Two approaches to issues of liability involving artificially intelligent beings

Abstract

We have yet to see an artificial entity that can be described as intelligent in the similar way as a human is. But we can look ahead to a scenario in which that prediction comes true—it’s just too bad that all the metaphors we have for thinking about the attendant issues seem to pit humans against artificial entities. I submit, by contrast, that we should view ourselves as sharing a space with these entities: we should consider our own liability for actions that could bring harm to them, and we can appreciate as much if we only look at our relation to animals. I argue that we can draw on that relation to better frame our relation to artificial entities.