How will life be in the future? Will we still have the jobs we have today? Will relationships be strictly between humans or will it evolve to include robots?
We have another of those jeremiads proclaiming the end of all that is decent and holy as the robots gear up to steal all our jobs. The real lesson from this being that perhaps we shouldn’t get a professor of computer science to do our economics for us. For there is one point that is undoubtedly true here, the rise of the sex robots, and the rest of it, worrying about 50% unemployment rates, the death of the job and the terrible problem of a rise in human leisure is simply nonsense. And the reason it is so is because this extremely intelligent and highly expert man is talking off his own knowledge base. Something which is sadly true of all too many who wish to contribute to this debate about what is going to happen as and when robots and AI get rather better than they are today. There’s three different economic problems with the assumptions that are being made here.
However, let’s start with the one part that is obviously going to be true:
Would you bet against sex robots?
Well, no, I most certainly wouldn’t.
He said that virtually no human profession is totally immune: “Are you going to bet against sex robots? I would not.”
It’s not exactly a startling insight that human beings are pretty interested in sex. But a quick survey of the history of sex aids shows us that we’ve been trying to mechanise it for quite some time now.
Dildos in one form or another have been present in society throughout history. Artifacts from the Upper Paleolithic which have previously been described as batons were most likely used for sexual purposes. There appears to have been hesitation on the part of archaeologists to label these items as sex toys: as archaeologist Timothy Taylor put it, “Looking at the size, shape, and—some cases—explicit symbolism of the ice age batons, it seems disingenuous to avoid the most obvious and straightforward interpretation. But it has been avoided.”
The world’s oldest known dildo is a siltstone 20-centimeter phallus from the Upper Palaeolithic period 30,000 years ago that was found in Hohle Fels Cave near Ulm, Germany.
A quick glance in any modern sex shop will show that we’ve not lost any interest in this mechanisation and also that some people are really alarmingly inventive about it. So, sex robots, yes, they will obviously arrive. To the extent that sex shops don’t in fact already provide mechanised versions of whatever happens to titillate. But we’ve also got the beginnings of a solution to Baumol’s Cost Disease here: something greatly to be welcomed.
The cost disease is the observation that average wages are driven by average productivity across an economy. However, it’s a lot easier to automate, and thus raise the productivity of, manufacturing than it is of services. Thus services will become relatively more expensive compared to manufactures as the society gets richer and average wages rise. Because the labour embedded in the service gets more expensive as the price of all labour rises, but the amount of labour embedded in the manufacture falls as that productivity rises. The normal example used, as I’ve mentioned already today, is health care, which requires a certain irreducible minimum of hand holding and bed pan emptying while we’ve just about got rid of human labour in something like manufacturing another copy of a piece of music. But since we’re on the subject of sex here this is also obviously true of the commercial version of that. It’s a fairly standard part of historical stories that, say in the 1940s, ladies of negotiable virtue might operate for the price of a pair of stockings. That these now cost $3 any- and every-where does not mean that that is the price these days of negotiated virtue. Rumpy has risen in value against the price of stockings we might say.
And there’s research showing that this has pretty much been true throughout recorded history in fact. If the average day wage was, in Roman times, 2 sesterces and the negotiation of virtue was also 2 sesterces, we would then find that in more modern times when the average day wage is $50 then that negotiation is also $50. But clearly the price of bread, wheat, cars, telephones, any and everything manufactured, has fallen precipitously against either of those wages over that period of time.
While it’s obviously fun to use sex, that activity of great human interest, to illustrate this the basic economic point applies to any and every service. Baumol’s Cost Disease is a real and unwanted thing. And if the robots are going to come along and mechanise services then this is going to make us all vastly, hugely richer. Just as increased mechanisation in manufacturing has done.
The second economic mistake is this:
Vardi said there will always be some need for human work in the future, but robot replacements could drastically change the landscape, with no profession safe, and men and women equally affected.
“Can the global economy adapt to greater than 50 percent unemployment?” he asked.
