I don't think the decoupling argument has been made as conclusively as everyone seems to be assuming. Look at the source in the OP for this [1]: These are some seriously misleading graphs.
First, plotting both labor productivity and GDP doesn't make much sense, as the former is almost exactly determined by the latter. They're just measuring GDP.
Secondly, they zoom in on the post-WWII era, then proceed to ignore the biggest long-term economic issues of the time period: The hangover from WWII, with the corresponding stable demand for US infrastructure and labor as the world rebuilt itself, and the housing/derivitaves bubble, as a combination of fraud and self-delusion caused the computed GDP numbers to shoot off into the hypothetical-wealth stratosphere even as the economy was faltering.
Once you account for the elephants in the room, there's nothing surprising left in the graph that would lead you to believe that the robotic obsolescence of labor is around the corner.
First, plotting both labor productivity and GDP doesn't make much sense, as the former is almost exactly determined by the latter. They're just measuring GDP.
Secondly, they zoom in on the post-WWII era, then proceed to ignore the biggest long-term economic issues of the time period: The hangover from WWII, with the corresponding stable demand for US infrastructure and labor as the world rebuilt itself, and the housing/derivitaves bubble, as a combination of fraud and self-delusion caused the computed GDP numbers to shoot off into the hypothetical-wealth stratosphere even as the economy was faltering.
Once you account for the elephants in the room, there's nothing surprising left in the graph that would lead you to believe that the robotic obsolescence of labor is around the corner.
[1]: http://andrewmcafee.org/2012/12/the-great-decoupling-of-the-...