A Study Finds Half of Jobs Are Vulnerable To Automation (economist.com)

The Economist reports of a new working paper by the Organization for Economic Co-operation and Development (OECD) that assesses the automatability of each task within a given job, based on a survey of skills in 2015. "Overall, the study finds that 14% of jobs across 32 countries are highly vulnerable, defined as having at least a 70% chance of automation," reports Economist. "A further 32% were slightly less imperiled, with a probability between 50% and 70%. At current employment rates, that puts 210 million jobs at risk across the 32 countries in the study." From the report: The pain will not be shared evenly. The study finds large variation across countries: jobs in Slovakia are twice as vulnerable as those in Norway. In general, workers in rich countries appear less at risk than those in middle-income ones. But wide gaps exist even between countries of similar wealth. Differences in organizational structure and industry mix both play a role, but the former matters more. In South Korea, for example, 30% of jobs are in manufacturing, compared with 22% in Canada. Nonetheless, on average, Korean jobs are harder to automate than Canadian ones are. This may be because Korean employers have found better ways to combine, in the same job, and without reducing productivity, both routine tasks and social and creative ones, which computers or robots cannot do. A gloomier explanation would be "survivor bias": the jobs that remain in Korea appear harder to automate only because Korean firms have already handed most of the easily automatable jobs to machines.

Comments

Hmm

[cascadingstylesheet]

This time it's half? Last week it was something else.

Nobody really knows.

Re:Hmm

[geekmux]

This time it's half? Last week it was something else.

Nobody really knows.

Actually, history has shown us the one thing we humans do know when it comes to predicting the future; we can underestimate the shit out of damn near anything.

gloomier?

[Wycliffe]

How is that gloomier? If Korea has already managed to automate away most of the jobs than can be automated away and they don't already have mass unemployment then that should be a positive sign that other countries can do the same.

Re:gloomier?

[ShanghaiBill]

If Korea has already managed to automate away most of the jobs than can be automated away and they don't already have mass unemployment then that should be a positive sign that other countries can do the same.

We should also look at countries that have almost no automation. Some examples are Ethiopia, Afghanistan, Mozambique, Guinea, Somalia, and Mali. If automation leads to poverty then by avoiding it, these countries should be doing GREAT! Are they? I haven't checked.

Re:gloomier?

[Plus1Entropy]

We should also look at countries that have almost no automation. Some examples are Ethiopia, Afghanistan, Mozambique, Guinea, Somalia, and Mali. If automation leads to poverty then by avoiding it, these countries should be doing GREAT! Are they? I haven't checked.

This is like saying there can never be too much oxygen in the atmosphere, because people with too little oxygen suffocate.

Pretty much everything in the Universe is non-linear. Almost anything that appears linear is only within a particular domain. Outside that domain, things become non-linear again. Clearly, if every job is automated, then everyone will be unemployed by definition. However, as you point out, there have been huge economic benefits to automation.

It's a fundamental law of mathematics that any function which is increasing at one point and decreasing at another contains at least one extrema in between.

Haven't you ever checked the oil in your car? You do know there is a minimum AND maximum marker on the dip stick right?

Re:gloomier?

[ranton]

The problem is when 20% of the population still has to work long hours while the other 80% have nothing to do. Currently we distribute wealth roughly based on the amount of work you do (or capital you control that does the work for you). That is obviously not going to work very well when only 20% of the population is working and only 1% controls the capital.

I don't agree it is obvious that what you describe wouldn't work. You are essentially describing a society where 1% are wealthy, 20% are synonymous with our current upper middle class, and 80% are working class / poor. That isn't much different than today. The only significant difference is that the 80% wouldn't have jobs at all, but would instead be part of more substantial safety net programs. Or more likely most would also be working in the part time gig economy for supplemental income.

I am currently part of the upper middle class, making over 4 times the median household income in the US. If I had the option for my wife and I to stop working but have our income cut by 75%, we would both choose to keep working. I doubt I am alone in this. In fact I would bet this is the majority opinion.

If the 20% of working adults were making a median household income of $200k, and the 80% of unemployed / underemployed adults had a median household income of $50k, there would likely be plenty of incentive for the 20% of economically useful adults to keep working.

Not surprising

[Trailer Trash]

100 years ago 95% of the US labor force was in the agricultural sector. Now, it's just a few percent. We don't have 90%+ unemployment, though, because now we have jobs that we didn't even know existed 100 years ago.