There’s two different problems or mistakes here. The first is about rates of change. 50% of all jobs disappearing in only 35 years looks just terrible. But this is to ignore the rate at which jobs already disappear. If we look just at firings and bankruptcies then some 10% of all jobs disappear each year. And then they’re recreated again elsewhere of course: what we see as the unemployment rate is the difference between these two rates and any mismatch between them. If we add in quits, where people voluntarily leave one job to move to another then the total jobs churn is about 20% of all jobs each and every year. And each such job change does mean a little bit of technological movement. So, over the next 35 years we expect 700% of all jobs to disappear and reappear (obviously, some will be exactly the same, others will appear and flame out like mayflies) and so 50% of jobs being at threat hardly even registers upon our concern scale.
The other one is this:
But Prof Vardi is unconvinced that a workforce of humanlike robots will be good for mankind.
“A typical answer is that if machines will do all our work, we will be free to pursue leisure activities,” he said.
“I do not find this a promising future, as I do not find the prospect of leisure-only life appealing. That seems to me a dystopia. I believe that work is essential to human well-being.
“Humanity is about to face perhaps its greatest challenge ever, which is finding meaning in life after the end of ‘In the sweat of thy face shalt thou eat bread.
“We need to rise to the occasion and meet this challenge before human labor becomes obsolete.”
“What’s different this time is computer scientists are working on building machines that can out compete us in everything we can do. If machines can do all the work or even 50 per cent of the jobs that we used to do, what will people do?”
The answer is, “something else”. And we can show this by historical experience. Take someone from perhaps 1750 and try to show them the modern world. Just about none of us have a job by what he would consider “work”. In his time there was a very thin indeed smattering of churchmen, lawyers and merchants, and everyone else was either growing food or making something with their hands. We’ve 1% or so of the people on the land these days, around 10% in manufacturing (although many of those, like the HR department, would not be doing a “job” by the standards of yore) and everyone else other than the military would be regarded as not actually doing any work at all. Take, as an example, the diversity adviser and try to explain that to our 18th century visitor. Someone is paid a good living just to talk to people about how they should be nicer to each other? Well, yes, they are actually. Because we’ve mechanised the food, the manufacturing, and so people are now available to do something else. And they do find something else to do as well, because human needs and desires are unlimited and the resources available to satiate them are not.
And then we have our final economic error:
Moshe Vardi, a computer scientist at the Rice University in US, expects that within 30 years, machines and computers will be capable of doing almost any job that a human can.
“We are approaching a time when machines will be able to outperform humans at almost any task,” Mr Vardi said.
The correct answer to that being “So what?” Or in more detail, to misunderstand Ricardo on comparative advantage. Not that such misunderstanding is uncommon: Bob Solow provides the tagline for this section with his comment that economics is all obvious or trivial except for Ricardo on comparative advantage. The implications aren’t trivial and the non-obviousness is proven by how many get it wrong.
Here the point is that even if the robots are better than us at doing everything then this does not mean that we shall necessarily have nothing to do. Just as the more usual application of Ricardo points out that even if someone is better at doing everything than you are it still makes you both better off to trade with them. Thus it could be that the robots will be better at absolutely everything than we are. Yes, even at being diversity advisers or zumba teachers (they’re already better than a certain form of economic writing as the robot produced company results announcements on this very site show). Even if this is so it will still be to our advantage for us to do what we are least bad at, perhaps that diversity gobbledegook, perhaps the rather more respectable teaching of zumba. That’s the very point of comparative advantage after all.
There is though a horizon at which this breaks down. Which is when we have in fact solved, sated, met, every human want and or desire. That is, we’ve entirely solved the problem of scarcity. There’s not really a lot economics can say at this point because the subject is about solving the problem of scarcity. Once we’ve done that it rather becomes redundant. But then again, a society in which every human desire and want has been supplied and satiated doesn’t actually sound like a bad one really.
The problems with all this worrying about the robots coming to take all our jobs are really about people not understanding the economic points about employment, jobs and so on. The coming rate of change is, even at the most extreme predictions, very much less than generally happens in the economy right now and has been for many decades. That we mechanise the production of services will simply make us all richer as all previous rounds of mechanisation have done. And as long as there’s unmet human needs and desires then there will always be something for labour to do: and if we’re meeting all human wants and desires then what on earth are we worrying about?
As to the sex robots I’ve no doubt they will arrive but am deeply unsure about how popular they will prove to be. We’ve had skilled professionals in this field for many millennia now but the vast majority of humanity still prefers amateur enthusiasm….