With renewed prominence of Explainable AI (XAI), many areas are revisiting early work on Explainability. Within the broader field of Artificial Intelligence (AI), Ambient Intelligence (AmI) has an advantage in the development of transparent and ethical systems because such work has long been an integral part of research, development and operations in AmI. In this paper we argue that, because of the paradigm requirements of system intelligence, social intelligence and embedded-ness, AmI is uniquely prepared to support the push for ethical and transparent technology development. We argue that Ambient Intelligent Systems are well suited to infer when an explanation might be needed (and of what kind), and the form that it should take. We further propose AmI devices as mediators between humans and machines because they can combine social and technical systems in a fully embedded way.
展开▼
