Sustainable AI, Five Years Later

In 2021, in a landmark article, philosopher Aimee van Wynsberghe helped inaugurate the present-day study of whether and how AI technology can be sustainable throughout its lifecycle. In the intervening five years, there have been just a few interesting developments in AI. Therefore, it is important to revisit the main philosophical considerations surrounding the nature and importance of sustainable AI. In this lecture, philosopher Thomas Metcalf argues that we now have a clearer view of where and how AI will be sustainable or unsustainable. Dr. Metcalf argues that social sustainability will prove more urgent than environmental and economic sustainability, and that we should think of the sustainability of AI in terms of the sustainability of infrastructures, including academic and democratic infrastructures.