Anecdotally, most of my extended family couldn't write a line of code if their lives depended on it; nor could they design a serious science experiment or write a novel. They aren't trained for those things, and that training has to start young or come with serious motivation, time, and effort. Not one of those people have any interest in learning those things.

I think it's possible to have a knowledge based economy, but we would have to drastically redesign our education systems and our culture in ways that I find it hard to even imagine how we would do so.

But, I think it's only a matter of time before we gain more skill in writing programs that automate out even the most complex knowledge based jobs. Watson is already showing the potential to be a better doctor than any human. The program named "Emily Howell" can already write symphonies that pass whatever the musical version of the Turing Test is. How long before we apply those things to all the other knowledge based jobs? What'll be left? More automation and maintenance programmers? How many of those will we even need?

If you accept that certain jobs[1] to be performed at a worthwhile level require a certain degree of general intelligence, or at the very least, the vast majority of people performing certain jobs have a measurable general intelligence above a certain level, and you accept that things like general intelligence, height, and other observable, measurable phenotypic traits of a species occur on a more or less normal('bell-curve') distribution, two things that are pretty much taken as a given outside the PC deluded, "I love me some outrage!," "Wow, just wow." crowd, then there is your answer. Is it not?

You might think you're doing good for the world by saying 'everyone is equally capable,' but you'll probably have more success in doing good for the world by understanding the world as it presents itself upon clear-eyed observation.

[1]"Software is an IQ business." -Bill Gates