You are being a Luddite. People will shift into cushy service sector jobs. Craft beer, massage parlors, landscape architects, mobile game programming, whatever wealthier societies might want more of.
That sounds fantastic, because service sector jobs tend to be local.
Massage parlors and landscaping -- pretty geographically limited, ie limited competition, and scale poorly, ie no large competitors with economies of scale.
Is that the overall trend? The things that wealthy societies want -- in the future, will they increasingly or decreasingly come from geographically limited services that scale poorly and are thus good candidates for employing large numbers of humans? Even if increasing wealth grows the pie, will enough of the things we want in the future be the kinds of things that lots of our individual neighbors can give us?
I wouldn't bet on it.
I mean, I think it'll all work out -- it always has -- but I could also see lot of things we take for granted, like high employment and industrialized education as sufficient to prepare people for work, revert back to the historical mean a bit.
I don't think one needs to be so pessimistic about it.
The historical trend is that as technology improvements have reduced the need for human labor in industries that used to be big employers, humans think up new and exciting ways to put the excess labor force to use.
When we look back on the era of cottage industry, we don't remember it as a golden age of high employment. We remember it as an era where the standard of living was lower because the huge amount of human labor that went into textiles meant that they were comparatively expensive. So expensive that after people were done paying to clothe themselves they had little money left for other consumer goods such as clocks, books, or glassware.
Similarly, we don't look back on the era of peasant agriculture as a time when every able-bodied person was virtually guaranteed a job in the thriving agricultural industry. We remember it as an era where most everyone was tied to the land and eked out their lives doing backbreaking labor.
As for whether the service economy can replace the manufacturing economy, I don't think that's a question we need to speculate on. Looking back on the past few decades of American economic development, it's clear that it already has. And that employment rates in the USA have remained relatively high throughout that process.
The argument for full-automation is and cannot be that simple. By just pretending that we just need more of something else you are over-going a lot of real world problems. Most of the people that are being put out of jobs now are not being given a reasonable transition route into those (promised) jobs. And they are also not equipped to take it by themselves, in a lot of circumstances.
If we truly want full-automation of boring and dangerous jobs to become a viable alternative we need a society that is ready for it. You need appropriate living and working conditions, even geographic-spread of those jobs across, education facilities, re-education facilities and so on.
The same as you don't just jump from Capitalist to post-scarcity in the blink of an eye.
Service sector jobs are not immune to automation. Eventually, inevitably, societies will put those who failed competition for highly creative jobs (something like 95% of population) on some form of welfare.
I don't think it's the most likely possibility, necessarily, but I'm totally open to the idea that near-full employment was a brief historical blip, caused by the momentary usefulness of humans as wet robots during the early days of automation, creating demand for them as 'employees.'
I sometimes suspect that both education and employment, at least as we know them now, are just blips caused by temporary conditions, and that may well revert to a less rosy historical mean. (I don't think that's a good thing, either -- history, uh, sucked).
Take education - what the heck chance does the current educational model stand of preparing people to be, say, a programmer?
It's not hard to learn programming. Not at all. And yet after 12 years of public school, 40-60% of the people who want to be programmers are weeded out of their very first year. That's a fairly damning indictment, given that programming is about the 'softest' engineering discipline out there.
But the truth is that most designers didn't start from zero in college, and most programmers the same. We came in knowing something, maybe knowing a whole bunch, maybe having done it since we were kids.
Industrial-age schools worked when it only needed to produce wet robots. What if that was a blip, and systems of apprenticeship, which are a natural fit for skills that may take several years of on-the-job experience to be really useful in, turn out to be the right model?
Well automated intensive agriculture can feed a country by using 1-2% of the workforce.
If we can get well-automated intensive consumer goods industry to provide basic necessities by using 3-4% of the workforce, and can finally make some progress in automating construction and other infrastructure, then I see no reason why all the "standard needs" can't be provided by ~10% of people, leaving the other 90% for these 'luxury' occupations. This can (and most likely will) happen after a few decades, still during my lifetime.
