Model‐free distributed optimal control for general discrete‐time linear systems using reinforcement learning | Synapse