A column about curious stories from the worlds of science, technology, and film. Today — a literary mystery more than one hundred and sixty years old. In mid-October, an article about a Jules Verne novel written in 1863 and left in a family safe for one hundred and thirty-one years unexpectedly took off on Hacker News. The novel is called Paris in the Twentieth Century (Paris au XXe siècle), and it is one of the first warning novels in world literature about what society becomes when it places technology and money above living people and human dignity. The book’s predictions are so precise that its first wave of readers in 1994 felt an almost mystical shiver. But the real story surrounding this manuscript is not about the accuracy of its forecasts. It is about how one editor altered the trajectory of all world science fiction with a single review, how publishers invented a beautiful legend about a “lost manuscript” — and why the article telling this story today was itself written eighty percent by a language model. The genre has irony to spare: a book about machines pushing literature aside became the occasion for a vast discussion of how AI has pushed writers aside.
Verne wrote Paris in the Twentieth Century in 1863 — he was thirty-five, with one success behind him (Five Weeks in a Balloon) and sixty novels ahead of him. The manuscript landed on the desk of Pierre-Jules Hetzel — France’s most influential editor, the man who turned Verne into a star. The reply became the most famous rejection in the history of science fiction. Hetzel wrote: “I did not expect perfection from you — I knew you were attempting the impossible — but I had hoped for at least something better. There is not a single question about the near future in this work that has been properly resolved, not a single criticism that has not already been made and remade. I am astonished by you... The work is dull and lifeless.” And he added advice that sounded like a sentence: put the manuscript aside for twenty years or so. Verne obeyed — and the manuscript went into a drawer.
The novel shows Paris in 1960 through the eyes of sixteen-year-old Michel Dufrénoy — a literature graduate in a world where a classical education has become cause for ridicule. At the graduation ceremony, the crowd boos him for his Greek and Latin: “Latin and Greek are not merely dead languages, they are buried.” Bookshops are crammed with technical manuals, nobody there has heard of Victor Hugo, and the main subject has become the “science of commercial progress.” Michel gets a job at a bank — copying documents in the Casmodage office, where his mentor is Quinsonnas, an eccentric accountant who writes music at night.
The novel’s list of predictions is embarrassingly long. Internal-combustion vehicles — “omnibuses propelled by explosions of gas” — and “facsimile telegraphy,” transmitting images and signatures over distance, more than a century before the fax machine entered every office. Electronic music and the synthesizer, recorded concerts instead of live performances, elevators, skyscrapers, power grids illuminating cities, the electric chair, high-speed trains. Even the cinematic metropolis of traffic jams and smog — it is all there. And the book ends with Europe gripped by an unprecedented winter and famine, the hero wandering through the freezing capital, pursued by the “Demon of Electricity,” before collapsing unconscious in the snow at Père Lachaise Cemetery.
The most precious thing in this story is not the predictions, but the alternative path that vanished from world culture. After the rejection, Verne returned to the safe track: Five Weeks in a Balloon was followed by Captain Nemo, From the Earth to the Moon, In Search of the Castaways. We got the genre of the voyage-and-discovery novel, where technology is a marvel and a gift to humanity. But we did NOT get the genre of the “warning novel” about the cost of that gift a century and a half before it emerged in Zamyatin, Huxley, and Orwell. Science fiction might have grown from an entirely different point — not from the inventor’s exhilaration, but from the moralist’s doubt. And no one will ever know what world science fiction might have looked like if doubt, rather than exhilaration, had lain at its foundation.
The second thread of the mystery is the legend of the “lost manuscript.” The beautiful version goes like this: the manuscript was considered lost until, in 1989, the writer’s great-grandson opened the family safe and found it there. A classic “treasure in the attic.” Except Wikipedia dryly notes that the manuscript’s location had been known ever since Verne’s death in 1905, and that the story of its “loss and rediscovery” was a publishing marketing stunt for the 1994 release (the “citation needed” footnote speaks more eloquently than any investigation). The French press in 1994 — L’Humanité says so outright — knew about the family safe in advance. The great-grandson was named Jean Verne. The publishers — Hachette and Le Cherche-Midi — built up the kind of information buzz around the discovery that today would be called a viral launch.
And here begins the modern layer of the story — the one that made the Hacker News thread spread. The article about the novel was published by Blue Continuum — a beautiful, well-structured text, with a powerful final thought. One commenter ran it through the Pangram AI detector — verdict: eighty percent of the text was machine-generated. The human contribution: only quotations from the novel itself. The comment became the top one in the thread: a text recounting how technology kills literature had itself turned out to be technology pretending to be literature. The article’s author, John Steele, publisher of Blue Continuum, has made no comment.
There is a temptation to end with the story of a visionary genius. I don’t like it. First, Verne’s forecasts were not all hits: the “scientific” foundations of the plot about the “Demon of Electricity” sound like steampunk today. Second, the “safe as surprise” is largely a marketing legend, and a good mystery story has to say so. But most importantly — the Blue Continuum story came full circle too perfectly: a book about a world where machines write books returned to us through an article in which a machine was caught writing a book. This is neither a reason for hype nor a reason for a moral. It is simply a reminder: the predicted thing came true so quietly that nobody was even surprised.