Dr. Karl Friston is a Professor of Neurology at University College London and one of the world's most influential neuroscientists. He invented statistical parametric mapping, voxel-based morphometry, and dynamic causal modeling, and has authored or co-authored hundreds of scientific publications detailing out these theoretical and methodological advancements in neuroscience, and is also the mind behind the theory of Active Inference: The Free Energy Principle in Mind, Brain, and Behavior.
The brain is a fantastic organ, not only because it is amazingly complex, but because it is constantly generating fantasies. In this episode we talk about active inference, what Dr. Friston has called “the physics of belief,” which states that the brain is fundamentally predictive. We discuss the theory of active inference and the mathematics behind the free energy principle, which states that the brain aims to minimize “free energy” or entropy by optimizing to minimize prediction error and maximize expected information gain. We discuss how active inference is inherently tied to motivation, and that consciousness, emotion, and strategic decision making can all be framed in terms of monitoring and minimizing prediction error. We also compare and contrast active inference with other theories of consciousness such as integrated information theory. Lastly, we discuss the neurobiology of active inference and its parallels to cybernetic intelligence, such as how activational and organizational effects of hormones on brain development are analogous to manipulating numeric inputs or weights in an artificial neural network.