Hopefully most of all current jobs can be automated so we can find new things for people to do.

Re:Not surprising

[apoc.famine]

I really don't understand how people can continue to make this argument. It does not appear that this time is going to be like the previous times. And it ignores the massive social upheaval during the switch that happens every time as well.

In the past, when we've automated low skill jobs, we've pushed people into higher skill jobs in the process. Farmers ended up working in factories. Factory workers ended up working in offices.

But right now, we're automating the higher skill jobs. And there is a very distinct limit to how highly skilled a large percentage of the workforce can become. We're on the cusp of automating away what a large percentage of office workers do every day. What are they going to do instead? Train to be doctors? Oh, wait, we're throwing machine learning and automation at medicine too, and that's showing a lot of promise.

We are fast approaching the time when we're going to be making robots and machine learning ("AI") that do almost anything better than the average human could do it. What do the average humans do then?

When we put all of the agricultural laborers out of a job, what are they going to do instead? What else are you going to train a migrant produce picker to do that can't also be done by a robot?

When we put several million truck, taxi, and bus drivers out of jobs, what are we going to train them to do? Stock shelves in the store? Cut hair? Make coffee?

When most of the accounting jobs go away, what do they do?

We've got no shortage of things for people to do. The problem is that inevitably, robots and machine learning are going to be able to do most of those things better and cheaper.

Re:Not surprising

[vux984]

"Hopefully most of all current jobs can be automated so we can find new things for people to do."

That only works if somehow they are also things robots can't also do better and cheaper.

If you develop AI/automation that generically approaches or exceeds median human capability, then it doesn't really matter what 'new jobs' you invent, because robots will do them cheaper and better than most people can.

Most people won't be able to find work at that point; it doesn't matter how many jobs there are. Either they'll be jobs they aren't able to do, or they'll be jobs the robots can do cheaper and more efficiently.

Think about it.

Heres another example. Most of the world's land mammal mass is in cattle, bred for slaughter. billions of them. Suppose we come up with vat-meat-substitute that is cheaper, needs less space, and tastes as good. What happens to the cattle?

Are we going to find a new use for the population? Sure a small number will survive, perhaps let wild, others in small organic farms for wealthy people to 'eat the real thing'. But the rest? There's nothing for them to do, we can't retrain them to help operate the vat-meat plants, they can't write novels... we'd pretty much wipe them out relative to their current number.

Humans are no different. If we come up with something that can outperform what the majority can do, the people displaced will not be able to find new work... without a fundamental change in how we think about work and wealth distribution there will be a revolution, war, and massive loss of life.

Re:Not surprising

[gweihir]

I really don't understand how people can continue to make this argument. It does not appear that this time is going to be like the previous times.

People are, on average, stupid. The ones that make this argument are afraid in addition and are trying hard to ignore reality by explaining it away. Said stupidity is the core reason this time will be different, because before it was always a shift to other jobs that the average idiot could still do. This time is different, because while most people can learn to read and write (making them capable to be bureaucrats as a last resort and we certainly have a lot of those today), this shift will only create a small number of new jobs and these will have requirements that 95% of the population cannot meet. The ones that are safe are engineers and craftsmen, because these do custom jobs were automation does not help, since automation scales with repetition of the same or similar jobs. ("Where "similar" is getting broader and broader...) Everybody else is fair game this time, with many jobs not about to be eliminated but to be done by a much smaller number of people because they only get to do the jobs that the automation cannot hack.

Re:Not surprising

[hazardPPP]

In the past, when we've automated low skill jobs, we've pushed people into higher skill jobs in the process. Farmers ended up working in factories. Factory workers ended up working in offices.

You're making a rather false assumption here, thinking that factory jobs are necessarily "higher skill" than farming. No. Especially in the early 20th century, working in a factory was a lower skilled job than farming. That's why any Joe could go to a city and find a job in a factory - they all used cheap unskilled labour en masse. In fact, the assembly line was invented for the exact purpose of using unskilled and low-skilled labour - divide the manufacturing process into a series of relatively simple repetitive tasks, and any idiot could learn to do them very quickly. That was the whole point. Industrialization allowed the use of unskilled or low-skilled workers to produce, en masse, cheaply, things that previously required highly-skilled workers that took years to train and made each product by hand from start to finish in a rather long process (in a word, artisans).

