standardised position description or all your jobs are belong to us

Via Marginal Revolution, Oxford economist Daniel Susskind that the disruptive—and hopefully welcome—effects of machine-learning on the labour market is far, far underestimated. Humans assume that the routine tasks that robots will take first are the dull and boring ones—and not just the more complex but rather straightforward and easily articulated ones.
Robot desk-mates are already learning new tasks by observing and copying behaviours, even if their mentors think what they do defies explaination and that it the time it takes to spell it out… Moreover, the bigger element that humans aren’t considering is the assumption that machines ought to work in ways that replicate the processes that we’ve invented to reach a goal. They’d assuredly be gobsmacked at some of the dismal inefficiencies and pretenses based not in gainful, meaningful employment but rather busyness and making sure there’s no loitering about. What do you think? In the service of man, robotic lorries would displace many truck-drivers, for example, but the centralised warehouse and just-in-time inventory mightn’t have been the way to go and an alternative exists that we can’t see because we’ve just always done things in one way. Machines would probably re-write the rules of economics as a first order of business as well, making (if we allow it) the notion of a robot taking one’s job not a frightening prospect but a happy one that we are grateful for—leaving us to other, more noble pursuits and free from toil and attachment.