This paper suggests to take seriously the idea, proposed by Le Cun, that AI could be seen at “a new phase of the Enlightenment”. It proposes the idea that an education for democracy with AI is possible under two conditions which are a basic feature of an ethics by design of AI and the two faces of the same coin: the expansion and enrichment rather than restriction and impoverishment of humanity at an individual, generational and cultural level. First, AI should neither be designed, nor developed or used for deskilling humanity but rather designed, developed and used for the preservation and extension of human capacities. Second, AI should not promote any form of technological colonialismbut rather aim at preserving and extending cultural and intellectual diversity. If we fulfill these two important criteria, we can thus think about AI not in terms of alarming risks but rather in terms of a constructive future promoting themotto of the Enlightment: “dare to think” and thus enabling an education for democracy.

The paper argues that recent advances in AI have challenged us to reevaluate how we define our relationship with these new intelligent systems and with this our understanding of what constitutes a sensory being. It identifies the need for a new multidisciplinary research agenda within which this reevaluation can be carried forward, while discussing some of the essential components of the research, including the concepts of sensory-centric and the details of communication that should create an experience for humans that is at heart educational. The paper highlights the implications of this conceptualisation in equipping citizens to become fully informed in order to render democracy effective together with corresponding approaches to AI regulation.