But right now, we're automating the higher skill jobs. And there is a very distinct limit to how highly skilled a large percentage of the workforce can become. We're on the cusp of automating away what a large percentage of office workers do every day. What are they going to do instead? Train to be doctors? Oh, wait, we're throwing machine learning and automation at medicine too, and that's showing a lot of promise.

The reason that so many jobs can be automated (and have been automated) is exactly because we turned them into a series of well-defined repetitive tasks...that's easy for any human to learn, but is also easy to design a machine for. It's therefore no wonder the first place industrial robots took off was the factory assembly line. Many office workers just shuffle paper (or computer files), so they are also automatable. In fact, we already had one huge wave of office worker replacement by machines a few decades ago - when computers became widespread. What one accountant can do today on a computer used to require a full room of people on abacuses and typewriters.

This stuff about AI and ML "showing promise" in medicine - is just you taking it too far. AI/ML will certainly be another diagnostic tool which will help doctors make decisions, but they will not replace them, just like MRI scanners have not replaced them. We just don't know enough about how the human body works to fully automate medicine, and that's not going to change in the near future.

We are fast approaching the time when we're going to be making robots and machine learning ("AI") that do almost anything better than the average human could do it. What do the average humans do then?

This is pure conjecture without any actual evidence to back it up.

When we put all of the agricultural laborers out of a job, what are they going to do instead? What else are you going to train a migrant produce picker to do that can't also be done by a robot?

When we put several million truck, taxi, and bus drivers out of jobs, what are we going to train them to do? Stock shelves in the store? Cut hair? Make coffee?

When most of the accounting jobs go away, what do they do?

We've got no shortage of things for people to do. The problem is that inevitably, robots and machine learning are going to be able to do most of those things better and cheaper.

You are showing a distinct lack of imagination (as are most people). Go back to 1750, and tell people how many millions will be employed in 250 years directly and indirectly, by the various entertainment industries (professional sports, TV, film, etc. etc.) - they would probably laugh at you, and if they manage to take you seriously, would probably say what a hedonistic and amoral society that must be. Yet here we are...doing much better than the folks in 1750, thank you. In 50 t

Re: Not surprising

[hazardPPP]

Today we have different jobs but on average they require a bit more brainpower than plowing a field did...

Right. I'd challenge any average office worker of today to go run a farm...not to make money, just to feed themselves. I'd wager they'd starve in a year.

Stop underestimating farming. Running your own farm required more brainpower than working on a factory assembly line. There are many studies which have shown that industrial work actually makes people dull, zombified, and stupid.

Re: Not surprising

[TheRaven64]

Right. I'd challenge any average office worker of today to go run a farm...not to make money, just to feed themselves. I'd wager they'd starve in a year.

Both you and the grandparent are making the same mistake: conflating education and intelligence. Most people are sufficiently intelligent to run a small self-sufficient farm. That's pretty much what 95% of the population was doing a few hundred years ago. The difference is that, back then, they were taught the basics of farming from when they were old enough to walk. Now they are taught to read and write instead. Most of those peasant farmers would have been able to do simple office jobs with the same training.

The real problem is the amount of time that the training takes. If you need skills that take 10 years to acquire to move to a new job, then that's not a good short-term solution for you. And it's also not a good long-term solution if that job isn't going to be around in 10 years and you don't know what skills will be required for the ones that will be.

Re:Not surprising

[drinkypoo]

This stuff about AI and ML "showing promise" in medicine - is just you taking it too far. AI/ML will certainly be another diagnostic tool which will help doctors make decisions, but they will not replace them, just like MRI scanners have not replaced them. We just don't know enough about how the human body works to fully automate medicine, and that's not going to change in the near future.

Lots of doctors will be replaced. We have a health care professional shortage due to the influence of the AMA. They're responsible for the nature of health education, which is designed to keep the supply of health professionals low in order to keep prices high. However, that only means there's more motivation to come up with automated health systems, so AMA member doctors (under 40% of doctors, mind you) are shooting themselves in the foot hardcore here. The doctor shortage will be solved primarily by arming nurses with expert systems.

Re:Not surprising

[Green Mountain Bot]

Say all the easy jobs go may away. I think people will get smarter and the average intelligence will rise.

