« A month after Minsk II | Main | Isaac Asimov’s easiest prediction for 2014 »

March 11, 2015

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

OK, since I've been name-checked as a Luddite, let me propose a defense for part of the construction job category.

Basic construction jobs are probably automateable. Moreover, unlike other professions, liability issues will likely favor automated worksites. So in general, I suspect the construction sector will experience a slow employment decline between now and 2040.

But there are a lot of construction jobs that don't seem amenable to algorithmic replacement, for two reasons. (1) The people who do the jobs have trouble phrasing them as a series of if-them statements; and (2) the jobs are in fact not routine. Carpenters would be an example of (1); masons working with poured concrete would be an example of (2).

Now, I don't want to go to the mattresses on this point, but it would not surprise if there are still significant numbers of non-artisanal human carpenters and cement masons in the year 2100. (I deliberately picked that year, on the assumption that robotics and A.I. will have matured by then -- we'll be on the opposite side of the s-curve. My contention is that it will never happen; not just "not for 20 years." And I wrote "non-artisanal" because I think that even if my contention is wrong we can all agree that there will always be small numbers of human artisans as long as Skynet refrains from genocide.)

Am I wrong? Tell me if I am! I've got two kids to steer away from certain trades.

"taxi drivers and chauffeurs...truckers and delivery drivers"

I'll give you many of the taxi drivers and chauffeurs, although there will still need to be people like the SCAT drivers who help my mother in and out of the vehicles.

However, I have friends and family in the delivery business, and they do more than drive. They deal with customers, carry packages inside to the proper locations (which vary from building to building), handle paperwork, and sometimes sell.

"Once code is written, it doesn’t depreciate." One word: COBOL.

Capital supplanting labor...that I don't have a good answer for. On the ameliorative side though, all of those robots are going to require operators and IT people. Lots of them, and few of them will work for minimum wage.

"But there are a lot of construction jobs that don't seem amenable to algorithmic replacement, for two reasons. (1) The people who do the jobs have trouble phrasing them as a series of if-them statements; and (2) the jobs are in fact not routine. Carpenters would be an example of (1); masons working with poured concrete would be an example of (2)."

My understanding of construction, however limited, is that a lot of the deviations and on the ground creativity in the process is a byproduct if the disconnect between plans and reality. But much like GIS is allowing farmers to run tractors and know exactly where they are laying down seed, we're seeing the integration of computer modeling and GIS to reduce a lot of the errors or discontinuities in the process. I don't know what that means for employment, if you're more efficient at planning your construction process does that mean fewer people, because some are just an over-estimate of what you need so you can adapt to the situation?

The challenge seems that it's hard for us to envision what the pull is for all of these workers, both low skilled and skilled, in addition to the technology driven push out of employment in their existing sectors. The skills adjustment would be a challenge even if there is something for them to go into, they may not have the skills, may be challenged in retraining, etc. When is this a problem? Before 2050, but maybe not the 2020s.

A lot to reply to here.

@Noel:

You're not a luddite. You're a technophile of questionable tastes. ;)

The transition to the Robopocalypse has already started in construction. I mentioned some time ago the robo 'steam roller' I encountered here in California (and really wished I was able to get pictures of).

There will assuredly be some people left even in 2100 doing construction. However, if I may, some of the new machine algorithms out there allow for the machines to learn by watching rather than by having it explained. Right now, somewhat basic. In the future, say ten years, probably not as much.

If I may, how we build will like change, too. 3d printed houses are not here /yet/. However, in 5 years, they'll probably be ready for regulatory approval. Between that and the structurally insulated panels built at automated factories (roof, floors, interior walls), delivered by robo trucks and mounted by bots, you can bet lots will change.

@David:

Lots of that can be automated. Sales and paperwork? Software. Carrying packages? delivery to the right place? Atlas and/or drones.

The machines will need a lot less operators than you think. Consider. The supercomputers of the 1950s/1960s required veritable armies of people working on them. These days, we have about 10 at the day job for machines which are millions of times more complex. The jobs will be well paying, but there will be considerably less of them than there are jobs doing the same work now. Consider the burgertron, too.

@Logan.

The transition has started. I bet at the next business cycle (we slip into recession after the current growth period is done), we'll see it start being adopted wholesale. Why? Because the companies will be able to do the same work for cheaper and try to eek out a profit even during the downturn. The problem is when the business cycle turns. Where will the jobs be? After all, Mr Ford, who will buy them?

Carlos may have some thoughts here.

"Lots of that can be automated. Sales and paperwork? Software. Carrying packages? delivery to the right place? Atlas and/or drones."

I have a friend who drives for an auto parts warehouse. He loads his truck in the morning, helps car shop owners carry parts inside, handles paperwork and returns, and calls the home office when the customer has a problem, and does all this for a barely living wage.

Everything he does could be automated, but it would be complicated, there would be customer resistance, and I really don't see much cost savings. I think his job is safe for longer than he'll be working at it. Like say until 2030?

However, what I'm really interested in are your answers, not the size of the problem. (And the silly ones too - can always use a laugh.)

BTW, the captcha thing is giving me blanks when I try to comment.

"However, what I'm really interested in are your answers, not the size of the problem. (And the silly ones too - can always use a laugh.)"

We'll get there.

Between 2020 and 2030 is when a good chunk of the switch over will happen. It'll start before then with the next economic downturn.

For the captcha, right now you can click the doohickey with the spiral double arrows. I'll try to get the problem fixed.

For automation, I find that I'm more interested in the long view: what will survive past 2040? I want to know what the world looks like when the transition is over ... I'm interested but less so in whether that will happen by 2040 or 2080 or 2100.

Although the latter question does matter for my kids: the difference between 2040 and 2060 will be huge for them.

Maybe this will help:

http://www.voxeu.org/article/robots-productivity-and-jobs

Probably not as an inspector. At least for jet engines:

http://aviationweek.com/mro/watch-robot-automate-inspection-procedures-engine-components

http://www.vox.com/2015/5/21/8630771/software-slow-wage-growth

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Working...
Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.

Working...

Post a comment

Your Information

(Name and email address are required. Email address will not be displayed with the comment.)

Categories