It would have gone unnoticed if it had been said by a second-rate writer seeking attention. But it was said by Olga Tokarczuk, Nobel Prize in Literature in 2018, author of "The Wanderers" and "The Books of Jacob", one of the most important and respected voices in contemporary European literature. And he said it not as a provocation or as a headline, but almost in passing, in a conversation lasting more than half an hour on the Impact forum in Poznań: he is using artificial intelligence to write his next novel.
The statement was brief and went almost unnoticed at the time. Controversy grew later, when that admission became associated with another announcement Tokarczuk made in the same conversation: that this new novel, scheduled for fall, will likely be his last. The combination of the two elements—AI and farewell to the novelistic form—turned what could have been a footnote into one of the most interesting literary and technological debates in recent weeks.
What exactly does "writing with AI" mean?
The first problem with the controversy is that "writing with AI" can mean radically different things. It may mean using an AI tool to generate complete drafts that the author then minimally edits, raising legitimate questions about authorship. It can mean using AI as a research and organizational tool, to process large volumes of information that the author then transforms into his or her own prose. It can mean something in the middle: using AI to explore variants of a difficult paragraph, to overcome creative blocks, to gain an outside perspective on the structure of a scene.
Tokarczuk did not specify which of these forms of work he is using, and that ambiguity is part of what fuels the debate. The harshest critics assume that this is something close to the first scenario. The most enthusiastic defenders read it as the second or the third. Reality is probably more complex and more interesting than any of those caricatures.
The controversy is not really about whether AI can write good literature. It's about whether it matters who—or what—generates the words, whether the reader's experience is equally valuable.
The nature of literary authorship
The debate that Tokarczuk has reopened has deep philosophical roots. The notion of authorship that Western culture takes for granted—a unique individual who, based on his or her experience and vision of the world, creates a work that bears his or her signature and is an expression of his or her subjectivity—is historically recent and culturally specific. For centuries, writing was a much more collaborative process: medieval copyists added to and modified the texts they copied, ancient scribes worked in traditions that valued continuity more than individual originality, 17th-century writers considered it perfectly legitimate to rewrite and improve the texts of their predecessors.
The Romantic notion of the individual creative genius—the writer as the unique and irreplaceable source of a worldview—is barely two centuries old, and it is not clear that it is the only possible way of understanding what is valuable in literature. The relevant question may not be "did Tokarczuk write these words?" but "is there in this text a vision of the world, a way of looking at reality, that could only have emerged from Tokarczuk's experience and intelligence?"
If the answer to that second question is yes, then the participation of an AI tool in the writing process is not qualitatively different from the participation of a good editor, a thesaurus, or rereading the underlinings in the books one has read. If the answer is no—if what comes out doesn't have Tokarczuk's distinctive voice and perspective—then the problem isn't the AI: it's that the result isn't genuinely yours, no matter what tools you used.
The literary market and its anxieties
Beyond the philosophical issues, the controversy has an economic and professional dimension that should not be ignored. Writers, as a collective, are in a particularly vulnerable position in the face of the development of generative AI. Unlike many other professions where AI automates repetitive tasks while professionals focus on higher value-added work, in literary writing the distinction between what an AI can do and what a human writer does is much less clear.
Current language models can generate perfectly technically competent prose. What they can't do—or can't do consistently and reliably—is inhabit a specific human experience, process the pain or joy or confusion of a specific life and turn it into something that another person can recognize as true. That is precisely what the best writers do that no tool can replicate.
The debate in figures: writers and tools throughout history
- 15th century: The printing press causes panic due to the "desacralization" of the book
- 20th century: The typewriter, dictated to secretaries, recorders: all questioned
- 80s-90s: Word processors "destroy literary style" (they did not)
- 2023-2026: Writers divided between rejecting and adopting generative AI tools
- 2026: A Nobel Prize winner normalizes the use of AI in the writing process
Why is it Tokarczuk who says it and not someone else?
There is something significant in the fact that it is precisely Tokarczuk who has taken this step. She is not a popular genre writer looking for efficiency in production. She is one of the most awarded authors and most committed to the literary and philosophical complexity of the last thirty years. His novels are dense, ambitious, formally risky. If anyone could use AI as a tool for exploration rather than as a substitute for one's own thinking, it is her.
The context in which he said it is also significant: not as a technological manifesto, but as an element within a broader reflection on his own creative process and its relationship with the form of the novel. The association with the announcement that this would likely be his last novel suggests that he is exploring the limits of what the novel form can do, pushing its tools as far as they will go, and AI would be part of that process of radical exploration.
To continue thinking: The Swedish Academy, which awards the Nobel Prize in Literature, has not made any comment on Tokarczuk's statements. Nor is it expected to do so: the Nobel awards already published works, not ongoing writing processes. But the question of whether a work written in part with AI could be a Nobel candidate in the future is one that the literary world will have to answer sooner than it thinks.
Ultimately, the scandal surrounding Tokarczuk's statements says more about our collective anxieties about AI than about the value of his work. If the book you publish in the fall turns out to be extraordinary—if it has the density and humanity that characterize your previous writing—the discussion about how you wrote it will quickly evaporate. If it is mediocre, the process will become the convenient explanation. Literature, deep down, has always judged works, not methods.
Share this article
