What can a future, human-centered relationship between humans and AI look like? Many concepts assume a collaborative relationship of mutual support that is characterized by symmetry. The demand for “human oversight” as control over AI-supported processes or AI-based decisions tends to assume an asymmetry on the basis of which humans should be able to intervene. The presentation will explain that there are a variety of constellations beyond the usual schemes that can be considered to design the relationship between humans and AI. This is not just about processing tasks, but also about supporting learning on the human side and continuously developing the technology on the AI side. What the relationship between humans and AI looks like in specific application situations ultimately depends on how technical characteristics and organizational practice are coordinated in a socio-technical design. To this end, basic variants can be identified that are relevant both for the practical use of AI and for research on AI.
Dieses Ereignis ist vorbei.