Are a few years of recession really sufficient reason to begin to believe that human labor has become – and will indefinitely remain – unprecedentedly inflexible? Are we to think that entrepreneurs have finally exhausted their capacity to creatively figure out ways to profitably employ the vast majority of adult human beings willing to work?
1. Activities wherein low-skilled workers can add value likely suffer from Cowen's low-hanging fruit problem. The easy things to find that low-skill workers can do have been found. And...we're using most of the capacity of those workers for available wages now. It may be that the number of people with zero skill that can be paid $20/hr ($10 wages, and $10 regulatory costs) is approaching saturation, given current realities. While the value of a worker keeps increasing, the time it takes for a low-/no- skill worker to move from $20/hr value to $21/hr value may be far higher than the time it took for the worker to move from $19/hr value to $20/hr value.
2. Innovation is costly. It may be that the risk of creating new categories of jobs in a bad economy is too high to even try the new paths, but that one the economy recovers, the risk inherent in building new types of jobs decreases sufficiently that new categories of work for low-skill workers will be created.
3. Regulatory cost is monotonically increasing. It could be that the costs of regulation are increasing faster than the value of low-skill workers is increasing. The functions could well have crossed in the last 5 years.
4. Up until 50 years ago, there were NO substitute goods for labor. Just recently, with the advent of the computer, instant global communication and global supply chains...the number of substitute goods for labor has increased drastically. It may be that for now, there is a glut of surplus labor-substitutes, that are MUCH cheaper than labor, and no one has found the $20 cases where local labor is semi-required yet.
5. Same scenario as #4...but labor is infinitely (or near-) substitutable. Effectively, computers/robots can now (or soon) do EVERYTHING a low-skill worker used to be able to do. The infinite supply of new jobs is contingent on there being NO substitutes for labor. Now that there are substitutes for labor, we should no longer expect an infinite use for labor, any more than we should expect an infinite use for whale oil (or an infinite use for petroleum, after the solar price point in ~2030).
6. Of course, the real choice is some combination of factors 1-5. I'd bet that suspending a pile of laws (Health care laws, Minimum wage laws, welfare laws, sue-the-boss laws) could decrease the cost of employing a zero skill worker from $20/hr down to $5/hr. I'd bet that a good economy, and decreased licensing/patent regulation could increase the cost/benefit/risk ratio around trying new approaches. Certainly banning computers/robots, immigration, and/or foreign trade could make the employability of many folks increase...of course the cure costs more than the disease in all 3 cases, so that sucks.
7. We should expect that #6, especially computer/robot #6 wins massively over time...and so even if today, there's no ZMP, we have no reason to believe that the future holds no ZMP...indeed, we have reason to suspect that ALL of us become ZMP workers sometime in the next 50-200 years...roughly as soon as desktop computers cross our particular IQ/EQ threshold.
8. This shouldn't worry us as much as it might, because one of the major defining limiters on wealth is that unlike most other natural resources, energy is presently expensive. On the other hand, as suggested before...inside of roughly 20 years (depending other factors), we expect that solar crosses the petroleum price point...and within another 20 years, energy in general starts costing what light costs us today...almost nothing. Free energy is probably as big of a difference as the other 4 big singularity-builders: AI/Cyborgization/Nanotech/Genetic Engineering.
Posts
7 comments:
First clarification:
The debate needs distinguish between ZMP in wealth vs. ZMP in dollars.
If the problem is ZMP in dollars, the problem is regulation, end of story.
Second clarification follows:
If we're talking about ZMP in wealth, then there's important second-order effects.
The point, the goal, of technological advancement, which everyone seems to forget, is for everyone to be a Patrician and not have to work at all.
Machine intelligence replacing low-skill labour basically means that even the poor will be able to afford maids and gardeners.
It's not a problem, it's mission successful. Ideal becomes reality.
For example, basically-free labour means food is basically free which means that even tiny wages are sufficient to support a human being.
And, I just realized, even those tiny wages can come from the paupers employing a robot to earn it for them.
Alrenous,
ZMP = Zero Marginal Product. The amount of real wealth that a person can add to a business.
ZNMP = Zero Net Marginal Product. The amoutn of real wealth that a person can add to a business MINUS the real total cost of employing them.
-------
I agree that the goal is for everyone to be a patrician. Well-reminded.
-------
If labor is mostly free -- good open source artificial stupids, and homebrew 3D printers to make robots -- then it becomes very difficult to determine what the average (low-skill) person can trade (to others) to create net value.
I think your tiny wages line is a bit over-facile.
OTOH, I think it doesn't matter. We'll shift to Star-Trek Socialism, or something approaching it...with guaranteed minimum incomes in real wealth.
What makes a business (un)profitable, and what on net destroys actual stuff, are different.
Recently, I've come to realize they're not only different, but wildly out of whack.
The debate brought up unemployment. Are they unemployed because they can't create net value, or because something is wrong with prices? 9/10 I'd bet on prices.
IF prices can be verified to be non-stupid, I'll start to consider that maybe we have a labour issue.
-----
It's facile, yeah. I'm trying to wave in the direction that changing this one thing is likely to drastically change many other things.
The problems are therefore not going to be readily predictable. However, it appears very unlikely to me that creating vast wealth is going to make much of anyone worse off, at equilibrium. I don't know how it'll work, but on average it'll work out.
even the poor will be able to afford maids and gardeners.
I saw a claim that the average person today has machines to perform the equivalent labour of 28 pre-tech servants.
It makes sense: I wake up in the morning when the alarm goes off (in the middle ages: because a servant who has been turning an hour glass to keep track of time shakes me awake)
I turn on a light (servant has been out gathering wood, lays and lights a fire, and keeps it burning brightly for as long as I want, then puts it out and readies it for next time.) Grab a cold yoghurt from the fridge (how many slaves are needed to bring ice from the frozen mountains?) and have a nice warm shower (more wood gathering, water heating, water pouring, disposal of the used water).
Then I go out to my car (horse and carriage that has been prepared and readied to go), hop in, and set off. Meanwhile it has a radio (a set of the finest musicians in the world are just waiting until, with a wave of my hand, they are told to play. AND a herald with the latest news, brought by runners, who is also waiting should I desire him to talk.)
etc. and so forth.
it becomes very difficult to determine what the average (low-skill) person can trade (to others) to create net value.
I can think of one thing: Status.
Having 20 human servants would (I assume) provide a sense of status that the robot equivalent of 200 servants would not.
At least until robots become indistinguishable from humans.
Dr. Pat,
Very insightful. thanks.
Post a Comment