In policy, corporate, and media discourses, AI is often envisioned as the solution to global challenges and as a threat to humanity. Dominant hype/fear imaginaries leave little room for picturing alternatives because the future of AI appears fixed and inevitable. This reflects a crisis of imagination: technology-driven futures are easy to conceive, but it is difficult to anticipate plausible ways in which technology and society could interrelate within future socio-technical systems. This crisis is particularly vivid among AI researchers and developers who are used to operating within computational systems and thus tend to neglect societal contexts. As many of them want AI to create better societies, they search for ways to overcome the crisis.

This project seeks to study and enhance how AI researchers and developers imagine socio-technical systems. Through a co-design process involving stakeholders and publics, it will explore how dominant and alternative ways of imagining socio-technical systems relate to each other and shape AI research and development. The concept of system imagination will be coined to analyze the process. This concept will scrutinize the potential within practices of individuals and small groups to connect seemingly conflicting ways to imagine socio-technical systems. The goal is to find out how the crisis of imagination can be addressed effectively so as to help AI researchers and developers better align technology-driven futures with society-oriented alternatives.

This research builds on Mareike Smolka’s experience in the NeuroSys Cluster4Future where she studies and enhances the imagination of scenarios for the future of neuromorphic computing. Her new project will start in January 2025 in collaboration with AI researchers and developers from OnePlanet Research Center. For more information about Mareike Smolka and future updates on the project, you can have a look at her website.