Not mentioned in the article but a possible source discussed at an Economics of AI conference in Toronto[0] (among other places) is the idea of a restructuring

Not mentioned in the article but a possible source discussed at an Economics of AI conference in Toronto[0] (among other places) is the idea of a restructuring lag[1]. In effect, we as a society are investing an enormous amount of capital in activities that augment or replace existing economic activity but have not yet completed the transition. For example, auto and tech companies are investing heavily in self-driving vehicles. At the same time, very few vehicles have been economically displaced as a result of that investment. In effect, we went from spending $X on cars to $X+$Y where $Y is self-driving research. Thus, productivity as measured in an aggregate economic sense, starts to fall. If anyone knows of more research on this argument, I'd love to know more. [0] https://www.economicsofai.com/nber-conference-toronto-2017 https://www.economicsofai.com/nber-conference-toronto-2017 [1] https://static1.squarespace.com/static/59c2a584be42d60a2772ba71/t/59c2b0f78dd041729ad41800/1505931530673/Brynjolfsson.pdf https://static1.squarespace.com/static/59c2a584be42d60a2772b...