MUHAI is a European consortium funded by the EU Pathfinder program that studies how it is possible to build AI systems that rest on meaning and understanding. We call this kind of AI meaningful AI in contrast to AI that rests exclusively on the use of statistically acquired pattern recognition and pattern completion. Because meaning and understanding are rather vague and overloaded notions there is no obvious research path to achieve it. The consortium has therefore set up a task early on in the project to explore how understanding is being discussed and treated in other human-centred research fields, more specifically in social brain science, social psychology, linguistics, semiotics, economics, social history and medicine. Our explorations have yielded a wealth of insights: about understanding in general and the role of narratives in this process, about possible applications of meaningful AI in a diverse set of human-centred fields, and about the technology gaps that need to be plugged to achieve meaningful AI. This volume summarizes the outcome of these consultations. It has three main parts: I. A general introduction, II. A series of chapters reporting on what understanding means in various human-centered research fields other than AI, and III. A short conclusion identifying key research topics for meaning-based human-centric AI.

This is the first volume in a series exploring how AI systems can construct rich models of problem situations for deliberative intelligence