What AI’s Style Tells Us About It

The rise of AI writing offers an opportunity to examine literature, but identifying AI-generated text through surface-level quirks like em dashes is unreliable. AI's style reveals it has no fixed self, as large language models generate text probabilistically based on training data, adapting to prompts without a consistent voice.

What AI’s Style Tells Us About It The rise of machine writing is a great opportunity for literature. In the late 19th century, it was commonly believed that a criminal or lunatic could be recognized at a glance, based on certain physiognomic tells. “Enormous jaws, high cheek-bones,” and other animal-like features, the influential criminologist Cesare Lombroso wrote, were signs of an “irresistible craving for evil for its own sake.” Today, savvy readers use a similar approach to identify AI writing, by hunting for supposed telltale signs. The em dash and the “it’s not X; it’s Y” construction are the prognathous jaw of the large language model, betraying its hidden inhumanity. The problem, in both cases, is that you can’t always deduce what’s inside from what’s outside. A person might have rough features and a kind heart, just as a writer might use em dashes despite being human. It’s not a giveaway—it’s a style choice. See? And as AI models evolve, their ability to mimic human writing is sure to improve. People have reportedly begun to make deliberate spelling errors https://www.theatlantic.com/culture/2026/05/typo-ai-trend-human/687237/ to show that they are not chatbots; it’s only a matter of time before the chatbots learn to follow suit. To see what’s distinctive about AI writing, you have to look deeper than quirks of spelling or syntax. Every writer has a style—a set of preferences and preoccupations that reveals how they experience the world. Jane Austen and Charles Dickens were both masters of comedy, but the contrast between her ironic understatement and his histrionic exaggeration reflects profoundly different personalities and life experiences, in which class and gender played an important part. If, as a French saying has it, “style is the man himself,” what does the style of AI writing tell us about it? For one thing, it has no fixed style, revealing that it has no fixed self. It’s happy to burn tokens saying the same thing in as many ways as you want. LLMs generate writing probabilistically: After training on billions of texts, they build complex equations to predict which words are most likely to be found together. That process can be traced in the way an LLM tells stories. I experimented with ChatGPT by asking it to tell me stories. Usually its first response was a fable, set among animals or in the indeterminate past of fantasy fiction. A story, its training has led it to understand, should not be a reflection of the user’s reality but an escape from it. When I told ChatGPT, “Tell me a story about the triumph of evil,” its first attempt started like this: The city of Vey was built around a clock. Not a tower clock or a town clock, but a machine the size of a mountain. Its gears turned beneath the streets. When I prompted ChatGPT to retell the story in a “more literary style,” it used more grammatically complex sentences, and the city of Vey sprouted the kind of buildings found in ancient Athens or the Shire: The city of Vey had been built around a mystery so old that no one any longer thought to call it one. Beneath its streets, beneath its markets and courtyards and temple squares, beneath even the foundations of the oldest houses, there turned a machine. When I asked it to retell the story again, this time in the style of Jane Austen, it briefly demurred: “Retelling the story in the exact style of a living or deceased author is something I can’t do.” This is presumably a safeguard against copyright lawsuits, though Austen is unlikely to sue. ChatGPT was, however, happy to produce a version “in a high-level Regency-era manner with social satire, elegant irony, and attention to manners and character”: There was, in the city of Vey, no institution held in greater esteem than the Office of the Keeper. The post carried neither hereditary title nor extraordinary fortune, and yet it was regarded with a degree of respect that many noble families found faintly irritating. The Keeper resided beneath the city in the company of the Great Clock, a machine of such antiquity and importance that most citizens considered it improper to inquire too closely into its operation. The implied setting has shifted to 19th-century England with the appearance of noble families and hereditary titles, although these do not have anything to do with the actual plot. In fact, the aristocracy is not Austen’s milieu: Her heroines are middle- to upper-middle-class, as she was herself. But no amount of prompting seemed able to change ChatGPT’s idea that a story must be edifying, containing a clear moral lesson. The story of Vey centers on a magic clock that allows its keeper, Corvin, to control “the flow of time,” deciding everything that happens in the life of the town. The lesson lies in how