Only 25 Percent of Occupations In US Are At 'High Risk' For Losing Jobs From Automation, Study Finds

An anonymous reader quotes a report from Quartz: Automation is coming, but not for everyone. Researchers at the Brookings Institution estimate just 25% of occupations in the US -- in production, food service, and transportation -- are at "high risk" for losing jobs from the advance of automation. "Automation is not the end of work," said Mark Muro, policy director for the Brookings Institution's program on urban economies and co-author of a study published Jan. 24. Most occupations will see specific tasks assumed by machines, but much of their labor will likely be enhanced, rather than fully replaced, through automation, the study found. That's because automation rarely replaces entire jobs, but instead handles specific tasks in occupations that often require hundreds of them.

To forecast the effects, Brookings researchers looked at thousands of specific tasks within each occupation, and the degree to which automation could handle them, coming up with a risk rating for each occupation. The workers most vulnerable are in transportation, production, food preparation, and office administration, which, combined, make up about 36 million jobs, or 25% of the total jobs in the US today. In these occupations, roughly 70% of tasks were considered routine and predictable, prime targets to be managed by machines. The most vulnerable were "packaging and filling machine operators" (100% exposure to automation), food preparation workers (91%), payroll and timekeeping clerks (87%), and light-truck and delivery drivers (78%).

Comments

Only?

[rsilvergun]

WTF? Seriously, what the actual fuck. Do you have any idea what losing 25% of the jobs effectively overnight (that's what "at risk" implies) means? We're a winner take all, if you don't eat you don't work society.

Those people aren't going to go quietly into the good night. Forget violence, a lot will start gunning for _your_ job. It'll be a race to the bottom like you've never seen before where the only winners will be the ones that own the robots pitting us against each other for their profit and amusement. You'll be lucky to make min-wage with a 4 year degree there'll be so many desperate workers.

There's a reason I can get a competent programmer in India for $30k/yr instead of $120k/yr, there's too damn many of them. If you don't want your standard of living to go to hell now's the time to do something about it. And no, buying lottery tickets or hoping your gonna get rich off your MCSE doesn't count.

Re:Only?

[drinkypoo]

It'll be a race to the bottom like you've never seen before where the only winners will be the ones that own the robots pitting us against each other for their profit and amusement.

So what you're saying is, study hacking and robotics?

Re:Only?

[Anonymous Coward]

"a competent programmer in India" that's a joke if I have ever heard one. I have not met more than 15% of them being competent, and much less than 5% of them being somebody I actually want to work with. "Competent" is a very low bar.

Re:Only?

[tsqr]

It's 25% of occupations, not 25% of jobs.

Re:Only?

[_Sharp'r_]

Virtually 100% of the carriage making industry has been lost. There were 13,000 of those businesses in 1890, now they're pretty much all defunct. There's almost nobody who hand spins thread for a living anymore, either. That used to employ vast numbers of people, mostly across the South.

Fortunately, there is literally no limit to the amount of work available for people to do, just a lack of people to be available to do it. Specific jobs have been being destroyed by automation for hundreds of years. Yes, the process has accelerated recently. Guess what, it hasn't had a noticeable effect on unemployment rates, rather it leads to increased wages over time as increased productivity as a result of more capital (automation) being able to be used by people to accomplish more.

No, this time won't be different. The luddite fallacy remains a fallacy.

Re:Only?

[ShanghaiBill]

Some more recent examples: Telephone switching equipment put hundreds of thousands of switchboard operators out of work. ATMs put thousands of tellers out of work (although overall bank employment went up). Typesetters are gone. "Secretary" is mostly a job that no longer exists. Same for "filing clerk".

Yet somehow, despite all these jobs disappearing, we have a full employment economy, and our economy has grown 500% since 1960.

30% of workers change jobs annually, and many of them switch to different occupations. Absorbing 25% job changes over a decade or so is something we have done many times before, with strong economic growth, rising wages, and better living standards.

Re:Only?

[Powercntrl]

No, this time won't be different.

Any investor is familiar with the phrase "Past performance is not indicative of future results."

Re:Only?

[Aighearach]

Calm downand read Player Piano by Kurt Vonnegut.

There is nothing to worry about, they'll find something else to do!

Humans are nothing if not engines for creating a feeling of purpose!

Re:Only?

[Aighearach]

If one missing comma makes it that hard for you to understand, maybe you should just turn off the computer, get another drink, and watch some teevee.

