Something from (almost) nothing: complex lessons from simplicity*

I describe the use of minimal renormalization-group inspired/justified models in non-equilibrium statistical physics and biology. Using examples chosen from my own work in pattern formation, turbulence and biology, I point out that minimal models are able to make specific parameter-free predictions of appropriately-chosen experimental quantities, an outcome of universality in somewhat surprising settings. I argue that these considerations account for the unreasonable effectiveness of theoretical physics in describing seemingly complex phenomena.