An Equilibrium of No Free Energy — LessWrong