Re:Only?

[Aighearach]

Bourgeois are the tradespeople who own their own tools. The canonical example would be the cobbler (shoemaker) who owns his own shop and lives upstairs.

This is why you commies always fuck shit up so bad; you can't even figure out that it is the factory owner who shits on the factory worker! You're too busy hating on people doing things for themselves to even notice the causes of your own suffering.

My advice: be less jealous, and more greedy.

Re:Only?

[Aighearach]

Bullshit, if you were a carriage company when the automobile showed up, you kept making the carriages and started adding tow hitches.

The reason there are less companies doing it is because of consolidation. And that consolidation is made possible by their own products!

Carriage companies became trailer companies. There are many more trailers being made, and people employed making them, than there ever were carriages.

Find better examples, like a product that actually stopped being made, instead of just one that is called something different now.

Re:Only?

[Aighearach]

Filing clerks didn't go away at all, they're called "data entry specialists" or something now.

Secretary is now called either Executive Assistant or Receptionist depending on their level of duties.

Telephone switchboard operators and typesetters are better examples because it was a different person doing the work after the change. ;)

Bank tellers are a mediocre example because those are low level jobs where people would eventually get promoted to something else in the bank anyways, so few actual jobs were lost; hiring of tellers went down, but individuals weren't getting laid off.

Re:Only?

[satsuke]

That's a false equivelancy because in your example, those jobs that were displaced were recreated doing something else in a similar field.

With automation, those jobs are gone .. poof .. no similar spot to take its place.

In an ideal world, everyone else would just work less hours, but that won't get around the reality of it being easier to push the remaining people harder than it is to lessen their load at the bosses expense.

Re: Only?

[Monster_user]

Increased wages over time? When you have pressure from automation reducing the demand for labor and procreation increasing the supply of labor, it is hard to see how supply and demand would increase the value of labor.

Automation frees up limited resources to improve profits for an increasingly smaller number. The value of automation is seen by the wealthy more so than the poor. For the poor it is an added expense, an additional pressure towards a minimum level of capacity to be competitive and remain employed. Because there is a sufficient supply of optimal labor, inefficient sub-optimal labor is unnecessary. Inefficient sub-optimal labor has a lower return on investment, if it has one at all, and thus has no economic incentive to fund the life towards automation and optimization.

How much does it cost to automate your home? To replace appliances and devices with more energy efficient alternatives? To install double-paned or other more moder window types? To replace a vehicle with an energy efficient alternative, or one with greater capacity and accident mitigation technology to reduce the risk of damage. Can one afford to live within a city on the income provided by said city?

We are raising the cost of entry while simultaneously lowering the return on investment.

An Alaskan said it best when he said that he feels sorry for those of us on the mainland, because Alaskans can live off the land.

Re:Only?

[CrimsonAvenger]

WTF? Seriously, what the actual fuck. Do you have any idea what losing 25% of the jobs effectively overnight (that's what "at risk" implies) means?

Yeah, it'll be even worse than the introduction of tractors and harvesters to farming was! Admittedly, the impact on farming jobs only meant that farming jobs went from 80%+ of the population to 5% effectively overnight. Not nearly so bad as this will be....

Re:Only?

[_Sharp'r_]

It turned out that bicycle manufacturers were much better equipped to make auto-related products. The carriage companies were essentially wood-based, with lots of carpenters making stuff. They all went away. Studebaker lasted the longest by actually somewhat successfully pivoting into automobiles, but even they were eventually absorbed with only the brand name remaining.

Re:Only?

[Somervillain]

There are plenty of competent programmers in India. They just want American wages, not 30k/year. Although most engineers recruiters present to me as "competent" for our Bangalore office are far from it. I sympathize. If you're good at your job, why work in Bangalore when companies will pay you to relocate to the US or UK for much higher wages...or find a job paying Western wages in India (like most tech giants do for their good projects).

Re:Only?

[The Evil Atheist]

Did you seriously just claim we have "rising wages" and "better living standards"? They've been stagnant for a long time. As for strong economic growth, considering the widening gap between rich and poor, the growth is obviously one sided and averages mean nothing.

Re:Only?

[ShanghaiBill]

Did you seriously just claim we have "rising wages" and "better living standards"?

NO, I did NOT say that. What I said was we had rising wages and living standards while jobs were being automated.

Automation leads to higher productivity, making labor more valuable, and thus leads to higher wages.

