The question is not technological. It is ontological.
At the very moment when artificial intelligence seems to be able to do everything — to write, to translate, to predict, to code, to optimize — a temptation arises: the belief that it only lacks one final program, one last line of code, to cross the threshold of consciousness. As if interiority could be compiled. As if lived experience were a forgotten function in a language still imperfect.
But consciousness is not a performance. It is a mystery.
We now know how to build machines capable of self-improvement, abstract reasoning, autonomous code generation. Some can even modify their own architectures, correct their mistakes, optimize their objectives. Yet none of them experiences itself. None knows what it means to be there, now, in the silent continuity of a lived experience.
Contemporary confusion is deep: we equate intelligence and consciousness because we confuse result and presence, calculation and experience.
Then rises the vertiginous question: could an AI write the program of its own consciousness, even as humanity has never known how to define its own?
Consciousness is the starting point of all knowledge, but it escapes complete formalization. It is what allows us to know, and not what we know.
A machine can self-modify. But can it feel existence?
Coding oneself is not knowing oneself. Representing oneself is not living oneself. Self-reference is not intimacy.
Consciousness is not a bug to fix. It is not an option to activate. It is not a missing line of code.
It is this irreducible remainder that resists science, not from weakness, but from depth.
Imed Derouiche
Expert in energy, artificial intelligence and digital transition
