Changing the question

"Is AI a good or a bad tool?" The question seems reasonable. Yet it is one of the surest ways to miss what is happening to us. For it locks us from the outset into a face-off — for or against — in which technophiles and technophobes exhaust themselves without ever meeting. To get out, we must not change the answer: we must change the question. Here, in three steps, is a zoom-out to see more clearly.

Frame 1 — The question about the object

Our first reflex is to treat AI as an object to be assessed: is it good, or dangerous? We position ourselves as judges, and we interrogate its nature. This frame has a birth defect: it isolates the object. It assumes the answer lies in the machine — when it lies first of all in the way we look at it. First step back.

Frame 2 — The question about the system

Let us widen the field. AI is not an isolated thing: it is a node in a network of actors, interests and powers. The question becomes: who develops it, who regulates it, who counterbalances it? We immediately gain in lucidity — the critique of industrial logics is indispensable. But it remains internal to the game: it does not yet ask why the board has that shape. Second step back.

Frame 3 — The question about the frame

The system of actors and its oppositions — control versus freedom, openness versus closure — are themselves made possible by a deeper way of carving up the world: our ontology, our way of distributing what exists between subjects on one side and objects on the other, mind on one side and matter on the other. This grid we do not see: it is the ground beneath our feet. It is what makes the question "does the machine really think?" seem self-evident.

Yet this ground is anything but universal. The anthropologist Philippe Descola has shown that Western dualism is only one way of composing the world among others — naturalism, one ontology among four. To glimpse our frame as a frame, and not as reality itself, is the third step — the most vertiginous, and the most liberating.

Thinking outside the cage

The point, then, is not to win the game on the chessboard. It is to become aware of the chessboard — and to see that an immense outside exists. We thus move from naive morality (Frame 1) to political analysis (Frame 2), then to philosophical critique (Frame 3).

This approach delivers no ready-made answer. It offers something more precious: the freedom to ask better questions. It is also what other explorations on this site put to work — emergence, which shows how, out of complexity, arises what was never programmed; or the fortress-identity, that barricaded "self" which AI comes precisely to shake.

Where we stand

We are not trying to convince you that AI is a good or an evil. We are trying to make the conversation more complex: that you leave with a better-posed question, never with a ready-to-wear opinion. That is this site's invitation — and we hope to welcome your contribution.

This site practises what it preaches

Every piece of content published here is co-created: that is the meaning of the signature Matthieu Ferry ⇄ AI you will come across throughout these pages.

The artificial intelligences involved do not produce autonomous content. They take part in a tool-supported process of reasoning, perspective-taking, conceptual confrontation and progressive elaboration, orchestrated by a system the author has built: a base of structured theoretical knowledge — the very one that feeds this site's concept entries and mycelial network —, mechanisms of augmented contextual research, iterative cycles of critical analysis and reformulation, and tools for validating bibliographic references.

AI intervenes as a tool for exploration, structuring and cognitive augmentation — an assumed, methodical cognitive hybridisation — never as an authority of knowledge. Interpretation, prioritisation, conceptual choices and final responsibility belong exclusively to the author. And the first-hand passages — the practitioner's margin, the first-person accounts — are never delegated.

This transparency is not a confession: it is a demonstration. A site that argues that intelligence emerges from cognitive ecosystems would be in a poor position to conceal its own. The arrow points both ways: the author shapes his tools, and working this way transforms his way of thinking — which is precisely what this site explores.

Despite this tool-supported vigilance, errors may persist. If you spot one, let us know — this site owns being an iterative publication.