I spent six years making tiny silicon membranes tell us — this is the story.

Curiosity, mostly. I've always been fascinated by science — math, physics, chemistry — not as disciplines but as lenses. Nanotechnology felt like the ultimate lens: what happens at the scale of the very small doesn't just matter, it governs everything we see at human scale. And the rules down there are deeply counterintuitive. I wanted to be inside that — not just reading about it, but contributing to it. The chance to run an experiment that no one had run before, on materials that barely anyone had touched. That was the pull.

No single moment — more of a confirmation that built across two internships. In my final master's year, I went to the Institut des Nanotechnologies de Lyon to work on foams, Janus particles, and microfluidic systems. What struck me wasn't just the science — it was how visual it was. Bubbles, emulsions, droplets: you could see the physics happening in real time. Then I was offered a second internship in Paris, with Dan Angelescu, building a micro-emulsion generator — a lab-on-chip that produces droplets and bubbles at the micrometric scale. By the end of that, I didn't have a revelation. I had a certainty.

I didn't finish my PhD — and that's part of the story. Over three years of research, I worked on micro-emulsion generation, co-developed an optofluidic inclinometry sensor capable of detecting bridge pylon deflections as small as 0.005 degrees, and contributed to Fluidion, a startup building autonomous water quality sensors for remote environments. Real work, real science. But the path ended there. What came next surprised me: I moved into consulting, helping Paris-based startups recruit German sales and marketing talent to crack the German market. Far from the lab — and yet I never stopped thinking about science and research, this time around topics like AI, neural networks, and the mathematics underneath. That parallel obsession led me, quite naturally, to computational creativity: the intersection of artificial intelligence, computer science, art, design, and psychology. The question that hooked me was simple and enormous at the same time — how do ideas emerge? Are there reliable, reproducible mechanisms to provoke creativity? That fascination became a research focus. And eventually, a paper.

Artificial intelligence — and the dizzying speed of it. Every day, a new model, a new capability, another frontier crossed. What obsesses me isn't the technology itself but the question underneath: how do we combine AI with science and research to create something greater than either alone? I want to see AI go beyond being a tool. I want to see it become a genuine research partner — one that helps scientists transcend their own cognitive limits, accelerate discovery, and eventually contribute to solving what has resisted human effort for generations. Cracking diseases. Unlocking nuclear fusion as a limitless energy source. Extending human lifespan. Enabling deep space exploration. That's not a fantasy to me. It feels like the natural next step of what's already happening.

What fascinates me most about artificial intelligence is the black box. We have a working theory of what's under the hood — neural networks, weighted nodes, fine-tuning — and yet something fundamental remains unexplained. We don't truly understand, at a granular level, how Transformers do what they do. What strikes me even more is the gap between the mechanism and the output. At its core, a large language model is doing something almost embarrassingly simple: predicting the next most probable word, one word at a time. That's the whole trick. And yet from that single, repetitive operation emerges scientific papers, original research contributions, full novels, science fiction, poetry. The output is extraordinary. The mechanism is elementary. That gap — between what it is and what it produces — is one of the most genuinely mysterious things I've encountered. And we're living inside it.

The Concept Collider is built on a simple but radical bet: that creativity can be mechanical. Take two distant concepts, collide them, and what emerges transcends both originals — a new idea that neither concept could have produced alone. What excites me isn't just the individual collision. It's the system. A reproducible, transparent process that generates creative output automatically, continuously, indefinitely. Knowledge production that doesn't stop. If this works — truly works — it means that the spark of creative insight, long considered unpredictable and deeply human, can be systematized. Not replaced. Systematized. That thought fills me with equal parts scientific satisfaction and genuine joy.

No single person. I'm usually deep in the problem itself — the task, the mechanism, the idea. But I always hold in mind the broader frame: the researchers who spent their careers on these questions before me, across different eras, different disciplines. Computational creativity didn't begin with me. It has a lineage. And I think every scientist feels this, at some level — the sense of building on something already there, picking up a thread someone else left, adding a small seed in the hope it helps something else grow. Being a tiny piece in a much larger gear. What drives me is the thought that, at my own modest scale, I might contribute something real to a discipline that others have given years of their lives to build. Being part of that history — even the smallest part — fills me with equal measures of pride and humility.

Computational creativity is the field that asks how far computers can go — first in helping humans think, create, and discover (thus acting merely as a tool), and now, increasingly, in thinking, creating, and discovering on their own. That's the deeper question: not just augmentation, but autonomy. Can a machine genuinely originate an idea? And what I'd want anyone to understand is that we're living through a genuine transition: from the computer as a tool to the computer as a creative partner. Co-creation. That shift is going to accelerate research in ways we can barely anticipate, with consequences across almost every domain of human life. What I'm trying to do, at my level, is to leverage that shift — to take this new kind of partnership and apply it to questions I find worth asking, to see what new concepts can emerge when human curiosity and machine capability genuinely work together.

Learning, mostly. When I'm not working, I find myself pulled toward new domains — a new language, a new culture, how people live on the other side of the world. Space technology, self-driving cars, eVTOLs, biogenetics and CRISPR, finance, nutrition, human psychology — the list has no obvious logic except curiosity. I cook, try new recipes, experiment with ingredients I don't know. I play racket sports obsessively: tennis, padel, table tennis, badminton — anything with a racket, really. I recharge by spending time with friends, going into nature, walking in forests, being around animals. I have a deep soft spot for dogs. But I'll be honest: I have a hard time truly switching off. When something captures me, the hours disappear. That applies to research as much as everything else — it's less a discipline than a disposition.

Distance, mostly. When I'm stuck, the worst thing I can do is push harder. The best thing is to completely change subject — let time pass, let my mind wander somewhere else entirely. When I come back, something has shifted. A different angle appears. I can't always explain why it works, but it does. The same goes for walks in the forest, classical music, white noise — or the sound of the shower, which sounds trivial until you realize that some of your best ideas have arrived there. These aren't distractions. They're the conditions under which my brain, left to its own devices, quietly finds new angles on things I couldn't crack head-on.

Two things that might seem trivial, and apply well beyond research. First: it's a marathon, not a sprint. And a marathon is long, and it hurts. Anyone who tells you otherwise hasn't run one. Second: the goal — getting the degree, finishing the thesis — matters less than you think. The real value is in the path. The questions you learn to ask, the failures you survive, the slow accumulation of understanding that happens whether or not you notice it. And underneath both of these: nothing of real worth is created without effort. Real growth requires passing through something difficult. There are no shortcuts to a new state. I wish someone had told me that plainly, early on — not as a warning, but as a preparation.