Problems with Prognostication, with a Meditation on Bifurcated Workforces

They do it every month. Over at Cato Unbound, they are once again having a discussion. This time it's about the future, and specifically the rich/poor divide involving the bifurcation of jobs into manual versus intellectual labors. They point out that the payoff for education and comparative advantage in the marketplace is accelerating at a pace that seems to foreshadow the obsolescence of manual labor and non-creative jobs, especially in America, in the near future.

Of course, the questions they ask manage to combine the dual questions of mechanization and outsourcing, which are really two separate questions. The first is a technological, predictive question, whereas the second is a policy and economics question. The point is that there are two separate groups that should address these problems, and in neither case is the prognostication particularly famed for it's accuracy. Why then should a group of people who are all well informed on the subject has definitive opinions on the subject? Because their published works make statements that have implications on these questions, that's why!

The next part of their question, of course, gives away the narrow perspective that leads to this grouping; they ask "what is America's comparative advantage?" It's funny, because it's the single most incongruous line in the article. First of all, technological change will obsolete things in transnational ways, so that comparative advantage is a silly mode of analysis. Secondly, in 18 years, when these newborn children are starting college and picking careers, what are the odds that the world will be in similar enough shape for this question to make much sense? (Remember now, 18 years ago, George Bush was president of the United States, The world was reeling from a war between a major power and Afghanistan where military might proved ineffective against the Muslim fighters, America was being challenged by the other world powers, and no-one knew when it would end. I was a bit young to have noticed at the time but I guess things haven't changed much. Except that now we can read about all of it online.)

Essentially, to respond to the question (and not to the original post, which was not too bad,) the game changes fast enough that the disjoint timelines that we need to use separate paradigms to analyze the two questions. Since I don't really care about American problems with outsourcing enough to write about them, and I'm convinced enough by Kurzweil's ideas that I think that considering geopolitical trends involving jobs 18 years from now isn't useful. So I guess I'll just sit on the sidelines and make fun of the posts in the thread.