Unlearning the mineral mind: algorithmic shadows, geologic memory and the plasticity of intelligence

This article develops the concept of unlearning intelligence as a critical response to the extractive, predictive and universalist paradigms shaping contemporary artificial intelligence. Rather than approaching AI as an immaterial system, it proposes understanding it as a mineralized, plastic and relational process, grounded in geological infrastructures, open to rupture and constituted through distributed socio-technical relations. Drawing on media geology, individuation, plasticity and cosmotechnics, the article introduces the notion of the fissure as a key analytical and speculative figure: a point at which extractive logics break down and alternative epistemologies may emerge. Unlearning is articulated not as negation, but as a practice of interrupting prediction, destabilizing optimization and opening intelligence to plurality. By reframing AI through its material conditions and transformative potentials, the article argues for a planetary ethics of thought grounded in technodiversity, opacity and care.