Two important things to note about the figures you quote

• The growth in labor force stopped in 2007 - this appears to be unprecedented in the data
• The labor force is measured in absolute numbers - a much more meaningful benchmark is the % participation rate

Most of the growth in participation rate in the last 60 years has come from bringing women into the workforce. That reversed to a massive extent, at least in the USA, in 2007, and hasn't recovered. See http://earlywarn.blogspot.com.br/2013/02/employmentpopulation-ratios.html for an interesting take - the whole blog is a valuable resource with some good thoughts on technological unemployment also.

Personally I am fairly convinced of the technological unemployment hypothesis, and think that even without strong AI there will be less and less things that only humans can do as time goes forward. I don't think this is necessarily a bad thing provided we can solve the massive inequality problems that are being caused by the increasing share of value of capital vs labor.

If we assume everyone is born (economically) equal (incorrect assumption, but a nice normative one at least. I'd be interested to hear the libertarian argument against us being born into equal circumstances/opportunities) all that each of us has to offer is our labor. We accrue capital from there - some more adeptly than others. If capital is more able to beget capital than labor is, then those that begin with a head start (ie own capital) will only see that head start increase, and those without capital will only fall further behind.

Redistributive economics is anathema to your average libertarian, but if my hunches about technology leading to increasing capital share of value is correct, what do we do about it?

My personal preference is a universal basic income. This would require high taxation, but I don't see how capitalism will survive without something like this to stimulate consumption.