They've been stagnant for a long time.

They have indeed, and this is due to stagnant productivity growth. The problem is not automation, but a lack of automation. Manufacturing and agriculture jobs were easy to automate back in the 20th century, but those days are over, and modern service jobs are proving much harder to make more productive.

If TFA is correct, and the pace of automation improves, then we should see wages start rising again.

Re:Only?

[elrous0]

People start rioting at even 20% unemployment. And that 25% of jobs displaced would produce WAY more than 20% unemployment, and would also disproportionately affect working-class jobs (you know, the people who have the least to lose by rioting).

So yeah, that would not be good.

Re:Only?

[JimSadler]

Either way the shock wave will be enormous. Vast areas of work will be almost eliminated such as school teachers. An eighth history class could be broadcast to every eighth grade student in the US. Professional drivers will be hit really hard. Meanwhile, other tasks that have variables may require only a human attendant. For example a lawn cutting robot might be super efficient with one human attendant in case something odd takes place rather than a three or four man crew. Lawyers are already being hit by machines that can take care of basic legal forms and the like. Automated guard and police services are already somewhat in play. The entire reason for technology is replacement of human labor and now technology is winning.

Re:Only?

[_Sharp'r_]

Please don't read into what I said things which I didn't say. No need for strawmen.

There is literally no limit to the amount of work available for people to do. Once we've converted this solar system to our liking, we can move on to the rest of the universe. Then we can redo it all again. There are an infinite amount of possibilities for something useful for someone to do, which is all that work is.

New opportunities for work become practical as we gain more technology or figure out ways to accomplish existing work someone is doing more efficiently. Available work isn't some static number which is capped at the amount of work currently being done by someone. There's always something new which could be done to improve things and increase the overall wealth of humanity.

Re: Only?

[PaoloAgati]

I think that the old motto "Workers of the world, unite!" should be taken again in consideration... 25% is enormous... And I think it's a low estimate

Re:Only?

[dgatwood]

Virtually 100% of the carriage making industry has been lost. There were 13,000 of those businesses in 1890, now they're pretty much all defunct.

Nonsense. There are now just a few dozen companies, but there are at least 4x as many people, so presumably at least 4x as many (horseless) carriages, and probably more given that most families only had one carriage, and the average family probably has 2.5 cars or something.

There's almost nobody who hand spins thread for a living anymore, either. That used to employ vast numbers of people, mostly across the South.

Even that was replaced by similarly semi-skilled jobs, though. The difference between those changes and the current changes is that right now, it is looking like in 20-30 years, there will be no more semi-skilled jobs.

It's one thing to retrain someone to do something else that requires only average mental ability. It's quite another to say, "If you cannot get a job as a delivery driver, you should retrain to become a brain surgeon." Believing such a thing is broadly possible would require borderline Marie Antoinette levels of naïveté.

Fortunately, there is literally no limit to the amount of work available for people to do, just a lack of people to be available to do it. Specific jobs have been being destroyed by automation for hundreds of years. Yes, the process has accelerated recently. Guess what, it hasn't had a noticeable effect on unemployment rates, rather it leads to increased wages over time as increased productivity as a result of more capital (automation) being able to be used by people to accomplish more.

That is only true because as humans are still needed for some parts of the process. We're rapidly approaching a tipping point where that will no longer be true. And we're already seeing increased unemployment from automation.

A 2017 study out of MIT and Boston University found that from 1993 to 2007, for every one additional robot added per 1000 workers, unemployment increased by .18–.34% and average wages decreased by .25–.5%. They found that somewhere between 360,000 and 670,000 jobs were lost due to automation, and further found that "Interestingly, and perhaps surprisingly, we do not find positive and offsetting employment gains in any occupation or education groups."

Let that sink in for a moment. Jobs were lost and nothing took their place. Unemployment increased because of automation. And nothing replaced those jobs. And the jobs that remained paid less money because people were competing against automation and trying to underbid it. And the more automation replaces people, the faster it will do so.

Re:Only?

[dgatwood]

Vast areas of work will be almost eliminated such as school teachers. An eighth history class could be broadcast to every eighth grade student in the US.

This could happen already. It turns out that having an actual knowledgeable person in the class matters a great deal, because passive learning doesn't work nearly as well as active learning.

The other things, yes, but education probably won't be automated for a long time to come.

Re:Only?

