Sean Carroll's Mindscape: Science, Society, Philosophy, Culture, Arts, and Ideas 87 | Karl Friston on Brains, Predictions, and Free Energy
15 snips
Mar 9, 2020 
Karl Friston, a leading neuroscientist from University College London, discusses his groundbreaking free energy principle, which proposes that our brains constantly model the world to minimize surprises. He explains how this principle connects neuroscience to mental health and decision-making. The conversation dives deep into predictive coding, entropy, and how our understanding of the brain could enhance treatments for disorders like schizophrenia. Friston also reflects on how insights from evolutionary biology shape our cognitive abilities and future planning.
AI Snips
Chapters
Transcript
Episode notes
Woodlice and Early Insights
- Karl Friston's earliest scientific insight involved observing woodlice.
- He realized their avoidance of sunlight could be explained by simple movement mechanics, not purposeful behavior.
Free Energy and Surprise Minimization
- Brains minimize surprise by creating a model of the world that aligns with sensory input.
- Free energy, an information-theoretic concept, not thermodynamic energy, quantifies this difference between expectation and reality.
Predictive Coding and Perception
- Predictive coding, a method for compressing data, minimizes prediction error.
- It's an instance of free energy minimization applied to perception, without considering action.