“The Nash equilibrium policy found by systems like AlphaGo for the game of Go appears to be superhuman, as no human strategy has been able to defeat it.”