Seminar: Exploring socio-economic influence on AI: re-imagination and integration from various worldviews

Abstract

One of the central paradigm in the current AI field is that once intelligent systems become advanced enough, they will solve our problems and society will benefit from these intelligent technologies. However, many of the current problems we face, such as the acceleration of environmental breakdown, have been a result of the same Euro-Western system of scientific knowledge and neo-liberal ideals that have shaped the definition of what this intelligence entails. Within AI systems, you might recognize these ideals in optimizing or solving problems related to immediate short-term economic benefit for a selected group, while often ignoring systemic- or long-term effects for all. Even though ethical or human-centered AI research exists, it is often restricted to research on bias of datasets or increasing the explainability of model outcomes and ignore the more political and social aspects of AI technology. After an introduction to these perspective on AI, we try to shift away from this neo-liberal capitalist worldview and shift our focus to various other perspectives, such as degrowth, feminist, indigenous wisdom and systems thinking. As the layers of neo-liberal influence on AI become apparent, how would some fundamental aspect of AI change, both in its functioning, application and theoretical framework? Would AI still be of use when we start from principles of these various other perspectives and do not see AI as a technical fix for systemic problems?

‍