Wednesday, December 26, 2012

Rise of the Machines

Interesting Krugman column here.

Since it is behind the paywall let me summarize: Krugman starts out with Bob Gordon's summary of the "industrial revolutions" of the last 300 years. In Gordon's (and Krugman's) view these were and are:

1 - (steam, railroads) from 1750 to 1830;
2 - (electricity, internal combustion engine, running water, indoor toilets, communications, entertainment, chemicals, petroleum) from 1870 to 1900; and
3 - (computers, the web, mobile phones) from 1960 to present.

Krugman then goes on to discuss his disagreement with Gordon that IR#3 - the digital revolution - has substantially run its course and will (or is already) yielding lesser and lesser economic gains. Krugman's primary issue with this is his conviction that the IT revolution has really not fulfilled its potential. He observes that
" turns out that there are other ways of producing very smart machines (and then details some of those ways) And this means that in a sense we are moving toward something like my intelligent-robots world; many, many tasks are becoming machine-friendly. This in turn means that Gordon is probably wrong about diminishing returns to technology."
I do not pretend to be especially tech-savvy so I am not sure of the actual practicalities of this. But I will defer to Krugman on the issue and assume that there is, in fact, a continued economic benefit to come. BUT...then he comes to what I consider the crux of the biscuit:
"Ah, you ask, but what about the people? Very good question. Smart machines may make higher GDP possible, but also reduce the demand for people — including smart people. So we could be looking at a society that grows ever richer, but in which all the gains in wealth accrue to whoever owns the robots.
And then eventually Skynet decides to kill us all, but that’s another story."
But I would go further than this.

General prosperity in such a society would actually depend on there being fewer people.

In such a society the remaining human work would fall on the top and bottom ends of the labor spectrum; the truly "creative" sorts of work - art, architecture, medicine, engineering - which require an inductive leap not programmable in a machine, and the physically-demanding-but-inductive sorts of work, like sorting trash from recyclables, performing bed-baths (and other medical dog-work), and waiting tables.

The latter sort would, as they always have been, be low-paid and low-esteemed. But that's the subject of another post.

But the point is that this world of Krugman's presumes that a huge volume of the work now done by humans - the "middle" sort, the kind requiring human muscles but not especially skilled human brains - will be done by machines.

But in that case...what the hell do you DO with those leftover humans, the ones too handless to be artists but too ambitious to be street-cleaners?

My point is and has been for some time that we in the First World are moving into an economy where we have too many people for too few jobs; that the sorts of massively labor-intensive industrial work of the 20th Century (riveting steamships, assembling buildings, constructing highways) have effectively vanished, replaced by the same work done largely by machinery.

But the human faculty for reproduction has not been similarly upgraded; we're still littering the Earth with our spawn. And, given that so many of said spawn will not or can not be trained to do those top-end creative jobs (high school graduation still lingers around the 70% range, tops, and college graduation is far lower), that many of these progeny look towards a fairly bleak employment future, even assuming that the most optimistic of Krugman's predictions comes true.
The bleak possibility that follows from this is the development of a sizeable group of unemployed and unemployable people who, at best, are pacified by televised entertainment and a dole but who, at worst, represent a constant threat to the society around them, a lumpenproletariat liable to be swayed by political rhetoric and shaped into a mob by unscrupulous politicians.

That, or exterminated by the machines.
But that is, indeed, another story.

Still, a rather dampening thought for a damp Boxing Day.


Ael said...


Standard economic drivel with so many implicit assumptions that it is useless for any purpose.

1) Ignores the internet revolution (which changes communication modes from broadcast to peer-to-peer) is only getting started.

2) Ignores 80% of the world, which is currently undergoing all 3 IRs at the same time. Also it ignores the imperial and colonial "shape" of the world economy even back before steam. (You are not an island).

3) Ignores environmental impacts of economic activity.

4) Assumes that they are measuring the right thing. I lead a life of incredible luxury compared to people 150 years ago (before electricity). However, I have to pick up my own socks and cook (at least some) of my own meals, which the rich folk of yore never did. Further my lifestyle isn't all that different from someone making half my income, or quadruple my income.

FDChief said...


1. "Just getting started"?? While I'd agree that ten years ago this might have been true, I don't think this is true today in the U.S.; much of what seems do-able with digital communications has been done. I can see speeds and volumes increasing, but the fundamental pattern of digital communications seems well-established by this time.

2. Immaterial; the issue is U.S. domestic economy. If anything the industrialization of the outside-the-U.S. world seems likely to make the sorts of work that the digital economy is destroying within the U.S. even less sustainable as a factor in the U.S. economy.

3. ??? Ummm...yeah, that's not the point of either the linked articles or my post.

4. Ummm...OK. Again, what does this mean? That the impact of the revolutions has been to raise the general economic level in the industrialized world? I'll buy that.

