Perspectives and Opportunities in Trustworthy AI

The trustworthiness (or otherwise) of Artificial Intelligence has been much in discussion of late. Perspectives range from how we might make people trust AI to AI being not possible to trust, with many points in between. In this talk, I will examine these perspectives and also question whether or not many of these discussions somewhat miss the point, which is: since people are going ahead and basically doing their own thing anyway, we should probably help them. Acknowledging that trust is a heuristic that is widely used by humans in a range of situations, and drawing on literature concerning how humans make trust decisions, I will present a general model of how people might consider trust in AI (and other artifacts) for specific purposes in a human world. I will explore how the model forms a useful basis upon which to develop intelligent systems in a way that considers how and when people may trust them, and in doing so empowers people to make better trust decisions about AI. Further, trust is not only about the mitigation of risk, it is an essential mechanism for the coordination of groups. With this in mind, I will also highlight some ongoing fundamental research in socio-technical systems that suggests that explicitly designing for trust brings with it new opportunities for technology-enabled social action.