Chapter 13

The Whirlwind of Intelligence

6 min read

13. THE WHIRLWIND OF INTELLIGENCE

We probably have very little time left before the singularity — or, to put it more prosaically, before AI surpasses humanity in every domain. And even if the singularity never arrives, the movement to externalize writing has been set in motion and is not likely to reverse.

If you have a novel in progress in this first half of 2026, and you plan to spend a few years writing it, publishing it somewhere in 2028, 2029, or even 2030 or 2031… What world will we be living in by then? Already in the nineteenth century, Balzac did not wait two years to send his texts out into the world: he jumped on the train of the press, publishing in serial installments, synchronizing the speed of his writing and publication with that of his era. How can we hope to match the temporality of the present using tools made for another speed, for another time?

This is why I am publishing this essay on a website programmed in less than a day using Kimi Code (a tool similar to Claude Code). Had I turned it into a series of articles for a journal or a book for a publishing house, it would not appear for ten months or a year, save rare exceptions.1

Will traditional publishers and literary journals, poorly attuned to the present and to the future taking shape, carry on as if nothing were happening? Perhaps, for a time.

At the turn of the 2010s, when we were exploring writing and publishing on the web — blogs, digital books published rapidly through publie.net in particular — I believed that literature was about to break away from the old continent of the print book. The book was not going to disappear, of course, but it would in all likelihood take on the same status as the vinyl record: a vintage object one loves to own out of nostalgia and love of the thing itself. The most vital part of literature was going to migrate into the agility of the web, where the present was being invented.

I had overlooked the enormous force of inertia that drives the literary world. There was a backlash, a triumph of the print-book industry — and of the e-book as mirror of the print book. To the detriment of authors, sure, but a victory nonetheless, symbolic capital remaining attached to the book in its traditional form, the literary press falling in line, and so on. Most blogs disappeared; publie.net closed up shop. In the English-speaking world, the e-book captured an enormous market share while remaining an ersatz of the print book, largely seized by mega-corporations like Amazon.

Why, then, should the traditional book, which has weathered so many eras, not maintain its hegemony through the intelligence revolution? In the age of the web-as-library, I had underestimated the importance of economic incentives. To put it bluntly: the rupture of the web-as-library was not devastating enough to force the change.

By automating white-collar jobs, AI is about to create an economic shock for which we are unprepared. An “Artificial Revolution” that will spare no sector, and that includes the book industry. This is a far less speculative possibility than one might think — a scenario that does not require the advent of artificial general intelligence, superintelligence, or the singularity; in fact, the process is already underway (the numbers show that entry-level jobs are being eaten away in the United Kingdom and the United States) and it should be running at full throttle by 2027. The result: the artificial economy will grow while the human economy declines. It is possible that, owing to the greed of AI companies, we will be largely shut out of the artificial economy; this is why it is urgent to think about the ownership and sovereignty of our data. Those who turn their backs on intelligence will be relegated to a world that is past and impoverished.

It is shocks that induce the most brutal changes in history. It took the Black Death (1347–1352) to accelerate the replacement of parchment with paper2; it took the Great Depression for DuPont to invent nylon and the synthetic materials that would transform the textile industry; it took the COVID-19 pandemic for remote-work tools, though they already existed, to become standard practice — a sectoral shock, admittedly, that did not have the magnitude needed to shake the book industry. The coming earthquake could be of an entirely different order. There is every reason to believe that the “white-collar bloodbath” — to borrow the words of Dario Amodei, CEO of the AI company Anthropic — will unleash a tidal wave far more devastating and transformative.

We are entering a deeply chaotic period in which AI threatens to break the back of the traditional economy — the book industry included — and in which governments, inundated with requests for support, will lack the fiscal capacity to keep entire sectors of the economy afloat. Even if we refuse the change, History may well force our hand.

To anticipate the rupture and reverse the dispossession, I have proposed avenues of exploration: becoming a master builder, reclaiming one’s data, elucidating one’s voice, multiplying actualizations, using translation as a lever, and so on. I sense that the regime of literary ownership is entering a period of transition in which value will shift below the reproducible text, into proprietary data. By deploying, from the dataset and the imprints it harbors, improbable prose and poetry, we have a chance — slim, perhaps, but do we have a choice? — of pushing literature a little further into the twenty-first century.

I know that most authors will not board the supersonic aircraft of artificial intelligence. I know that the majority will curse — rightly — the big AI companies and the leaders who give them permission to unleash upon the world this torrent that redraws everything. I know that many will prepare their own Luddite revolt.

There is a certain beauty in this gesture of refusal in the face of radical, violent change. One has the right to reject the power loom and to weave by hand; and to watch through the window as the world transforms and grows ever more mad and incomprehensible.

Or one can plunge into the whirlwind of intelligence without knowing where it will lead.

  1. The essay is roughly 17,000 words, 25,000 counting the appendices: the length of a short book. ↩

  2. By decimating the monks who copied manuscripts, the Plague caused the cost of book copying to explode. At the same time, the dead left behind mountains of used clothing, making rag paper abundant and cheap. The conjunction of prohibitively expensive copying and abundant paper paved the way for Gutenberg’s printing press the following century. ↩