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

We introduce the _economic efficiency of the energy_ defined as the ratio between GDP produced and energy used at country level. We study its behaviour over time both from historical data and recent detailed databases. We observe that at the start of the first industrial revolution such quantity decreases dramatically. The comparison of the efficiency between underdeveloped, developing and advanced countries shows that the first have the highest efficiency and the second led on the third until around the year 2000 when a switch occurred due to a constant growth in efficiency of the advanced economies in the time window 1980-2018. The current AI industrial revolution is forecasted to have a growth impact on GDP several times larger than the first postwar (WW2) decade in western society. We argue that, since AI is still an energy eager technology, the danger of a collapse of the efficiency now is real in analogy with the first industrial revolution. A central question revolves around whether, and if so in what time scale, the AI revolution will follow an efficiency growth strong enough to enhance the global one.