Economic Topology, ASI, and the Separation Equilibrium — LessWrong