The French writer Philippe Janoda had just won the Goncourt Prize when he admitted the next day that he had used artificial intelligence several times while writing this novel.
The award-winning work "The Nameless Woman on the Banks of the Yvette River" explores an unsolved murder case of a young woman that occurred in Paris at the end of the 1940s.
He said that he included a sentence generated by ChatGPT in his story, with the intention of “playing a little trick on the readers.” He only told the judging panel which specific sentence it was; “If it gets out, I know who said it,” he joked.
In his words, using AI in some cases is actually a way to mock AI itself.
The attitude of writers towards AI has always been divided. Some have openly resisted it, while others signed petitions against using their works to train models. There are also those, like Janada, who treat it merely as a small tool in the writing process – something akin to hiding Easter eggs within texts, using machine-generated sentences to test whether readers and judges can distinguish them from human-written ones.
The Goncourt Prize is the most prestigious literary award in France. If a winner had voluntarily admitted to using AI, it would have caused a huge uproar two years ago. Nowadays, however, people are discussing whether a single sentence even constitutes a problem at all.
The subject matter of this book is itself a mystery, and the author has added another riddle within it. Readers will likely begin to try and figure out which sentences seem least likely to have been written by a human being.
What’s really interesting is that that sentence hasn’t been recognized yet.