The ever-increasing rate of new technologies, once science-fiction, might lead us to believe that anything is possible. Recent advances in Artificial Intelligence that allow us to talk to our virtual assistants by simply calling out "Siri", "Hey Google" or "Alexa" can lull us into a sense that we are indeed talking to an intelligent entity. However, all too often things don't go quite as we expect and we end up with a frustrating experience. This raises questions such as ‘What impact do the apparent human qualities have on our expectations?’ and ‘What shape or form might the interaction take to improve communication between the AI and the human?’
In this UX SIG event we will explore these issues with help from Iulia Ionescu from the University of Arts London who has been carrying out research on the way people perceive virtual assistants and how this affects their behaviour; Libby Miller from BBC Research & Development who has been exploring how subtle gestures and appearance can communicate expressivity in robotic puppets and Tristan Ferne, also from the BBC, who has been investigating methods of making AI more understandable and explainable, and how that might improve trust.