Steady State Learning and Nash Equilibrium