Hacker News new | ask | show | jobs
by mschuster91 3911 days ago
This might be true for high-skilled jobs, yes.

But the vast armies of uneducated, second-or-third-generation-jobless people will have their jobs slaughtered by automation. Truck/cab/freight delivery will be automated, freight handling and shipping will be automated, cleaning of public and private areas will be automated.

The only field I can see which will depend on low-skilled worker armies is construction.

And with the refugee crisis we just magnify the problem because the refugees will have to be integrated, they have to learn German and still lots of them will be lower educated than the rest :/

6 comments

"The only field I can see which will depend on low-skilled worker armies is construction."

The trend in construction for the past decade has been towards less labour intensity.

So no good news here either.

Construction is already beyond the big wins in labor force though. Construction sites used to have armies of workers. Hundreds. Now they have perhaps a dozen for large skyscrapers, 5 for normal skyscrapers and one or two for houses.
On which construction sites have you been recently?

A year ago when I was working in construction, I remember for a standard home at least:

Planning (architects), demolition workers, excavation, foundation builders, bricklayers/concrete builders, plumbers, electricians, woodworkers, window workers, fireplace/chimney builders, painters, inner decoration guys, telecom/smarthome engineers. On one single home site. Oh, and add for at least the "big and dirty" parts of building a house at least two or three trucks with drivers.

But, of course, the companies did rarely work at the same time.

I see you point that the amount of labour involved in building projects has decreased however I disagree that no further labour decreases are possible. A typical house in the UK is currently build from bricks and wood with a tiled/slate roof and a concrete subfloor. The interior is made to look nice by application of plaster. A typical block of flats has a metal frame but the walls are still brick/breeze block (in the US cinder block I think). This requires several skilled workers commanding fairly significant wages (bricklayer, carpenter, roofer, plasterer) in addition to the plumber and electrician.

If houses were to move to a prefab model with e.g. plastic construction and standardised room sizes and plumbing and electrical connections between the rooms most of these skilled workers would find themselves redundant...

Fortunately the UK insurance industry inadvertently protects the building industry by making it difficult to insure so called 'non-traditional' constructions.

> If houses were to move to a prefab model with e.g. plastic construction and standardised room sizes and plumbing and electrical connections between the rooms most of these skilled workers would find themselves redundant...

Still, the prefab blocks have to be built by adequately skilled workers - gas, water and electricity at least will always require a certified worker simply due to the inherent risks of fire, explosions, water leaks and electric shocks.

This is true but a lot of labour intensive processes can be done off-site in a more controlled environment.

For instance, partial pre-assembly of mechanical/electrical fittings through to modular kitchen/bathrooms with almost everything done before it gets to site.

Of course you'll always need people to put all of these elements together but the numbers needed becomes less and less.

The certified worker who signs everything off used to do almost everything himself, we are now moving to a situation where this is almost all he does.

You are missing service industries in general and elder care (poignantly relevant with the article's emphasis on demographic shift) specifically. A large older population with a smaller worker pool to draw on will drive up demand for caregivers, exhausting the rapidly devaluing savings of the elder generations.
Most people would prefer being wiped in the ass by a robot.

There are really strong incentives for automation in healthcare.

When talking about the immigrants and refugees working with elders; Unfortunately, even the "human touch" and the social part of the job is sometimes problematic because of cultural and language barriers... (Also applies to the opposite situation, elder immigrants receiving care.)

I didn't get why the savings are supposed to devalue. The elder care costs for the super-rich is a drop in the bucket for them, and they can and will still use the rest of their wealth to pay poorer people (and robots) to generate them more wealth.
Truck drivers are in short supply. There is also a trend in mainand Europe to schedule more freight into the hinterland from ports on trains and waterways - so called Synchromodal Transport [1]

The new APM automated continer port in Rotterdam [2] is set to handle 4m containers per year using automated vehicles and cranes - with few operators on the dockside (just for the two train cranes and one barge crane) - the other 8 cranes are operated by remote control.

