For an illuminating perspective about different paths automation may take, which also uses history as a guide, see this: http://www.economist.com/news/leaders/2

For an illuminating perspective about different paths automation may take, which also uses history as a guide, see this: http://www.economist.com/news/leaders/21701119-what-history-tells-us-about-future-artificial-intelligenceand-how-society-should http://www.economist.com/news/leaders/21701119-what-history-... My key takeaway is that, often automation can drive prices down, which increases demand, which in turn increases employment. This counterbalancing effect can eventually increase total employment, although the new jobs may go to different people. AI replacing services is a bit different than robots replacing workers in manufacturing, in my view, because often the entire job is not replaced. It can have the effect of allowing the human to focus on (currently) non-automated portions of the job. Accounting is ripe for automation, for example, but I predict it will be a good thing for most accountants. Well, predictions are dangerous, but, there, I've said it. As far as manual labor is concerned, it seems to me that ship has sailed.