Brain as Hierarchical Prediction Machine
The brain operates as a hierarchical prediction machine that continuously generates top-down predictions about expected sensory input and propagates only prediction errors upward for model updating.
This is an axiom because it represents a foundational theoretical framework for understanding brain function—the predictive processing or predictive coding theory. It's irreducible in proposing a unified computational principle governing perception, action, and learning across the entire cognitive architecture.
The predictive processing framework, developed by researchers like Karl Friston, Andy Clark, and Jakob Hohwy, proposes that the brain is fundamentally a prediction engine. Higher levels of the cortical hierarchy generate predictions about activity at lower levels; only when predictions fail (prediction error) does information flow upward. This explains diverse phenomena: perception (minimizing sensory prediction error), action (minimizing proprioceptive prediction error by moving), attention (precision-weighting prediction errors), and learning (updating generative models). Neural evidence includes repetition suppression (predicted stimuli generate less activity) and the hierarchical organization of cortical processing.
This axiom is foundational for the curriculum because it unifies many cognitive phenomena under a single principle and explains why prediction is central to cognition. It connects to schemas (sources of predictions), surprise (large prediction errors), attention (precision control), and learning (model updating). Understanding the brain as a prediction machine explains why we're blind to the expected, why violations of expectations are salient, and why active engagement (generating predictions) enhances learning compared to passive reception.