[1] http://www.dinalog.nl/en/themes/synchromodal_transport/

[2] https://www.youtube.com/watch?v=cHVF6xEjArA

Luckily they're on it :

http://www.dailymail.co.uk/news/article-3258829/Driverless-t...

Everybody's missing the really bad thing about these trucks and self-driving cars though. Everything is patent encumbered. So they will transform a bottom up industry (dependant on people to drive the trucks, trucks which can be produced by any of a 1000 companies) into a top-down patent-encumbered capital-intensive industry (because only one company will get self-driving cars thanks to our patent system). There will be one company driving these trucks, not many like today.

This will destroy truck drivers, yes, absolutely. However something it will also do is sink every KMO that has trucks driving, which is a a lot of them here in Germany.

So give it another 10 years and "oh no hell no to self driving trucks/cars" will be one of the few things socialists and capitalists/liberalists agree on, because it will destroy the constituency of both parties. I also seriously doubt they can reliably deliver cargo in the central streets of Germany's cities. The ones built in the 14th century which are about wide enough for a single horse carriage (very tight for a car, and truck drivers actually manage to squeeze in (you -often- have to forgive some damage of course))

> I also seriously doubt they can reliably deliver cargo in the central streets of Germany's cities.

I'm always pretty curious about why people, even on HN, seem to equate "this needs high precision" with "automated systems won't cope" when it comes to driverless cars. Saying something is really hard for a person because it requires careful, precise control sounds like exactly the sort of thing you want machines for.

So what is it about the task that is complex for a machine to solve?

Even in your patent-encumbered worst-case scenario I don't think DHL et al will have much problem paying licensing fees.

Urban transport will not be HGV. City limits trans-shipment points will split packages into inner-city deliverables.

Last mile is an active area of research investment e.g. electric cargo bikes [1]

The buzz phrase in the literature is massification to atomisation [2]

[1] https://k4rgo.files.wordpress.com/2013/12/k4rgo-ups-germany....

[2] https://people.hofstra.edu/geotrans/eng/ch1en/conc1en/atomiz...

> The only field I can see which will depend on low-skilled worker armies is construction.

I don't think that the construction industry will be immune to automation. Google up "house build robot" or "construction robotics" to see what's already being done.

As automation continues to improve, we see that fewer and fewer people are sufficiently skilled that they cannot be replaced at least in part by technology. Even tasks demanding high cognitive abilities gradually shift into the domain of what advanced machines can do. I wonder what kind of leverage average software engineers will have when machines can write better code than an average programmer? Or allow one developer to do the work currently requiring a hundred.

And how can there be a labor scarcity when productivity has risen to levels where we can overproduce almost everything we can think of, far outstripping our natural resources?

A lot of our inability to automate comes down to not being able to operate at a big enough scope. For example, all commuter transportation problems are ultimately caused by having too many people, too distant from one another. But we can't simply move them closer together for a variety of reasons, and so we have to think about trip planning instead. But we could also reduce the need to make trips and reduce the average length of each trip.

The future is going to be full of complicated thoughts like this - problems to which automation can solve some things but policy and design solves others. We'll have to be really good philosophers to avoid catastrophic error.

> And with the refugee crisis we just magnify the problem because the refugees will have to be integrated, they have to learn German and still lots of them will be lower educated than the rest :/

The EU already has 23 million unemployed (and a lot more in minimum-wage bullshit jobs nobody wants or needs), a million refugees won't make problems significantly worse than they already are.

> a million refugees won't make problems significantly worse than they already are.

Oh yes, they already do. The housing market in German cities is already fucked up beyond limits because our politicians have slept for years, and 1.5m refugees now needing winter-resistant housing (so, no tents) bring the market to its knees.

But I'm thankful for the refugees, because they show our politicians they can't sleep forever.

That's a different problem from employment, though – but yes, it's nice to see that, after 10 years in office, Merkel actually has to do something for a change.
<3