New Perspectives on Growth and Sustainability. Glimpses into the AI Industrial Revolution

We study the concept of _Economic Productivity of Energy_ (EPE), defined as the ratio of GDP produced and energy consumed at country or groups of countries level. We examine its behaviour over time using both historical data and recent, detailed databases. Our study compares the EPE for three different groups: underdeveloped, developing, and advanced countries. The underdeveloped countries exhibit the highest EPE, while an inversion occurs between developing and advanced countries around 2004, driven by a steady growth in EPE of the advanced economies from 1980 to 2018. Notably, during the onset of the first industrial revolution in England and Wales, the EPE decreased dramatically, with the trend reversing only decades later. We argue that, given AI’s current status as an energy-intensive technology, the risk of a collapse in EPE is significant, in analogy with the first industrial revolution. However, AI is projected to generate GDP growth several times greater than the one seen in the first postwar (WW2) decade in Western societies. Additionally, AI’s impact and adoption are expected to be more pronounced in advanced economies. We stress the need to investigate further these crucial questions: whether, when, and how the AI revolution will affect the currently positive global trend in EPE.