That takes time, and assumes that the people who can't make a living just go away peacefully. But history is littered with the corpses of societies where they did not.

Even if people don't get smarter, they will still need to find a place to live, eat, etc.

Our current economy only cares if those people have the money to pay for places to live, eat, etc. Need is only demand if it is coupled with the ability to spend. Without that ability, it doesn't get addressed.

This is about the 8th or 9th of these

[rsilvergun]

maybe 10th. I've lost count. But at least half, maybe more, of all jobs are going to be automated in the near future. Hell, even even half of that is true it's 25%. Now would be a real good time for us to figure out what we're going to do when a quarter of the population is unemployable. In America if you don't work, you don't eat. And when people don't eat, they get violent and prone to suggestion. And we've got a _lot_ of bombs....

All I'm saying is, If the rest of the world doesn't want that 25% to start looking for some kinda strong man to get them jobs of the military persuasion maybe they should start doing something. Maybe stop destabilizing our politics (Russia, I'm looking at you) and stop encouraging right wing, authoritarians from getting into power.

Or don't. Nobody bothered much with Germany in the lead up to WWII.

Re: This is about the 8th or 9th of these

[Opportunist]

Yes, Bismark wanted a war with France, and France had to start it so the defensive pacts with the other German states would come into effect. And Napoleon III was enough of a fool to hand him that gift. There was exactly zero reason for Napoleon to start this war. None.

World War 1 was a different beast, though. While Wilhelm II was someone who had a spleen for uniforms and military hoopla, he was not the reason for WW1. That war started because Vienna got greedy and wanted that war more than anything to stabilize its already dissolving empire. The Austrian-Hungarian empire sent an ultimatum to Serbia after the attack on their heir apparent that was absolutely impossible to accept by any state. It was essentially the demand to cease to exist as a independent country and become annexed by the A-H empire. They accepted pretty much all of it, including that Austrian troops may search for culprits of the attack (to give you an idea just WHAT they accepted: Foreign military conducting search&seizure on your soil without you having any say what they can or cannot do), pretty much anything short of surrendering independence, which A-H did not accept as sufficient and declared war. Which set the whole system of alliances into motion.

Wilhelm himself said later that, had he known just WHAT Sarajevo accepted, he would not have supported that war, that (what Serbia accepted from the conditions) was more than anyone could sensibly expect.

If you're looking for a culprit of the war, look further south. Berlin was dragged into this mess.

Also, please remember that the general attitude towards war and starting one was completely different to what it's today. War was something you'd do if you run out of other options, not something that you got ostracised for by other nations. WW1 changed that attitude, but until 1914, wars were something where you'd go and kill a few soldiers, then trade land and everyone goes home eventually to prepare for the next. War was something nations do when they compete.

That attitude had changed by 1918, yes. But still the demands Clemenceau made were insane. If those contracts stood, Germany would have paid back its "fine" up until the 2000s. How exactly did he expect this to work? It's understandable that after this war people were pissed. On all sides, don't worry, I doubt that there was a huge amount of love for France in Germany either. Only one side had the means to flog the other, though.

You cannot found peace on oppression, though. Would you accept it? Imagine a war between your country and its neighbor, your neighbor wins (which you consider fake, too, because you still had guns and ammo when the ceasefire came into effect, you didn't lose, you were "assassinated" from the back), then makes ridiculous demands that your grandchildren will still be paying for. Will you pay? Or will you follow whoever promises you to get rid of that?

You don't need to be Germany for this to work. This works in any country. Anywhere on the planet. Anyone would try to stop this kind of treatment as soon as he even remotely sees any chance to do so.

How hard would it be to ...

[140Mandak262Jamuna]

... to take some report from some organization with an axe to grind and dress it up make news story out of it?

Is that job vulnerable to automation?

Look backwards.

[0100010001010011]

Rail museums are fascinating once you realize all the bits that humans had to do. Prior to the self lubricator being invented it was someones job to go around and make sure all N hundred points were properly lubricated. You had to have people physically down on each switch. The locomotives themselves had a 50% duty cycle.

All of it has been 'automated'. No one is pining over not being able to fire a tinder box. A modern locomotive may take a handful of people to do what used to take hundreds if not thousands.

The same goes for every other industry from food production to transportation. Humans are industrious creatures in that we'll find something else to do and new ways to be lazy. 50 years ago making your living in eSport or drone racing would have been unheard of.

