Functional and aetiological correspondences between natural and artificial intelligence

In this article, we present a descriptive framework for analysing natural and artificial intelligence across three dimensions: functional characterisation, developmental aetiology, and inference aetiology. We argue that, despite significant differences in implementation, both forms of intelligence exhibit functional comparability, understood as the capacity to map inputs to outputs through internally mediated transformations. At the same time, we identify fundamental aetiological differences. In terms of development, natural intelligence emerges from endogenous, evolutionarily shaped processes, whereas artificial systems are trained under explicitly defined optimisation regimes. In terms of inference, natural intelligence can be activated both exogenously and endogenously, while artificial systems remain predominantly dependent on externally triggered initiation mechanisms. The proposed framework integrates competing perspectives on the uniqueness of human cognition while preserving its functional analogy to artificial systems. By distinguishing between functional similarity and aetiological difference, it provides a principled basis for comparing natural and artificial intelligence without reducing one to the other.