OK, if you all insist I'll break my New Year's resolution one more time since I already have today.
When I first saw this post by Russ Roberts about an Econtalk with Robin Hanson about "technological singularities" I was very interested in it and almost posted on it. I've heard Robin Hanson lecture on this before. The timing is always speculative, but the logic of the argument - particularly when you look across the length of human history - is persuasive.
I was tempted to write a very quick post in response, simply noting that if the technological singularity Hanson describes here ever happens it would probably be sufficient to turn me from a market Keynesian to a socialist.
I decided the admission would haunt me for the rest of my blogging career so I didn't post it, but what the hell.
I post it now because of this post by Nick Rowe fleshing out the implications of the technological singularity in more detail, and this post by Karl Smith. "Overproductionism" inevitably leads people to socialism for good ethical reasons - if providing for demand can be easily achieved without contracting for the services of a large portion of the population, you're going to have extreme inequality and suffering not as a result of differential compensation from differential marginal products, but simply because the labor content of output is negligble. The problem with this line of reasoning has always been that our demand always outpaces technological unemployment (this is why concerns about underconsumption are much more defensible than concerns about overproduction). If there is a paradigm shifting technological singularity that makes the overproductionist fallacy no longer fallacious, then we're going to have to rethink a lot about the way society was organized.
Yes, this is very Marxist. What of it?
The point is, though, it's all predicated on the assumption that everything we know to be true about the economy is no longer true as a result of this sort of Hansonian "technological singularity". That assumes the singularity even happens, and it assumes it happens in the asymptotic, extreme way that it's being discussed by Hanson, Rowe, and Smith.
I think it would be a good thing, if we could manage the transition. Whether it's likely to even happen is a different story. I'm personally inclined to bet against Marx and crude overproductionism, but time will tell.
I think about these issues a lot... I think about the social implications (e.g. - "this would make me a socialist") less thoroughly than I think about the intellectual history of these overproductionist ideas. So thoughts on my thoughts on the social implications would be appreciated.
Have you read Player Piano?
ReplyDeletePersonally, I always have a hard time with the whole idea that massively efficient production is going to impoverish the working class. I can see that inequality might grow. But certainly the cost of provisioning a basic level of sustenance would be ludicrously cheap.
And what of the idea of a service economy? If the production of physical things becomes totally automated, so that product designers, manufacturing engineers (and no low-skilled workers) and the like are all that's needed to make stuff that people want, doesn't that leave a whole bunch of cheap labor available for services that are harder to automate? I don't know that I like the prospect of moving toward a society where the wealthy hire the less wealthy as servants, but that is still distinct from the massive impoverishment of the working class. I'm always just a little surprised to realize that many people can, in our current economy, earn comfortable livings maintaining yards and pools. That doesn't seem to be any more demeaning than assembling cars or electronics.
Definitely Robert - I don't see this as likely at all. Even if there is a technological singularity it is doubtful it will furnish the preconditions that would render human labor unvalued. Overproductionism is utopianism or dystopianism depending on who is selling it, but that doesn't change the fact that utopian or dystopian stories aren't a good read :)
ReplyDeleteThe service economy is an excellent point too. And we don't even have to think explicitly in terms of "servants". There are lots of instances where I'd imagine we'd prefer interactions with humans to interactions with machines.
It's also a good point that even if inequality becomes unprecedented, that doesn't mean the poor will be living badly. Is that the only thing that matters? Obviously not - otherwise the welfare state would not emerge with the modern industrial economy. But it does have some import.
I am with Keynes insofar as I think if something like this occurs it will be because we don't want to work anymore. We will "content ourselves out of work", we won't "invent ourselves out of work". But even that, I think, is highly unlikely.
One point, though - the reason why the pool cleaner can live comfortably is because (1.) the market economy produces amazing prosperity, and (2.) the pool cleaner has a job. If, after the technological singularity, the poor do not own any of the robots all that market-driven prosperity will be of little use to them. That's why I am tempted to say I would be a socialist - not out of hostility to the market economy, but out of a realization that if human labor ever becomes obsolete the only thing that can ensure even a minimal standard of living is the ownership of the means of production by the proletariat - the ownership of the robots, in other words.
And, as long as we're speculating so boldly about technological advances that might one day happen, why not assume that the non-labor inputs will also collapse in value as productive methods become infinitely more efficient? If I can make enormous quantities of energy from a handful of matter, and if I can then use that energy to freely reorganize other matter into forms and patterns that I desire, then what resources are scarce? Who is rich and who is poor?
ReplyDeleteI believe you have just refuted Nick Rowe's call to own land... you should click through the link and post that sentiment there.
ReplyDeleteI think such ideas of technological singularity are typical in stories of Isaac Asimov.
ReplyDeleteIn his Aurora planet in the story of...The Naked Sun, I think...people had machines taking care of every one of their needs and they had become spoilt and negligent. They also became individualistic and did not need other people anymore. To drive this point home, Asimov demonstrated that one Auroran woman did not mind walking around naked in front of a dressed man (the Swedes are already there).
Of course, Asimov had a limited understanding of economics and a limited understanding of something more important - human nature. The law of diminishing returns obviously hurts the capacity to make such "scarcity-eliminating" inventions which would be too expensive to make in the first place. And then human nature - people often just like working and often even like physical work (eg.agrarian philosophers). And people often just settle for less. Progress for the sake of progress does not happen, else we would have been using robots to scrub and bathe us.
True Prateek, which is the source of my skepticism. But you're really begging the question aren't you? A technological singularity is by definition some sort of unforseen repeal of the law of diminishing returns. You can't cite the law to refute its own repeal. And we've seen temporary instances of increasing returns to scale - we just know they peter out. I consider it highly unlikely, but not impossible. And of course it would be a singularity with respect to human demand. If a giant elder god that eats planets floated by and gobbled up the earth it clearly wouldn't be a singularity from his perspective (that, by the way, is the very definition of "capital consumption").
ReplyDeleteI like your point that people just like work too. This is a point that Keynes made in the general theory (actually I believe in the passage where he talks about pyramid building). I think that shows a maturation between his publication of Economic Possibilities for Our Grandchildren and the General Theory.
1) Marx was not an over-productionist; and
ReplyDelete2) Over-production is possible.
Gene - I am not an expert on Marx, so I will take that under advisement but with skepticism. Got any more info?
ReplyDeleteOn number 2 - my point is persistent overproduction is something I seriously doubt (although I wouldn't rule it out as impossible if this technological singularity pans out). Certainly short term over-productionism is possible. I would never challenge that idea.
I need to review your discussion in more detail again - but from what I remember from Jonathan Catalan's discussion of it (let me know if you need me to track down his link), I agree with him you're talking about malinvestment not really overproduction. You seem to be refering to inventory build-ups due to mistaken calculations. That's possible - I would never challenge you on that - but it seems relatively uninteresting. More interesting to me is the traditional "technological unemployment" version of overproductionism. Now, as I've stated above this has often been overstated by its proponents. Technological unemployment is real, but it is always temporary because with productivity comes higher real income and with higher real income comes higher demand which puts people back to work. A temporary dislocation worth remarking on, but not a doomsday scenario.
A permanent problem with overproductionism would loom with Robin Hanson's example (if it were to ever happen) and that is very interesting to speculate on.
Thanks for coming to he blog!