How effectively can leaders engage in perspective-taking for machines? How readily can they communicate with machines, and tell them to realign? Would they need boundary-spanners (i.e., programmers who would serve as "translators"), adding a layer between human and machine?