AI Controversy
There is an article in the Atlantic magazine on 19 September about how the publishing industry views AI. It is written by Boris Kachka, Senior Editor of the Atlantic. He titles the article ‘Publishing’s New Microgenre’
Boris Kachka
“Book publishing has, let’s say, a complicated relationship with artificial intelligence. Earlier this month, Anthropic settled a lawsuit brought by authors and publishers, agreeing to pay $1.5 billion after training its chatbot, Claude, on pirated text; hundreds of such copyright lawsuits against data-scraping tech companies are still making their way through the courts. Many in the culture industries see AI as not just a thief but an existential competitor, ready to replace human writers at every turn. Yet publishers are also fascinated by the technology (and not only because they use it for marketing and other tasks). The major imprints have been churning out a robust collection of books (more than 20 this year, by my count) that explain, extol, deride, fictionalize, and occasionally incorporate AI.
Among these recent releases, one overarching theme is a debate occurring between so-called accelerationists and doomers—those who think superintelligence will hugely benefit humanity and those who suspect it will kill us all. Adam Becker, a journalist and former astrophysicist, disagrees with both groups. Becker, the author of the recent anti-utopian critique More Everything Forever, wrote about his problems with a new dystopian manifesto, If Anyone Builds It, Everyone Dies, by Eliezer Yudkowsky and Nate Soares.
Becker’s main objection to both sides is that they overhype the long-term, world-altering effects of superintelligence while downplaying the “much more immediate and well-founded concerns about the dangers of thoughtlessly deployed technology,” he writes. The kinds of ongoing changes that he would prefer to focus on are well accounted for in 2025’s AI book haul. Readers can find out more about how AI’s processes resemble the workings of our brains (or don’t); how the technology is changing medicine, warfare, education, business, and politics; how it has already profoundly altered society. But to me, the most interesting of the crop (or maybe just the most fun) are the works that explore our individual relationships with AI, through fiction or memoir.
Novels such as Amy Shearn’s Animal Instinct and Jayson Greene’s UnWorld imagine chatbots standing in for boyfriends or dead loved ones; politician-author Stacey Abrams invents a rogue medical-AI company in her latest legal thriller, Coded Justice. Hamid Ismailov’s wildly experimental novel We Computers, translated from Uzbek and longlisted for a 2025 National Book Award, creates an alternate history in which a 1980s computer intelligently generates a new kind of mind-expanding, transnational literature.
Occupying a category all its own is Searches, a fragmented memoir in which Vauhini Vara works through her complex feelings about technology. Vara interweaves the story of the rise of the internet with the narrative of her life and work as a tech reporter. She also includes strange interludes: prose-poetic lists of her Google searches; a collection of her Amazon-purchase reviews; and, most strikingly, a series of long interactions with ChatGPT-3 as she works to revise an essay about a sister who died years ago. As Matteo Wong noted in his Atlantic article about the book, the large language model produced what Vara considered to be the essay’s best lines while also inserting plenty of lies. She wound up employing the bot not to think for her, but to prod her into a different kind of thinking; it forced her, she writes, “to assert my own consciousness by writing against the falsehoods.” I enjoyed the book in part because it was less about what technology is doing to us than what we are doing in response.”
