In a recent WIRED magazine article, reporter Emma Madden notes that the monolithic tone of AI chatbots is prompting many writers to adopt first-person narratives and distinctive stylistic quirks so they won’t be mistaken for chatbots. As chatbots turn out prose that is polished and predictable, a growing number of authors are leaning on personal storytelling, unusual phrasing, and small deliberate quirks to make their writing unmistakably human.
Here are the main Takeaways:
- Writing against AI. Madden opens her piece with novelist Laura Brooke Robson, who last year began a novel about the founder of an AI dating app. While working on it, she made a conscious effort to avoid the patterns people associate with AI. She dropped what she calls wishy-washy metaphors. She took a winding path through her sentences to avoid an em dash or a list of three. Robson told Madden the process strengthened her writing. She now leaves small, intentional imperfections in her work as a quiet signal that a human wrote it.
- A new writing style is taking shape. Much of this emerging style is defined by what writers choose to avoid. They are moving away from the familiar tells: lists of three, passive voice, the “not X but Y” construction, and em dashes. In their place come personal stories, unexpected observations, and highly specific detail. Novelist Ellena Savage told Madden she misses the conventions she has given up. She misses over-using the em dash, and she misses the rule of three, which she found deeply satisfying.
- Authenticity has become a professional concern. Madden points to several cases where sounding like AI carried real consequences. In March, Hachette withdrew Mia Ballard’s horror novel after online accusations that it relied on AI, which Ballard firmly denies. She says authenticity is now her first priority, and that she edits her own work after an editor she hired ran the book through AI software. Madden notes Hachette did not immediately respond to a request for comment. She also reports that Steven Rosenbaum later admitted his book about AI and truth contained fabricated quotations.
- Editors want writers to sound more human. Richard Fisher of the culture magazine Aeon told Madden he now encourages writers to be more human and more conversational. He wants the rough edges left in, because AI cannot write prose as spiky or idiosyncratic as a person can. At Pitchfork, Mano Sundaresan has pushed contributors toward their own voice, and the magazine has sent a memo barring AI from review copy and encouraging first-person writing. As Madden notes, Pitchfork is owned by Condé Nast, which also owns WIRED.
- AI was easy to spot in the submissions. After opening Playboy to pitches in May, editor-in-chief Philip Picardi received roughly 1,000 submissions. Having experimented with AI tools himself, he judged that most had been generated. The experience convinced him the magazine needed first-person human experience. He now encourages writers to experiment with form, pointing to bell hooks and her refusal to capitalize her name as the kind of writing AI cannot replicate.
- Even mistakes can feel more human. Clickhole writer Jessye McGarry filled one article with spelling mistakes while repeatedly reminding readers a person wrote it. She told Madden the error was an exaggerated joke, and that keeping it in took work, because she had to undo autocorrect. Madden also features Ben Horwitz, whose anti-Grammarly app deliberately leaves typos in emails and signs them off from an iPhone. When he tested plain emails against messy ones on Fortune 500 executives, the only messages they opened were the misspelled, lowercase ones.
- There may be a silver lining. Writing teacher John Warner argues that AI could make good writing more valuable rather than less. He spent years trying to kill the formulaic student essay, which is close to what chatbots now produce on demand. Madden’s own reading is that the new anxiety about style may push writers in the middle of the pack to improve, if only to avoid being made redundant.
Madden ends with an important observation. Once any anti-AI writing style becomes popular, it also becomes training data. AI can eventually learn to imitate intentional imperfections, personal warmth, and other human signals just as it learned earlier writing patterns. The cycle will continue. In the end, the one thing AI cannot easily imitate is genuinely strong writing. As Jessye McGarry says, “The only true counter style is good writing.”
Acknowledgement: This summary was generated using AI writing tools as part of my experimental learning process. For more information, check out the original article listed below.