Re:Look backwards.

[jeff4747]

The difference is there was some other job the workers could move to.

The trick with this round of automation is if we develop advanced enough AI to do the automation, there isn't going to be something else. Because that new job would also be automated away.

What we're bickering about is when it's going to happen, and what is "advanced enough AI". But it's going to happen. We might want to plan for it instead of hoping that the 1890s repeat themselves.

Mass retraining

[tepples]

During this shift away from agricultural labor after World War I, who funded mass retraining of the workforce? That might help us figure out who will retrain the current workforce for the age of automation.

Re:Mass retraining

[hazardPPP]

During this shift away from agricultural labor after World War I, who funded mass retraining of the workforce? That might help us figure out who will retrain the current workforce for the age of automation.

You didn't need a lot of retraining to work on a factory assembly line in the 1920s.

It is the pace of the change that matters

[Laxator2]

Whenever a story about automation comes up most of the replies are of the type:

"It happened 100 years ago and then again 50 years ago, it we ended up better. We'll figure it out this time."

What these posts don't say is that the _pace_ of the change was much slower back then. Also, we keep on comparing mostly physical industries (e.g. railways) with the current tech industries.

In the tech industry the pace of the change is much faster and during the 80's 90's and 00's it was accelerating. I remember growing up in the 80's when studying electrical engineering and making money from repairing TVs was a perfectly good way to make a living.

Then TVs became almost disposable and computers came along and in the late 80's and early 90's many people made a living writing stuff in BASIC and Pascal. Try making a living from those skills today, only 15-18 years later.

The point is that in the past the major changes took longer or at least as long as the turn of the generations. You could learn a trade and it kept you going until retirement.

However, nowadays you can expect 3-4 major changes throughout the employable years of a person, and not everyone is able to keep up the pace with such change.

Re:They are going to pay me for not working anyway

[geekmux]

A lot of people are saying that the government is going to pay us for not working anyway,

They do already. It's called welfare. Yeah, I'm sure we'll give it a fancy name in the future like "UBI" to appease the Starbucks hipsters and make it feel like something other than welfare, but in reality it won't be any different. The trillionaire owners and rulers of the future who will be asked to fund UBI will dodge that responsibility like they do taxes today, so certainly don't expect UBI to pay any better than welfare today.

Even Nobel Prize Economists, like Krugman, is saying the same

I would certainly hope a Nobel Prize winning economist would know a thing or two about welfare. As automation and AI create a global welfare state and put billions of unemployable humans in it, the real question will be mental health and stability.

Do I need to worry that AI is going to take away my job?

Yeah, but feel good knowing it will only take automation or perhaps "good enough" AI to take your job away, and that will be the same for the other 80% of the human population.

Do I have to worry?

Probably, but look at the bright side. You won't have a pesky job consuming 40 hours a week, so humans will be free to pursue their life's dreams. Or at least as free as welfare recipients are.

Re:They are going to pay me for not working anyway

[Kiuas]

The trillionaire owners and rulers of the future who will be asked to fund UBI will dodge that responsibility like they do taxes today, so certainly don't expect UBI to pay any better than welfare today.

You're missing something quite important here: consumption. Companies need customers, and if and when due to automation a significant chunk of consumers can no longer work because they have no marketable skills, consumption will come crashing down, which will hurt the companies and thus the bottom lines of the trilionaires. Think about the major tech companies for example: Google, Amazon, Facebook, Apple, etc. In a world where the majority of people would be making just enough money to survive these companies would collapse, because the goods and services they're selling all require consumers that can actually be marketed to effectively, meaning consumers with disposable income.

Put another way: the benefits of automation (reduced manufacturing & logistics costs) will help no-one if as a result of automation and reduced purchasing power the consumer-base will collapse and is replaced with a massive amount of people living on just the bare necessities because that will slash demand for most consumer products and services and cause a lot of these companies to go under which in turn will reduce the demand for B2B products and services.

This is not a matter of empathy for the poor, this is a matter of game theory. The economy is essentially a giant game that requires people to be buying things for it to stay viable. The more corporations embrace automation and cut the amount of jobs, the more inevitable it will be that the amount of UBI/income transfers per person will have to go up in the future. That is unless one maintains that the rich will suddenly stop caring about making money and will be fine with seeing their profits collapse.