles in your imagination.
In academic research, tools like SciSpace, Elicit, and Consensus merge the capabilities of search engines with large language models, finding relevant papers and summarizing or synthesizing them—cutting literature review time dramatically. Elsevier’s ScienceDirect AI even boasts on its site: “Goodbye wasted reading time. Hello relevance.”
Perhaps—but in the process, you lose the chance to decide for yourself what’s relevant and to make your own connections between ideas.
Unwelcoming To Readers?
Even before generative AI became widespread, book reading—both for leisure and for school—was already in decline.
In the U.S., the National Assessment of Educational Progress found that the share of fourth graders who read for fun almost daily fell from 53% in 1984 to 39% in 2022. For eighth graders, it dropped from 35% in 1984 to just 14% in 2023. In the U.K., a 2024 National Literacy Trust survey showed only one in three 8- to 18-year-olds enjoyed reading in their free time—nearly nine points lower than the year before.
The pattern is similar among older students. In a 2018 survey of 600,000 15-year-olds across 79 countries, 49% said they read only when necessary, up from 36% a decade earlier.
Declining Reading in Higher Education
College students fare no better. Recent reports highlight declining reading in U.S. higher education. My research with literacy scholar Anne Mangen found that faculty are assigning less reading, often because students refuse to do it.
Cultural commentator David Brooks captured the issue with a telling anecdote: “I once asked a group of students on their final day at their prestigious university what book had changed their life over the past four years. After a long, awkward pause, one student finally replied, ‘You have to understand, we don’t read like that. We only sample enough of each book to get through the class.’”
The trend extends well beyond students. A YouGov survey found that only 54% of Americans read at least one book in 2023. In South Korea, the figure was just 43%—a steep drop from nearly 87% in 1994. In the U.K., The Reading Agency also reported declines in adult reading, noting one possible cause: in 2024, 35% of adults identified as “lapsed readers,” meaning they once read regularly but no longer do. Of these, 26% said they had stopped because they spent more time on social media.
Today, the term “lapsed reader” could apply to anyone who sidelines reading—whether due to waning interest, the pull of social media, or the habit of letting AI do the reading for them.
Everything Lost, Overlooked, and Forgotten
Why read at all?
The reasons are countless—and so are the books and websites advocating for it. People read for pleasure, stress relief, learning, and personal growth.
Research links reading to childhood brain development, greater happiness, longer life spans, and slower cognitive decline.
That last point matters even more as more people let AI handle mental tasks for them—a phenomenon known as cognitive offloading. Studies show that when people rely on AI to do the work, they view themselves as using less of their own thinking ability. EEG research even found distinct brain connectivity patterns when participants used AI to help write an essay compared to when they wrote it entirely on their own.
It’s still too early to know how AI might affect our long-term ability to think independently. Current research has mostly examined writing tasks or general AI use, not reading. Yet if we stop practicing how to read, analyze, and form our own interpretations, those skills will inevitably weaken.
And it’s not just cognitive abilities at stake. Letting AI do our reading also means missing the joys that make reading worthwhile—being moved by a line of dialogue, savoring a clever turn of phrase, or feeling a bond with a character.
AI’s promise of efficiency is tempting, but it comes with the risk of eroding the rewards of literacy.
Read the original article on: Phys.Org
Read more: A Study Finds AI Chatbots Can Be Tricked Into Revealing More Personal Information