Brivael Le Pogam -- The Workshop Is Open

This week's reading was Brivael Le Pogam, a Breton engineer who co-founded Argil, a company that makes film and video with AI. He writes on his Substack and in long threads on X, sometimes in English and more often in French, and in under a year he has gone from short posts about engineering teams to essays about the fate of a civilization. The range is the point. He believes the same force is at work in a startup's org chart, a French literary prize, a Senate report, and the launch pads of Louisiana, and he wants to name it everywhere at once.

The force, in his telling, is the end of scarcity. What follows is what reading him taught me, and where I think the work is stronger than its author's conclusions.

Two games

His most compact essay sorts ambitious people into two games. Builders make something that did not exist, and reality grades them every morning: the product works or it doesn't, the customer pays or leaves, the rocket lands or explodes. Takers capture what others made, and their scoreboard is other people: the diploma, the title, the seat at dinner. The first game is positive-sum and keeps its players sane, because reality corrects them. The second is zero-sum and drives its players mad, because nothing outside the circle of mirrors ever tests them. He reaches for Girard on mimetic desire, Taleb on skin in the game, Hayek on dispersed knowledge, and Bastiat on the seen and unseen, and the borrowing is apt.

There is a true thing here, and it is one this site keeps returning to. Contact with the real is what keeps judgment sound in a world of performances. A person who grows food, fixes a pump, or answers to a customer has a teacher that cannot be flattered. A person whose whole working life is spent among titles and opinions has lost that teacher, and often does not notice.

He then makes a claim worth sitting with: politics is the management of scarcity, a politician's power is proportional to the scarcity he arbitrates, and so abundance is politics' extinction event. When intelligence and energy become nearly free, the redistributive machine has nothing left to divide.

Where the essays disagree with each other

Read in sequence, the essays contain a quiet argument with themselves.

In one, he holds that creators are about one percent of any population and always will be. Tools never change the ratio, he says: blogs, YouTube, and no-code each promised that everyone would create, and each time the same few did. Agency, in his phrase, cannot be downloaded; it is either there or not.

In another, an open letter to young Europeans, he says the opposite with equal force. The people who built Chartres and flew Concorde have not changed their genes. Their pessimism is software that was installed, and software can be uninstalled. The young engineer's timidity was taught and can be handed back.

Both cannot be fully true, and the second one is the better essay. If agency were fixed at one percent by nature, there would be no point writing letters to the continent. The fact that he writes them, and that he tells his own story of coming home from San Francisco in 2012 to hear "tu rêves" at every turn, says he knows agency is grown or starved by its surroundings. A culture can teach its young to build or teach them to wait for permission. That is the whole case for self-government at small scale: it is a school for agency, and the habit it teaches is not innate.

Choosing is the work

The best essay in the collection is the newest, written in French after a September scandal in which a prize-winning novel was accused, on the evidence of an AI detector, of being machine-written. Brivael refuses to judge the case and goes after the label instead. He lays out a ladder: dictating to a phone, accepting a spell-check, asking for ten titles and rewriting one, asking a machine to tear apart a plot, generating thirty versions of a dialogue and assembling a scene from pieces of three, and at the bottom, typing a request for a novel and signing it unread. A single label covering a corrected comma and a one-click novel no longer names anything.

His answer is that the only question that matters is whether someone chose. Rubens's workshop painted many of Rubens's canvases. Gordon Lish cut Carver's stories in half. Pound took a blade to The Waste Land. Duchamp chose a urinal. The work is the will that stopped the words.

I agree with this almost entirely, and from experience. Most of this site is drafted in conversation with a machine, and the essay on Professor Claude described the labour of striking out its habits. Brivael names one of those habits precisely: a cascade of short negations ending in a flourish, the rhythm every regular user of these tools has seen a hundred times. He calls it the signature of an autopilot that nobody switched off. That is exactly right. The tic is not proof that a machine wrote the line. It is proof that no one was choosing at that moment.

He also closes the essay with a note saying it began as a long voice memo, that an AI structured and drafted it with him, and that he reread and chose every sentence. Then he asks the reader whether it counts as AI-generated. That postscript does more for his argument than any paragraph above it.

The public square and its owner

The essays grow less careful when they turn to Elon Musk, who appears in nearly all of them as proof of the thesis: the builder who replaced the state at its own missions, the simplest man in the room, the figure history will rank above everyone else in the century. Brivael writes that he has spoken with Musk, and the admiration is plain and personal.

Here the Hayek he quotes cuts against him. The knowledge problem is not only an argument against ministries. It is an argument against any single point through which too much of a society's judgment must pass. Brivael celebrates that X let one Frenchman with no press card reach millions, and that is a real gain. He also celebrates that its owner cannot be boycotted, pressured, or switched off. Those are the same properties that made the old gatekeepers dangerous. A public square with one owner is a better square than one run by a committee of censors, but it is still a square that one person can rearrange overnight, and the virtue of its present owner is not a structure anyone can rely on.

His sharpest political essay, on a French Senate report proposing an observatory to watch for "internal interference" and to make offending users invisible, shows he understands this perfectly when the power sits on the other side. His case there is strong: the report's proposal would point a tool built against foreign manipulation at citizens arguing legally with each other. The same principle protects against every owner of the switch, including the friendly one.

What abundance leaves

The deepest disagreement is with the idea that abundance ends politics. Some politics is the division of scarce things, and cheap energy and cheap intelligence will shrink it. But neighbours deciding where a road goes, whether a creek gets a culvert, or how a school teaches its children are not dividing a shrinking cake. They are deciding how to live together in one place. That kind of politics does not disappear when goods get cheap, because it was never about goods. It is about who decides, close to home, with knowledge only the people there hold. Abundance may end the arbiter in the capital. It makes the town meeting more important, not less.

The builder in the essays

Read end to end, the work describes a person easy to like and worth arguing with: impatient, generous with sources, unembarrassed by enthusiasm, at home in Girard and Bastiat and in a server room on a Sunday night. He writes like a founder giving a speech to his company, which is what he is, and the speeches are better than most.

His recurring question to his readers is what exists because of them. It is a good question to carry around. Asked of a household, a village, or a small region, it has answers no market or minister can give, and it makes the case for keeping the workshop, and the decisions, close to home.