Communications of the ACM || Andrew Kahng of UC San Diego weighs in on reinforcement learning for chip placement.