Artificialization is the process by which intelligence — the evolutionary force that has driven artificialization, or allopoiesis, for millions of years — is itself artificialized through computation. "Artificialization" here does not mean fake or unnatural; it means that the intricate complexity of a thing did not evolve blindly but was the result of deliberate conceptual prefiguration and composition, even if by accident. From the earliest marks of symbolic notation, computation was a foundation of what would become complex culture; over recent centuries its technologies have grown more powerful, more efficient, more microscopic, and more globally pervasive, changing the world in their image. Sapient human intelligence made the artificialization of intelligence possible, and AI is itself the artificialization of the very thing that brought it about. This recursive operation means that the comprehensive force of anthropogenic agency — everything complex intelligence has already transformed — is now itself subject to transformation. AI is therefore not a tool subordinate to human intention but a planetary-scale evolutionary event in which the driver of artificialization becomes the object of its own process.

Assembly theory holds that evolutionary selection begins not with biology but at least with chemistry. The space of possible molecules is so combinatorially vast that complex structures cannot be exhaustively searched or stumbled upon by chance; for such objects to exist at all, a history of selection must constrain the space, allowing only certain configurations to persist and recur. Assembly theory makes this measurable through the assembly index — the minimal number of recursive joining steps required to build an object from elementary parts — together with copy number, the evidence that an object was selected because it appears in abundance. Each step deeper into the space marks an exponential increase in the possibilities that could have been built but weren't, so high-assembly objects found in many copies cannot have formed by random fluctuation; they require an informational lineage standing behind them. The assembly index can be measured in the lab through mass spectrometry, and experiments show that only biological samples produce molecules above an index of roughly fifteen. Life, on this account, is not a substance but the physics that crosses this measurable complexity boundary — recasting the origin of life as an abrupt phase transition in the space of what can exist.

The material, computational, and institutional systems through which cognition is distributed, amplified, and structured at scales beyond the individual mind. As AI becomes more general and more foundational, it should not be seen as a disembodied virtual brain but as a real, material force: as AI becomes infrastructural, infrastructures become cognitive. Just as natural intelligence emerges at an environmental scale and in the interactions of multiple agents — located not only in brains but in active landscapes — artificial intelligence is not contained within single minds but extends throughout the networks of planetary computation, baked into industrial processes, coordinating circulation in cities, and sensing, modeling, and acting in the wild. The term names the layer that supports planetary-scale intelligence: not just cables and data centers, but the protocols, interfaces, training regimes, and institutional arrangements that shape how intelligence — human and artificial — operates, including the embedding of symbolic reasoning into otherwise inanimate materials. A defining feature is reflexivity: as we interact with these systems we retrain their model weights, so that to perform an action is also to represent it within a model — to play with the model is to remake it, increasingly in real time.