I feel his article leaves out some fundamental political aspects of the situation like I brought together in "Beyond a Jobless Recovery: A heterodox perspective on 21st century economics"). His article of course assumes consumer demand is infinite (despite Maslow's hierarchy of needs suggesting people more to more low-cost self-actualization activities over time). It assumes that the business benefits of employing a human will always outweigh the costs for many jobs (despite strikes, lawsuits, quality, illness, turnover). It assumes humans will always have special advantages over AIs and robots. It ignores whether some aspects of the economy (like long pipelines to become a professor) are really needed or are just protectionism. It ignores the social impact of rich/poor divides on working conditions and the operation of a capitalist economy itself. It ignores the value to the worker of the intrinsic nature of the work (i.e. some people may just be less happy in service jobs compared to agriculture or manufacturing). It ignores deeper issues of rethinking work as play (like Bob Black wrote about). It also ignores (incidentally, in relation to humans vs. robots) that "comparative advantage" only applies theoretically when you have "full employment". The article jumps between proving some points with numbers and then making other points as "strong hunches" or by quoting suggestions about technological unemployment from fifty years ago (quoting Herbert Simon). His prescription is of course mostly just more "education" — which is nice job security for a professor.