Predictive Processing

Predictive Processing (PP), also known as the predictive brain or predictive coding framework, proposes that the brain is fundamentally a prediction machine that constantly generates models of the world and updates them based on prediction errors. Developed through contributions from Karl Friston, Andy Clark, Jakob Hohwy, and Anil Seth, among others, PP has become an influential framework for understanding perception, action, and consciousness.

The core mechanism involves hierarchical generative models that predict sensory input at multiple levels. The brain continuously generates top-down predictions about incoming sensory data, and only the discrepancies—prediction errors—propagate upward through the hierarchy. Conscious perception, on this view, is the brain's "best guess" about the causes of sensory signals, constructed through a process of hierarchical inference. Precision weighting determines which prediction errors are given more influence, allowing the system to balance prior expectations against sensory evidence depending on context.

For consciousness specifically, PP suggests that what we experience is not raw sensory data but a constructed model designed to minimize surprise. Anil Seth's extension of this framework proposes that the self and emotions are also predictive constructs—we perceive our bodies and emotional states through the same inferential machinery. This "controlled hallucination" view explains phenomena like perceptual illusions, the sense of presence, and even the feeling of being a self as products of the brain's predictive modeling. PP has been particularly influential because it provides a unified computational framework that connects perception, action, attention, and consciousness.