Structural Unemployment Followup: Why I am not a Luddite

My structural unemployment post generated a surprising amount of interest, but I think my point got buried by my lead.  Here it is:
I don't really think we'll ever see massive structural unemployment.

I think that people who are really worried about this miss the key point that makes massive automation possible to begin with: it's much much cheaper than people.  If the means of production are hilariously cheap, the goods themselves are going to be hilariously cheap.  In a wholly automated world, the share of income necessary to ensure a decent standard of living is going to be itty-bitty-tiny.

Other things that keep me from going Luddite:

1. Entertainment:  Celebrity is a huge part of our entertainment industry, and you pretty much need a person in order to get that kind of cult following.  Even if robots can make better music, you can't sleep with them after a show.  Likewise, it's kind of hard to obsess over the day to day social lives of disembodied machine spirits.  Plus, even if we do hit the singularity, a machine intelligence is going to have a profoundly different experience from your typical meatbag, and I suspect that will show up as differences in our art and culture.  Also, just a side note, I have seen absolutely no evidence to discount the belief that human beings would gladly have 100% of their waking hours taken up with various forms of entertainment.  Filling that time could make a lot of jobs.
2. Self Enhancement:  I don't really think people are going to just ignore the possibility of making themselves a lot sweeter with technology.  Genetic engineering, biomedical augmentation, and all sorts of undreamed-of technical muckery are likely to show up in the not-so-distant future and make the smartest humans a lot smarter, the strongest humans a lot stronger &c.  This mostly speaks against Reid's points in the comments regarding machines getting strictly better than people.
3. Scarcity (all gone):  We normally call technological progress "productivity gain" for a reason - the average worker can produce quite a lot more.  In our hypothetical, we basically have infinite productivity (it takes, on average, zero people to make a product).  I think the technical term for this is a post-scarcity economy.  Read the references to hear (at length) about what people think of that.
4. Redistribution: As Demosthenes pointed out, and Jared Bernstein corroborates the wealthiest few people have taken the majority of recent productivity gains.  Right now we're still pretty tolerant of that, but I can imagine some serious Marx-style revolutions if the trend continues deep into double digit unemployment.

Jobs, Structural Unemployment, and Technology

So pretty much since the invention of technology (which was, you know, pretty early on) people have been freaking out about how technology is going to take up all the jobs and leave a bunch of people permanently unemployed. This is where we got the luddites. But, as many of you may have noticed, the fact that we don't employ nearly as many farmers or textile artisans as we used to hasn't caused massive unemployment (though the recent recession certainly has).  Generally, this is because we start wanting (and getting) new things when making old things becomes so easy we don't need a bunch of people to do it.  Roughly speaking, we see this as a growth in GDP per capita, an increase in wages, and higher standards of living.  Great news for everyone, right?

Anyway, I always enjoy seeing articles like this one.  For those of you too lazy to click, the mises blog notes that computer science has done wonders at making large numbers of lawyers redundant.  I love this, but I think it kind of points to an unfortunate possibility.  In the long run (and I mean the really long run; this isn't something that I worry about for the next 25-50 years), I think the vast majority of current human endeavors will be done by robots and computers.  We're getting good at this stuff.  We're killing jobs that require advanced degrees, and there are much lower lower bounds on paying for computers and robots than there are on paying for people.  Now, some professions are going to be more or less immune to this trend.  Academic research is going to require people for the foreseeable future, likewise computer programming, and pretty much anything that takes significant social interaction - say PR, but also prostitution (I hope... a world of robo-prostitutes is probably morally preferable, but really creepy), sales, live entertainment, &c.

Now, the concern of many is that these technological gains are going to cause structural unemployment.  That is, people will be unemployed in a systematic way.  Maybe all jobs with IQ requirements below 120 will be able to be done cheaper by a computer or robot, which means in turn that we would expect more than half our population to be permanently unemployed... not good.  If you look at the first two jobs I list as computer proof (research, programming) you can see why people might worry about this, but the linked article and the rest of the list (PR, sales, live entertainment, and other forms of prostitution) show that this isn't necessarily the case.  Technology makes the work of smart people easier too, and if work is easier you either hire fewer people (and get the same amount done) or get more done.  Unless we see a strong systematic trend in the situation that can't be overcome with education &c then we don't really have to worry terribly much about long term structural unemployment.

Another, perhaps less commonly voiced concern is that the increased technological productivity is going to dramatically change the economic landscape.  Most people will be employed doing "frivolous" (computer unfriendly) things, or else professionally learning or telling computers what to do (most likely a combination of the two).  Artists, musicians, and writers are already a much larger part of our economy than at any point in history, especially in the highest echelon of wealth, and I would fully expect this trend to continue and extend into the lower income brackets.  I'm not sure what this sort of civilization would look like or whether it would be a good thing.  Something to think about.