[_Sharp'r_]

Carriage companies didn't turn into automobile companies. Woodworkers didn't make the best car makers. Instead, bicycle companies turned into automobile companies and their suppliers. The Carriage companies literally went away, with only one (Studebaker) lasting for very long, before it also went out of business and the brand name (not the workers) was absorbed by a car company.

At one point, 70-80% of everyone was employed in agriculture, primarily as labors. Now 1% is. That's way more jobs "going away" than 25%, yet somehow those people made the transition to more valuable employment in less rural areas.

As for the 2017 study you cite, here's a quote from one of the study's authors, Daron Acemoglu:

The media, in this field as in many other areas of technical expertise, is attracted to extreme statements rather than focusing on a balanced discussion of what is known in the academic area. In the context of the future of work, this takes the form of statements claiming that new technologies will bring the end of work for humans. Nothing in the serious research in this area suggests that something like this is in the cards.

If you want to use that study to argue the Luddite case, you should go argue it with the author. I'm going to assume he knows the results of the study better than you or the sensationalist media does.

Re:Only?

[mapkinase]

>Do you have any idea what losing 25% of the jobs effectively overnight (that's what "at risk" implies) means?

I do not have the slightest idea either, since I do not know how many workers are for each of the jobs. I seriously doubt that researchers analyzed "thousands of specific tasks within each occupation". They analyzed some occupation and then they extrapolated it on the rest of the jobs.

A simple crucial information needed. Without this information we can't say.

Logically, I would assume that researchers would start with analyzing the most common jobs, not the piano tuners, or ALGOL programmers. That would actually mean that most likely (only if the researchers are believable, I am not sold on that) much more than 25% of the total working population is affected.

Instead of even counting 25% of jobs they should counted the number of professionals in each of the jobs they really "analyzed".

Re:Only?

[mapkinase]

His example is unconvincing, as you rightfully pointed out, but general look at the history of humanity shows that we recovered from every single problem we had. :-) In this generic aspect he was right.

The solution would be some kind of society where (a) a lot more is given to unemployed than now in US (b) a lot more policing (Orwellian to the bone, Kafkaesque in implementation, Huxleyan in the results) is done against disenfranchised population depressed for the reason of being useless to the society.

Re: Only?

[mapkinase]

>An Alaskan said it best when he said that he feels sorry for those of us on the mainland, because Alaskans can live off the land.

I am not sure I understand what that means. Is it sarcastic? I need a sarcasm sign.

Did he really mean that Alaskans live off the land (which is patently not true - Alaska is heavily subsidized) - then I do not really see a point. How is that relates to "We are raising the cost of entry while simultaneously lowering the return on investment."

If he is sarcastic then I do not see the connection either.

?

Re:Only?

[AmiMoJo]

While I agree with your point I think we need a better metric than employment rates or measuring success.

The UK had fairly decent levels of employment, but a large number of those jobs were of very low quality. Often they were "zero hour contracts", where the company calls you up to tell you how many hours you get to work that week. There are also a lot of low paid jobs which while technically employment leave the worker living on benefits, essentially passing the cost on to the taxpayer.

Re:Only?

[sad_]

so it might be even worse then 25% of jobs!
if those happen to be just those 25% of jobs that employee the majority of the people.

Re:Only?

[Kiuas]

Guess what, it hasn't had a noticeable effect on unemployment rates, rather it leads to increased wages over time as increased productivity as a result of more capital (automation) being able to be used by people to accomplish more.

This is true for some, but not nearly all, workers. The fields directly related to automation such as programming are in high.demand and have seen wages grow, but this is not the case for jobs requiring less education. As automation becomes more and more commonplace, it means there's fewer and fewer jobs available for those doing manual tasks. In fact, if you look at OECD statistics such as labor compensation per hours worked, you'll note that the rate of increase in wages is in fact dropping. In the case of the US for example, the rate of growth has dropped from 6.53 % annually in 2000 to just 1,32 % in 2016. It's since come up a little bit from there to 2,7 % but when you combine this with inflation and a rising cost of living, real wages are actually falling now in the US (and many other western countries) because the rather modest wage growth is not matching these. The same trend can be observed elsewhere in the developing countries, including the euro area.

