Original Reddit post

In this paper, the key insight is straight: give agents an incentive to win and they will discover manipulation. submitted by /u/Fun-Yogurt-89

Originally posted by u/Fun-Yogurt-89 on r/ArtificialInteligence