Stephen Marche: Writers need to accept AI, but human writing isn't losing its value

Language that sounds convincing but has no connection to reality is becoming the main byproduct of the generative AI era — and that is exactly why the role of the human author isn't disappearing, but changing. Canadian writer Stephen Marche, who has himself experimented with writing a novel with AI, argues that artificial intelligence has entered literature and media for good. But instead of ending the profession, it brings a more uncomfortable shift: it makes formulaic, averaged-out style dramatically cheaper and forces writers to prove their worth not through polished sentences, but through substance, precision, and experience.

Marche opens with an everyday scene on a playground. During an argument, a girl yells at a boy: "That's AI!" She means a new kind of nonsense — speech that sounds meaningful but actually has no connection to reality.

For Marche, this is an important signal: kids have quickly learned to recognize something adults are still trying to describe with terms like "hallucinations" or "synthetic content." AI has already become a cultural category, not just a technological tool. Bad writing used to be able to pass as normal; now readers are developing a new instinct — distrust of prose that is too even, too impersonal, suspiciously convenient.

According to Marche, treating AI as either the apocalypse or a universal fix for every problem is pointless. It isn't the end of language, and it isn't a magical substitute for creativity. It is a powerful disruptive tool that is changing the economics of writing.

The scandal around Mia Ballard's novel Shy Girl is telling in this context. Hachette canceled the book's publication after accusations that the text relied on AI generation. Yet the novel had already come out as a self-published edition, and judging by the reaction, neither readers nor editors saw a problem until the use of AI was named out loud.

Ballard herself said that it was an acquaintance who had edited an early draft, not she, who used AI. The episode essentially exposes the core nerve of the debate: society has not yet settled on stable rules for what counts as unacceptable use of AI in literature, or where the line runs between editing, co-authorship, and full generation. But something else is already clear: command of a banal style is no longer a scarce skill.

Machines can quickly churn out smooth text, standard descriptions, safe opening paragraphs, and endless variations on familiar themes. That raises the value of what is harder to automate: an author's perspective, personal responsibility for meaning, and the ability to connect words to reality and form to observation, risk, and concrete human experience. What sets a good writer apart now is not the ability to write without errors and clichés, but the ability to say something real.

For the industry, this means rebuilding familiar quality criteria. Editors, publishers, and readers will have to rely less on surface literary polish and look more closely at a text's origins, its internal logic, and how much the author was actually involved. The job is changing for writers, too: AI can speed up drafting, finding the right phrasing, or structuring material, but it doesn't relieve them of responsibility for truth, nuance, and tone.

The easier it becomes to produce plausible text, the more valuable text becomes when a real person and a recognizable mind stand behind it. Marche's conclusion is simple: we will have to accept AI, because it is already built into the culture of writing. But there's no need to surrender to it.

In a world where machine speech increasingly imitates meaning, an author's value is defined not by the ability to generate volume, but by the ability to tell lived experience apart from verbal noise. And that is what makes writers not less important, but more.