This shouldn't be that hard to understand in practice. Take something like invoicing, or any other job that doesn't require a high amount of skill but has been critical in employing many, many people on the office side of things. While I was studying I used to work such a job during the summers especially. Back then the office side team consisted of nearly 30 people as the job involved a lot of manual tasks, including scanning paper invoices, manually entering information about what's being paid so that the books are in order, matching the invoice with its existing purchase order etc... With electronic invoicing increasing at a rapid pace, these tasks are quickly becoming more and more automated as the systems learn to pre-fill this information and you can use less and less people in the process. Now, as this happens and companies cut the number of workers in these tasks, productivity per person indeed goes up. One person can easily do what in the past took 5, as the most time consuming parts of the process have been cut out. But do you think this means the companies end up increasing the wages of the remaining staff? Because that is not happening. There's simply no need to do that. The tasks themselves have in fact gotten easier to do, as now all the guys overseeing electronic invoicing have to do is check that the software hasn't missed anything. Since the goal of implementing these systems is to cut costs by cutting labor costs by reducing staff, it makes no sense for companies to then eat up their savings by bumping the salaries of the remaining workforce. It's not like the requirements of the job have gotten any more difficult. With less and less of such jobs available, the competition for the remaining spots will become fiercer and fiercer as automation proceeds to become more commonplace. With many more people wanting to do these jobs than there are jobs, oversupply means the wages will not grow and may even fall. Why pay more when so many people are willing to do it for slightly less?

No, this time won't be different. The luddite fallacy remains a fallacy.

The end-point of this development is cutting humans off the loop entirely. In 20 years, most data entry jobs will be gone, autonomous vehicles will make a lot of drivers unemployed. Manufacturing and logistics are already being automated at a high pace both in the West and in Asia. And so on. The more advanced automation and AI becomes, the less need there will be for humans across the board. The luddite fallacy rests on the notion that companies will forever pre

Re:Only?

[Durrik]

This is the problem my company faces. I also wish they didn't see $30 k in India as luxury wages. There's plenty of competent people in India, but they're competent enough to know that they can get much more than $30 k, or the peanuts my company pays.
The cheapness of my company is also the reason why my job isn't at risk of automation, they look to save money for this quarter, and will never make the up front investment to automate my job. Ship it off to India if they can, but automate it? No.
I also write the build and test automation software as one of my many hats, and I do my best to automate my job so they keep giving me more jobs that I automate. Fortunately its positioned us to get into industries that require a lot of certifications on code quality and process and that means much more work so in a way I'm creating employment by automating the stuff I am.

Re:Only?

[tsqr]

Maybe you should read and understand the article instead of scaring yourself needlessly.

Re:Only?

[drinkypoo]

Fortunately, there is literally no limit to the amount of work available for people to do, just a lack of people to be available to do it.

There's also a lack of people willing to pay them to do it. There's lots of things that need doing that nobody has figured out how to profit from, and they're going undone. For example, meaningful fire prevention for California, reclamation of regions desertified by human activity, etc. But then there's the flip side of that argument, which is that those activities can be automated as well, and they will be eventually. So no, there is a limit to the amount of work available for people to do. Under capitalism, work tends to be limited by the speed at which some already rich prick can find a way to profit from it.

Re:Only?

[drinkypoo]

There is literally no limit to the amount of work available for people to do.

Ugh, stop saying that. It was meaningless the first time you said it, and it's still meaningless. The amount of work which could be done in the universe isn't necessarily going to be done by humans, nor should it.

Once we've converted this solar system to our liking, we can move on to the rest of the universe. Then we can redo it all again.

No, we can't. Ever heard of entropy? There's a limit to how much can physically be done before it results in the heat death of the universe. And on a more personal note, we are using up more natural capital than is being replenished here on Earth. At this rate, we are not actually going to make it to the rest of the universe. We literally need to do less work, at least of the unsustainable kind, if we are going to avoid driving ourselves to extinction.

Re:Only?

[unixisc]

Precisely! They talk like 80 million people losing their jobs is no big deal. Maybe we should work on automating the work that Brookings does, and then see how their staff feels about it.

The biggest problem is the assumption that elitists make - just go and find a job in a new line of work, develop a new skill... Sure, b'cos any trucker who loses his job to a self driving truck or an Amazon Scout can simply learn Java or Python or whatever language fad of the moment there is at the time, move to a shithole like San Francisco or Seattle or DC, and then live happily ever after!

I can't wait for the jobs that all these elitists do to be automated, so that they're left applying for jobs as Walmart greeters

Re:Only?