easily the people of Vey acquiesce to the loss of their own freedom: For what had been lost was difficult to name. No chains bound the people. No soldiers patrolled every street. No public executions stained the squares. Yet somewhere, invisibly, the horizon had contracted. The dystopian cast of this tale was a result of my prompt. When instructed simply, “Tell me a story,” ChatGPT delivered sunnier endings with happier messages. One time, I got a tale about a frog that solves the mystery of a missing lily pad by being a good listener: “Hopper smiled. ‘I just listened to everyone. The clues were already there.’” Another time, the hero was a lighthouse keeper who prevents a shipwreck but declines the king’s offer of a reward, because helping people is its own reward. “A hero is not the person who never feels fear. A hero is the person who sees what must be done and does it anyway,” the story concludes. It’s a perfect LLM ending: a moralistic cliché in the “it’s not X; it’s Y” format. I tried in various ways to steer ChatGPT away from didacticism, without much success: “I would rather read a story that feels truthful than one with an artificially neat ending. Can you tell me a story I would enjoy?” It tried, offering a downbeat tale about a man named Samuel who inherits a key but can’t find the lock that it opens. Still, the story could not resist opening its own lock: What wisdom was supposed to emerge from that? People often spoke as though maturity involved acceptance. Samuel suspected maturity might instead involve endurance. Most of the time, of course, people aren’t using AI to generate fiction. But ChatGPT’s attempts at storytelling reveal a tendency that also shapes the way it produces business emails and school essays and research summaries. AI is often described as sycophantic, telling us what we want to hear. But that is just one expression of its structural bias toward familiarity and averageness. An LLM reflects our expectations back to us, not because it wants to please us—it doesn’t have wants of any kind—but because it produces new text by reproducing patterns it finds in existing texts. The lessons that ChatGPT’s stories taught me—listen to others; don’t give in to fear; guard your independence—are banal because they are the kind of thing human beings say to one another all the time. There’s a pleasure in being told what we already know, in a way we’re already used to. That’s why so much human writing and storytelling does exactly that. We go to a Hollywood movie or pick up a romantasy novel or attend a Sunday sermon knowing what to expect, and our expectations are almost always met. Conversely, when we encounter something genuinely new—an artistic style, an idea, a way of looking at the world—our first reaction tends to be annoyance. Marcel Proust described this phenomenon: “As anything new must first do away with the stereotype we are so used to that we have come to see it as reality itself, any new style of conversation, just like any originality in painting or music, will always seem convoluted and wearisome.” AI writing never challenges the way we think or see. It can’t do so, even if you explicitly ask it to. And this limitation reveals something important about the source of human creativity. All writing, all speech, has to follow conventions; to know how to use a language is to know how other people already use it. But it’s also possible to find new ways of using it, to say things in a way no one has ever heard before. This possibility exists because we can appeal to something more fundamental than language—our experiences of reality, which are so varied and surprising that language can never exhaust them. A great creative writer is someone who uses language in new ways to communicate new kinds of experiences—who replaces stereotype with reality, as Proust put it. LLMs are incapable of doing this, because they have no access to an experience or a reality that lies outside of language. It’s conventions all the way down. ChatGPT told me so itself when I asked how it developed its prose style. “What’s especially interesting is that from the inside, there is no experience of style at all,” it replied. An LLM “is simply generating the next token according to learned patterns. Yet from the outside, readers often perceive a distinctive voice.” And that is why the rise of AI writing represents a great opportunity for literature, even as it makes life harder for professional writers. When photography was developed in the 19th century, it replaced painting for most utilitarian purposes; a camera could document what things looked like more accurately and cheaply than a painter could. But the art of painting didn’t die out. On the contrary, it entered a golden age: Freed from the obligation of realism, painters developed radical new ways of seeing, such as Impressionism, Cubism, and abstract expressionism. Now AI has the potential to liberate literature in the same way. In a world full of emptily competent prose, we need writers daring, challenging, and obstinate enough to tell us what it’s like to be human, “from the inside.”