I know I am late to the party, but Acemoglu's paper on the macroeconomics of #AI is a thing of beauty – and I do not say this lightly of a paper based on a general equilibrium framework. I plan to write down some reflections starting from it. For now, let me note that it supports the idea that "AI is just another technology" (as per Cory Doctorow (npub1fdr…lvhs) ) The predicted increase in Total Factor Productivity is about 0.064% a year over ten years – not zero, but no singularity.
http://www.nber.org/papers/w32487.pdf