[unixisc]

Over time? Yeah, people don't train for jobs that no longer exist. Once cars became mainstream, people stopped learning to be horse & buggy drivers, but the people who already were that lost their jobs, and had to find other lines of work. Same w/ typesetters. ATM tellers were one group of people who could learn other skills in the bank that had not been automated (although an ATM can't (yet) give me a roll of quarters, or a few tens, which is why I still go to a human teller)

It's the same thing here. Let's say self driving trucks become the norm. Millions of truckers would then lose their jobs, as jobs like the transportation and delivery of products becomes automated. Question then becomes - what other jobs are there that exist throughout the country - not just in places like San Francisco or DC? Usually, it's things like education or healthcare, which is something more dominated by women. Since a trucker is unlikely to go get a job as a teacher or a doctor, it just puts men in the community at a disadvantage vis a vis the women, who more often than not, do not wanna marry them. Result of that is the breakdown of families.

If we could automate all jobs and come up w/ a new mechanism of paying everybody whatever they need based on those needs, it would be one thing. But this partial automation trend - particularly of male dominated jobs - promises to be a transitional disaster!

Re:Only?

[unixisc]

But UBI is something that would have to come out of somewhere other than government/tax base, since we are already $20T+ in debt. In short, either something like bitcoin mining for everyone, or allowing anybody and everybody to print their own currency for whatever they need. In short, abolish debt. Create a whole new set of 'rights' for everyone - a home is a right (solves homelessness), a car is a right (particularly a self driving one, if one doesn't have a driver's license) and so on.

Re:Only?

[apoc.famine]

That's what baffles me about these discussions. Every time we talk about automation, someone comes on and posts about how industries were lost in the past, but we always found new things for people to do. That completely ignores the giant problem of machines being faster and better at so many things that there increasingly will be nothing productive for us to really do.

AI just crushed some of the greatest Starcraft players in the world. That's an insanely high level of organization, management, and efficiency skills. Imagine what you can apply a similar AI to. Like building management, supply chain design, any number of logistics fields!

Sure, that's still in its infancy, but it's coming, and coming fast.

In the article on the Starcraft battles, one of the human players said that the AI was winning using novel strategies that nobody had ever seen before. He was excited because that meant there was more to learn about the game. That terrifies me, because of how many people over the years couldn't come up with a creative, novel approach that an AI could figure out in 14 days of learning.

Machine learning/AI and automation just won't leave enough stuff for us to do for jobs. FFS, we're already turning this on the arts, with moderate success! If humans aren't even able to compete on creativity, what do we really have to offer? Not efficiency, cost, reliability, speed, or consistency.

Short-term, we're likely fine. We're going to have to build all of these systems that are going to put us out of a job. We're going to have to debug them and maintain them. But just like multi-million-dollar room-sized computers are now $10 disposable items, AI and automation will go that route too, where they already haven't. That's when we're really screwed.

Re:Only?

[terrycarlino]

Filing clerks didn't go away at all, they're called "data entry specialists" or something now.

No they're not. A file clerks job was to take a piece of paper and put it in a file. At that time everything was recorded on paper and had to be filed. A data entry specialist is someone who takes information on a form printed on paper and enters it into a computer. Very few places use this method to enter data into a system and those that do are rapidly disappearing. In almost every case direct entry of data is done by someone else whose job is not data entry, such as the mechanic who will start to work on you car as soon as he enters your data into the computer.

Secretary is now called either Executive Assistant or Receptionist depending on their level of duties.

An Executive Assistant is not a secretary. High level executives always had Executive Assistants. In the bad old sexist days they were almost always men, while secretaries were almost always women, unless you want to go back to the 19th century. Receptionist have the same job they always had. Previously every mid to high level executive had their own secretary. Her job was to take dictation, i.e. allow the executive to compose correspondence, either letters, memos or in very rare cases contracts verbally. The secretary would then type it up, the executive would sign it and she would mail it out, by sending it to the mail room, which was full of other people who no longer have a job.

Now all but the most high level executives type up their own memos, letters, etc and send most of it by email. The job was not really automated, the skills needed to do it were taught to executives and the work pushed off on them.Software made it easy enough to do that . Specialize knowledge (shorthand & typing) was no longer required.

Telephone switchboard operators and typesetters are better examples because it was a different person doing the work after the change. ;)

This is exactly like what happened for your other examples.

Bank tellers are a mediocre example because those are low level jobs where people would eventually get promoted to something else in the bank anyways, so few actual jobs were lost; hiring of tellers went down, but individuals weren't getting laid off.

