This article develops the thesis of an asymmetric ontological complementarity between artificial intelligence and human creativity and draws out its consequences for the production of knowledge. Building on a theoretical framework developed in prior work—the probabilistic turn (Rosiñol Lorenzo, 2026a, 2026b)—the article argues that AI performs what I call subjectless inductivism: exhaustive extraction of regularities at an unprecedented scale, yet constitutively incapable of transcending the statistical framework from which it operates. The human, for their part, possesses the capacity for abduction (Peirce): the generation of hypotheses that violate prevailing plausibility equilibria. The difference is not one of degree but of kind, and it is precisely this mutual irreducibility that makes the combination productive. The article identifies two consequences of this complementarity. The first is epistemological: AI, by navigating transversally the probabilistic infrastructure of multiple disciplines and languages, dissolves the plausibility silos that historically compartmentalised knowledge, turning interdisciplinary exploration—once serendipitous—into a systematic possibility. The second is civilisational: the combination of ontological complementarity and silo dissolution opens the possibility of an Aufhebung of knowledge—not a return to the premodern unity of knowledge but a supersession that preserves disciplinary depth within a new symbiotic integration between the trained human and the navigating system. The article concludes by noting that the same categorical difference that grounds this potential also grounds its risk of inversion: if the complementarity collapses, the outcome is not acceleration but plausocratic intensification.
José Rosiñol Lorenzo (Thu,) studied this question.