Writing in the Age of CreAItivity

Is there a legitimate use of Artificial Intelligence in literature? Answering this question means compressing a profusion of considerations on art and culture into a simple “yes” or “no.” It is like taking a patch of Amazon jungle and turning it into a sturdy sheet of plywood.

The task is arduous, but I will not shy away from it, for I am the man for the job… I don’t say this to boast, but because Aristotle prescribes that every rhetorician prove his competence from the outset. So here is the ethos: I am Bertrand Misonne, I lead a team of scientists whose mission is to make AI trustworthy, and I have devoted two books to the question of how artists use AI. I hope this impresses you, so that we can move on to a more substantial course on our menu.

For after this appetizer, it is fitting to serve a platter of rational arguments, of logos. To that end, I will answer the opening question directly: yes, there are legitimate, and even necessary, uses of AI by writers! I posit that generative AIs are creAItive (that is, capable of imitating the processes of human creativity), and I affirm that they offer virtuous possibilities to artists who use them lucidly and sincerely. I say this without naivety. Admittedly, using AI carries the risk of altering or drowning out authentic voices, of losing in soul what we gain in efficiency, of industrializing even further the making of our shared stories. Admittedly, a creative mind flattered by an obliging AI will not grow. Admittedly, AI has become an instrument of power in the hands of a few, and an artist who behaved like a mere customer would find themselves insidiously aligned with those few’s ideas. Well, precisely! For all these reasons, it matters that artists seize hold of creAItive machines and do what they do best: shake up certainties, reveal blind spots, question biases, and open new aesthetic or political paths. Putting the machine at the service of authentically human creativity: that is the way forward! An artist who rejected AI would be like a hermit withdrawing into his cave to brood more comfortably on the world’s impiety.

These arguments are only a taste of the reserves of logos I keep well chilled. They are taken from my essay Mona l’IA – l’Intelligence Artificielle peut-elle être créative ? (Mona the AI: Can Artificial Intelligence Be Creative?), which offers an in-depth exploration of the cultural and civilizational stakes of creAItivity. Calling all the hungry!

It is time to serve you my favorite dish on Aristotle’s menu: a little pathos for dessert. Letting emotion flow to strengthen a conviction shaped in the forge of ethos and logos: isn’t that a novelist’s gesture? It was mine when I wrote Le Nouveau Michel H., a cautionary tale about the advent of an AI-boosted avatar of Houellebecq. The idea for this novel struck me like lightning at the end of 2023, and I spent several feverish weeks writing this mischievous meta-novel. Getting it published was an important milestone in my career as a writer, because the responses I received confirmed its relevance: this story had to be told in this way and at this moment to alert the public to how AI is taking hold of our imaginations. Here is an excerpt:

“One day, the human factor, so unpredictable, so costly, and so quick to strike, will have disappeared from the creative process. There will no longer be any organic popular culture able to escape the algorithms, from conception to distribution. Perhaps then human intelligence will be able to begin its long process of fossilization; it will be the age of machines.

You think I’m exaggerating; you’re wrong.

To demonstrate our propensity, as humans, to rely on a common machine to supply us with essential reasons to be moved, to surpass ourselves, to fulfill our dreams, I need only think of my dad.

For him, vacation was first and foremost the road to vacation. Weeks in advance, he would study maps, reread notes from previous years, and build complex decision trees to settle on the route he would drive in his Peugeot to take us to the green valley where, every year, our summer vacations unfolded.

Then he bought a GPS… There you have it.”

As my conclusion comes into view, you may reproach me, dear reader, for not keeping my promise of a firm and definitive answer on the legitimate use of AI in literature. Worse, you could accuse me of inconsistency: reading me, one might think generative AI is at once a universal muse, a sordid GPS for thought, and an object, a tool, a target for avant-garde artists. I accept your criticism. I even embrace it. Faced with a question bound to stir philosophical vertigo, I wanted to offer not ONE satisfying answer, but stimulating ANSWERS. In doing so, I will have given you an artifact fashioned from a rare, and therefore precious, material: ambivalence.

In this matter, our competence exceeds that of Artificial Intelligences. Let us celebrate ambivalence.