Entry Point#2: AI & effort

Posted on Oct 6, 2026

Notes on the Thélyson Orélien affair

For a few days now, a controversy has been occupying social media. On 21 September, Thélyson Orélien received the Fnac novel prize for C’était ça ou mourir. A few hours later, an anonymous X account, “Balance ton Claude”, accused him of having written most of the book with an AI, relying on the Pangram detector. The author denies it. Grasset and Boréal, his French and Quebec publishers, maintain their confidence in him, but Boréal suspends the book’s promotion and the Académie Goncourt removes it from its selection on 25 September (Le Monde, Livres Hebdo).

I resisted the urge to comment in the heat of the moment. I took the time to read and listen to what was being said, in order to hear several points of view. I do not know whether Thélyson Orélien used a generative device, and the purpose of this note is not to settle the question. What interests me are the arguments the controversy brought to the surface.

The positions fall into two camps, which I will call, for lack of a better term, the conservatives and the progressives. The conservatives sacralise writing and the “pact” an author is said to make with his readers. They refuse any use of generative devices, which they consider incompatible with the creative act. The progressives see in it a way of living with one’s time: the generative device would be a new instrument of expression serving the author’s creativity. I have already written on these subjects, first on the question of transparency, then on this new instrument of expression.

Going through the exchanges, I was reminded of a passage by Boris Cyrulnik, in Le laboureur et les mangeurs de vent:

“Designating an aggressor brings a strange well-being, a good opinion of oneself, a clarity that needs no validation. The current that carries these ideas is enough to bring happiness to the wind-eaters who feed on ready-made phrases.”

It is this comfort that kept me from reacting right away. In both camps, many reactions were more a matter of rallying than of analysis. I would rather take up some of the arguments put forward, to see what they are worth.

The breach of the author’s pact

Some speak of the breach of a “pact” between the author and his readers. By signing a work, the author would commit to being its authentic and sincere author. I understand this argument, but it seems to me to rest on a belief and a tradition rather than on a tacit pact. This guarantee went without saying as long as no automated writing tool existed. Ghostwriters, for their part, have existed for a long time, and I do not recall such offended reactions about them.

There is therefore clearly something else, and finding the deep and/or real reasons for this indignation is not an easy exercise. I doubt that the answers one could put forward would settle the question, so much does it rest on each person’s own sensibility, but I find the exercise interesting to a certain extent. I continue.

The myth of the isolated author

Another argument regularly comes up to justify the use of a generative device: the artist has always been influenced or assisted. I share the idea that the reclusive author, sealed off from any hold, is an illusion. The artist who creates has always been influenced, and sometimes even assisted, whatever the work.

By influence, I mean what one receives from a cultural education, from one’s tastes, or from a form of inspiration (without falling into plagiarism). Pictorial movements illustrate this quite well: from one period to another, one sees painting techniques and ways of representing being passed on and transformed, with impressionism or cubism for example.

By assistance, I mean help provided by third parties, whether a ghostwriter or a production workshop. Those who put forward this argument cite the workshops of Rembrandt or Jeff Koons. Others stretch the analogy as far as cinema, where the director relies on a large team to make his films while the director alone takes on the “role” of the creative.

I insist on the “and/or”. Assistance is not always present, and there will always be authors who work alone. For them, only influence comes into play.

A question of effort: laborious knowledge

I nevertheless remain sceptical about the relevance of the influence analogy. It leaves aside a notion that seems to me important to dig into in order to understand these reactions: effort.

Influence is, originally, the fruit of an effort. Whoever reads, watches or listens absorbs what they consume to build their framework, their frame of reference, “their cultural and literary memory” to borrow Alexandre Gefen’s expression. Until now, this influence went through work: sorting one’s sources, reading them, appropriating a material. The assistance of a ghostwriter or a workshop also remains a matter of effort, except that the realisation is done by someone other than the author. In current debates, this dimension of effort and appropriation is (almost) never considered.

With generative devices, this effort is short-circuited. In Bernard Stiegler, the short-circuit designates the mechanism by which technical memory (the memory deposited in objects) bypasses and eventually neutralises the long circuits of thought, desire and individuation. Access to simulated know-how is, moreover, I think, one of the reasons for the massive adoption of these devices.

Perhaps that is what shocks, at bottom. The effort needed to produce a text shrinks. Anyone who has tried to write knows what I am talking about. One has to think about the message and about how to express one’s ideas, put on paper what is going round in one’s head, find the right wording, decide which author to quote and how to connect one point to another.

The author who writes with a generative device can short-circuit this process, or even use it to appropriate a cultural memory, in the broad sense, that he has not built up and does not possess, accessing it through the device. In the absence of a warning, the discovery of this use, or mere doubt, produces an effect of usurpation. “He cheated”, some would say. A few months ago, several authors were claiming the “AI-free” label. Those who devote time and energy to building up a cultural memory almost see it as a matter of competition. It also results in a relative devaluation of texts written with the help of a generative device, notably as regards their form. I refer here to what I wrote on attentional ablation.

Painting allows me to take up the question from this angle. Today, in supermarkets, for a few euros, one finds works printed in large series. One also finds canvases painted with a brush by an artist, assisted or not in a workshop but necessarily influenced. The production effort is not the same for an image made on a computer and printed in many copies as for an oil painting two metres by three.

Are we not confusing things? Should we not also question the notion of the artist’s or author’s effort in the face of the use of generative devices?

Note on the use of a generative device

Given the subject, it seems only right to me to specify that the ideas, references and outline of this text come from my personal notes. I asked Claude (Anthropic) to put them into shape. I reread and amended the result before publication.

References

Notebook entries:

Glossary terms:

These notes are part of Duality, my digital notebook. If you would like to receive them directly in your mailbox, you can subscribe to the newsletter.