But my point in disagreeing with Krugman is that I think he, and the other "optimists" overestimate the benefits of the digital revolution. Rather, my concerns are that we may well have reached the inflection point of the digital revolution, and that not only will the wealth produced by increased digitization tend to go exclusively to the rentier and wealth-possessing class but that the implication of the internet revolution is the actual destruction of much of the economy which provided employment to what we think of as the "working class"; the people who got paid to do industrial labor, the rivetheads, the guys with the high school education working on the shop floors and the assembly lines.

And, frankly, I don't know where these guys are going to go. Call centers? Nursing homes? That sort of work has never and will never pay well enough to support a family at anything like a middle-class level.

The industrial West has purchased social peace by dragging as much of its population into the middle class as possible. IMO the big reasons we don't see Homestead Strikes and IWW-style industrial agitation is that the guys on the lines see themselves as "middle class" and middle-class people don't rock the boat, don't make trouble.

But...if a whole bunch of these guys are going to end up dropped back down to welfare level...

That can't be a good thing.

Lisa said...


This bifurcation is surely a concern. While the entrance to this stage may be staggered amongst nations, surely this seems the endpoint of capitalism -- an inverse bell curve.

I do wonder whether people will simply be enervated as they drop out of the lower middle class, everyone simply bleakly, meekly accepting the crumbs their overlords allow them. Accommodation and projection seem to be hardy human behaviors.

Barry said...

Ael: " Further my lifestyle isn't all that different from someone making half my income, or quadruple my income."

With quadruple your income, your retirement would undoubtedly be a *lot* nicer. Or these days, your retirement would actually be retirement, as opposed to being booted out of the paid labor force.

And that's just a start.

Barry said...

FDChief: " much of what seems do-able with digital communications has been done. I can see speeds and volumes increasing, but the fundamental pattern of digital communications seems well-established by this time."

First, speed, especially ubiquitous speed, has a quality all of its own, to paraphrase Stalin. For example remote driving of cars and remote operation of many things requires highly reliably high-speed bandwidth available every where.

And we've only had half-way decent high-speed cell networks for a few years in the USA. Go back a decade and the idea of being able to stream video on your phone as you are moving around the streets would only be found in 'Wired' magazine; now it's frequently possible.

We're still figuring out what to do with it.

Barry said...

What I find hilarious (in a sick sense) about the original article is that we're watching software move through the US economy on a year by year basis. Functions once thought unlike to be computerized a few years ago are now (e.g., law is getting hit by AI), and monitoring/control of work ramps up steeply.

Now, it's not unreasonable for an economist to be ignorant of that fact, since they are protected from the consequences.

Ael said...

1) Twitter, Facebook, inter-organization communications like e-Health. Global state "surveillance". Robot truck drivers. All large successful systems were small successful systems at one point. We are just getting going on the "bootstrap" cycle.

2) They can't cherry pick UK or later the USA and then make a statement about how economies work in general. Its like the blind guys only feeling the trunk of the elephant and thinking that they know what it is shaped like. Much of the resources and markets needed to fuel the high growth were external to what was being measured.

3) The globe is bumping up against limits to growth. The industrialization of Asia and climate change are probably going to be bigger factors in future growth.

4) My feet are dry, I'm warm in the winter and cool in the summer. I get to listen to the finest musicians in the world. I can read books from the best authors in the world. I don't ever worry about my next meal. My kids are getting a good education and I don't have to worry about losing my house if I or the wife gets sick (thanks to the govt of Canada). These things would not change if I made half my income. If I made quadruple my income, I could travel more often to Europe, or own a big fancy car. Better for sure, but that big an improvement over having your feet dry.
The point is that by measuring per capita growth you miss out on the quality of life improvement offered by the electrical revolution (i.e. Lights on at night, clean water, flush toilets) Cell phones and computer games don't even come close.

Swanditch said...

FDChief said...

"These things would not change if I made half my income. If I made quadruple my income, I could travel more often to Europe, or own a big fancy car. Better for sure, but that big an improvement over having your feet dry."

Oooh. NOW I'm envious.

Because if I made half my income I'd be within an ace of the federal poverty line.

THAT's the reality for many of us out here; we're half a year of no paychecks away from living under bridges and eating out of the dumpsters.

And that was my point; regardless of the economic niceties of the Gordon/Krugman posts I think it addresses something seriously troubling about the Rise of the Machines; that for a lot of us here in the US it implies the loss of our jobs and their replacement with...nothing.

To a certain extent I'm insulated because it would be hellaciously difficult for an AI to do my job; there's simply too much inductive reasoning and observational analysis involved. Sure, a pattern-recognition computer could be taught to identify landslide features on an aerial photo but would be hard put to do everything needed to perform a slope stability analysis such as actually logging the boreholes and performing the reading of the inclinometers.

But as Barry points out, even the professions are getting hit; digital commo has made it possible for law firms to outsource their work to legal secretaries in Mumbai and for hospitals to outsource theirs to radiologists in Guangzhou.

If anything I see increased complexity and speed as making this globalization less favorable to the high-wage nations rather than more.

As for the external resources and markets, and the environmental limits of growth, well, like the eventual self-awareness of Skynet, those really ARE another story and not the one I am concerned with here...