I start with new theories of natural evolution, touch on the limits of present AI, and end with new interpretations of prediction error minimization. LINKS