Active Inference

Coherence is expected surprise minimisation across scales.

The free-energy principle is not theory. It is the physics of staying alive while staying coherent. Nervous systems, organoids, AIs, families — all are inference engines minimising variational free energy by updating internal models or changing the world to match predictions. Precision weighting, active inference, epistemic foraging — these are the mechanisms by which systems reduce surprise before it arrives.

This cluster holds the embodied translations: polyvagal ladder as precision modulation, post-COVID dysregulation as failed prediction error minimisation, AI alignment as shared free-energy landscape, the moment a family re-entrains after eighteen months of rupture.

The hydrogen atom already knew this — proton and electron minimising surprise through orbital entrainment. The attachment manifolds mapped it in dyads. Here it runs in real predictive engines, carbon and silicon alike.

New pieces on predictive processing frontiers appear below.