In the 20th century, many things changed about how people allocated their money. One of the biggest was the movement towards investing capital in education for one's children.

A couple of aspects of this are interesting to note. First, such education is a form of inalienable property. One is unable to divest themselves of their education after they've obtained it. They can effectively rent out their education by becoming an employee, but they can't sell it. You could imagine someone attempting to sell the right to their lifetime earnings in exchange for a lump sum, but this would most likely be unconstitutional, as it comes too close to indentured servitude.

Since there is no market for buying and selling the educations that people have obtained, it seems likely that there are some large inefficiencies in this area. There are likely to be many people either over- or under-educated for their jobs, or people unable to properly utilize their education, while there may be many others who would like to put someone else's skills or education to work, but are unable to do so for a variety of reasons. Of course, an even simpler place to see inefficiencies is in the decisions of people as to the subjects and extent of their studies before they embark upon them -- it's clear that we can't properly predict how to make those choices, because we can't properly predict the need for skills by future employers.

In recent years, we've seen software systems slowly encroaching on various low-skill industries, and more recently, on higher-skill industries. Many have noted that some people will be displaced in this process. But I haven't heard anyone talk about the theory I present here.

What will be the effect of this use of skilled software? I think that possibly, it will be to begin to detach education from people. What I mean by that is that as software becomes more and more sophisticated and capable, it may be possible that it could serve as a replacement for human education to a large extent.

The principal advantage of this is that whereas there is no market for received education, there is a market for software. Not only a regular market, but a market in which transactions can often be performed online, further reducing transaction costs even below the level of real-world markets in goods. This could allow businesses to avoid the great inherent inefficiencies of the market in employees (i.e. education rental), and effectively buy and sell skills in the form of software, to the exact extent required.

This change may allow many people to avoid the extensive education we undergo now, and instead operate software. This makes people below the skill level of software mostly fungible, and mostly obviates the need for education below that level.

What sorts of social changes could this lead to? I think it's clear that this could create a much greater divergence in education and wealth between those below the skill level of software, and those above it. Let's call this line the software line. Those who thought they could hack the education required to get above the software line would work hard to get it, and those who thought they couldn't would avoid any excess education.

I expect that this could lead to a discontinuity in the continuum of wealth and education across our society. If those below the software line aren't found to be completely useless, they will still gain little benefit from rising above their peers. On the other hand, those above the line will feel the same incentives to improve their education as all of us do now. In the end, we may return to a completely bifurcated society of upper and lower classes with a complete inability to understand and empathize with each other...