Thermodynamic neural networks and intersection theory: an ontological hypothesis of emergent intelligence
摘要
This article proposes a new ontological framework for describing cognitive processes, grounded in intersection theory and an entropy-based model of the space of truths. At the micro-level, we show that neurons and neuronal populations function as thermodynamic systems that reside in regimes of fluctuations, relaxations, and entropic transitions. At the macro-level, these processes manifest as ordered structures–the topologies of truths and their intersections–that jointly shape the cognitive landscape. We introduce the notion of ontological truth as the minimal quantum of reality, as well as the epistemic agent (individual or collective) who performs septation (partitioning), dividing the universal space of truths into known and unknown zones. On this basis, we formulate the entropic trajectory of cognition, which characterizes the evolving balance between the known and the unknown over time. We further show that synchronization of individual agents via the operator of temporal velocity