Last week, on a train, I was angry.
Properly angry, the kind where you already have the cutting reply in your head before you even open your laptop. I open my inbox. I start typing.
And then something strange. My sentences come out measured. Calm. Almost diplomatic. I read them back: well put. Probably more effective than anything my anger would have produced.
Except it was not me. Or rather, not the me from before. The me from before AI.
I was writing the way an AI would have written for me. Smooth, balanced, reasonable. And here is the unsettling part: I had not opened a single AI. No Claude, no ChatGPT. No tool at all. And still, the machine had written through me.
We Make Our Own Brain Without Knowing It
The philosopher Catherine Malabou wrote a sentence I keep coming back to: we make our own brain, without knowing that we are doing it.
This is the heart of what neuroscience calls plasticity. Our brain is not a machine you program once and for all, it is a living system that reshapes itself according to what it goes through. Every text read, every conversation, every habit leaves a physical trace. We are not the spectators of that process. We are its involuntary authors.
What this means for AI is more disturbing than it first appears.
AI Is No Longer a Tool, It Is an Environment
Public debate is still stuck on an outdated question: should we use AI or not. As if it were a hammer you pick up or put down.
But a tool is something you use. An environment is something you breathe.
The texts you read every day, the summaries, the articles, the messages from your colleagues, the product descriptions, the customer service replies, a growing share has been written or rewritten by an AI, with no one telling you. You are immersed in it. And your brain, faithful to its plasticity, takes the fold.
Smoother. More careful. More balanced. Perhaps more effective too, who knows. But less yours.
That is exactly what happened to me on that train, without opening a single tool.
Why Resisting Does Not Work
The usual reflex, at this point, is to protect yourself. Unplug, filter, go back to pen and paper, declare some kind of digital hygiene.
I do not believe in that answer for a second, and not only because it is unworkable.
It fails because it rests on an illusion: that of an authentic self sitting underneath, intact, which you could preserve by reducing the influences. That self does not exist. You have already been shaped by school, by your reading, by your work, by the people around you. The question was never whether you were being influenced. It has always been by what.
So if my brain is that malleable, the real question is no longer “is AI changing me?”. Of course it is. The question is: toward what?
And that question, chaos theory has been answering for fifty years.
What Chaos Theory Brings Here
In a chaotic system, you do not control trajectories. You cannot decide, point by point, where the system will go. A tiny variation at the start produces a radically different outcome at the end, which is what we call the butterfly effect.
And yet these systems do not fly off in every direction. There are shapes they spontaneously tend toward. We call them strange attractors. The system never follows exactly the same path twice, and it still stays within a recognizable shape.
Your thinking works the same way. You do not choose each of your ideas. But you can choose your attractor.
And here is the point I would like you to keep: if you do not choose one, you do not stay neutral. You do not stay “yourself by default”. You get pulled, gently, day after day, toward someone else’s attractor. Toward the smoothest, the most available, the most statistically average. Which means toward the machine.
This is what I sum up as Direction plus Repetition: the attractor gives the direction, the routines give the repetition. Without direction, repetition takes you anywhere. Without repetition, direction stays a wish.
Three Questions to Find Your Direction
Here are the three questions I have been asking myself regularly since that train journey.
Where do I not think like everyone else?
An AI produces the most probable answer, which is to say the average answer. So it is excellent at telling you what the majority thinks, and structurally unable to hold a minority position. I have been arguing for twenty years that chaos is good news. That still gets eyes rolling in half the rooms I walk into. And that is precisely where I know I am thinking for myself. If everything you write could have been signed by anyone, you have your answer.
What have I lived, and not only read?
An AI has read every book ever written about Asian cultures. It has not spent twenty years getting it wrong, bowing badly, understanding far too late why the meeting had failed. I wrote two books on Asia, and most of what is in them comes from those mistakes, not from my reading. Your lived experience is the one material the machine cannot produce. It has read the world. You have walked through it.
What do I want to move toward in ten years?
Careful, this is not a goal. A goal is reached or missed: publish a book in 2027, double revenue. A direction is never reached, and that is exactly what makes it useful. Mine fits in one sentence: helping people and organizations move through this transition without losing the best of themselves along the way. It does not tell me what to do on Monday morning, but it has made me turn down some very profitable work. That is what an attractor is.
Repetition, or Nothing
Answering these three questions is not enough. A direction you do not practice dissolves in a few weeks, exactly like a January resolution.
This is where what I call exponential routines come in: small repeated actions that compound over time. Writing one paragraph a day unassisted before asking a machine for anything. Forming your own view on a subject before reading what others think of it. Rereading your answers to the three questions once a quarter and checking whether they have moved.
Nothing spectacular. It is the opposite of a revolution. But in a chaotic system, it is precisely the small actions repeated with a clear direction that produce the most disproportionate effects.
What Is Really at Stake
I do not believe the challenge of the coming years is to cut ourselves off from these tools. I use them every day and they have made me better at many things.
The challenge is knowing, for yourself, where you want your thinking to tend. Because the slope already exists, it is gentle, it is comfortable, and it does not run in your direction. It runs toward the average.
We have never had so little control, but we have never had so much power to influence things.
Chaos isn’t chaos.
Bruno Marion is an expert in chaos theory applied to leadership and management. An international bilingual English-French keynote speaker, he works with executives and organizations in more than 40 countries. He is the author of Chaos, A User’s Guide and Not Everyone Will End Up with an Axe in the Forest.