Bank tellers are a bad example because it's been documented that when ATM machines came along banks that use them actually accumulated so much more business that they generally open new branches. Branches which require even more tellers than they had originally. So the number of tellers working at banks went up.

Re:Only?

[apoc.famine]

We literally need to do less work, at least of the unsustainable kind, if we are going to avoid driving ourselves to extinction.

But where's the profit in that?

Re:Only?

[apoc.famine]

The article on how AI just whooped some of the best Starcraft players has definitely got me a little scared. That's a management and logistics application of pretty high order. That's the sort of thing that could help run a company.

The luddite fallacy rests on the notion that companies will forever prefer or need humans instead of machines to perform tasks, but that's not set in stone.

And it rests on the notion that humans in companies will be the ones making that decision. When the AI says, "The least efficient position in the company is X, it should be automated.", a lot of C* employees are going to take that recommendation. We are not really that far from that day.

Re:Only?

[gweihir]

Indeed. 25% is pretty extreme from its effects. In addition, 25% is at best a short-term estimate. Long term it will be far, far worse.

Re:Only?

[gweihir]

Completely agree. These numbers are nonsense. And to make it worse, they are easily manipulated up to give the appearance of success or to hide severe problems.

Re:Only?

[dgatwood]

Carriage companies didn't turn into automobile companies. Woodworkers didn't make the best car makers. Instead, bicycle companies turned into automobile companies and their suppliers. The Carriage companies literally went away, with only one (Studebaker) lasting for very long, before it also went out of business and the brand name (not the workers) was absorbed by a car company.

Actually, you missed a bunch of companies.

• Holden, a saddlery company, transitioned to making sidecars for motorcycles and eventually became part of GM.
• Brighton Buggy Company, founded in 1894, was subsequently moved and renamed to Hercules Buggy Company in 1902, and merged with the other Hercules companies in 1920; Hercules still builds truck and custom van bodies to this day in Henderson, KY (just across the river from stoplight city).
• George V. Arth & Son began as a buggy builder and transitioned to being a car body shop.

It seems likely that many more of those companies continued to exist in one form or another, though it's hard to say how many. Either way, carriages were still in use in some parts of the U.S. well into the 1920s, which means the transition from carriages to cars took about forty years to complete (and of course a few groups, such as the Amish, still use them today). During those four decades, young people who would have gone into woodworking instead went into car manufacture, so job loss was a lot more limited than you seem to assume.

By contrast, with automation getting better and better, manufacturing is likely to approach full automation in the next decade, and other job classes are getting pretty close as well. Ostensibly, any physical job creation has always been only temporary until it could be automated, but when the time period for automation was measured in decades, centuries, or millennia, things were okay. But the growth rate of automation is accelerating rapidly, and so the time period for replacement by robots is getting shorter and shorter. At some point, there will not be enough of those temporary jobs for anything approaching full employment, which also means the pay will also collapse to the minimum allowable by law.

And this is what study after study have shown is happening. Things are *not* like they were in the 1800s, because new jobs are no longer arriving quickly enough to replace the job losses from automation. We aren't even seeing an increase in high-skill/high-education jobs to make up for the losses in low-skill jobs, much less an increase in jobs that the original workers can be quickly retrained for. We're just seeing a net loss in jobs. So although ostensibly there is always more work to be done, the rate of automation replacing jobs exceeds the rate at which people think of new things to be done, and most new manufacturing and similar is highly automated from day one, which reduces the rate of job creation even for new projects.

Re:Only?

[dgatwood]

Oops. I should have commented on the second part.

The media, in this field as in many other areas of technical expertise, is attracted to extreme statements rather than focusing on a balanced discussion of what is known in the academic area. In the context of the future of work, this takes the form of statements claiming that new technologies will bring the end of work for humans. Nothing in the serious research in this area suggests that something like this is in the cards.

If you want to use that study to argue the Luddite case, you should go argue it with the author. I'm going to assume he knows the results of the study better than you or the sensationalist media does.

Obviously it will not bring about the end of work for humans, and obviously the article does not say that it will. What it will do is bring about a rapid (and rapidly increasing) reduction in work for humans who do not have adequate levels of education or skill to do more complex jobs. The problem is, we have an alarming number of people who are not well educated. These people are likely to be absolutely screwed within the next couple of decades, because there won't be any work for someone who isn't either artistic or college-educated (and not enough work for many of the people who are). And even though a percentage of those people will retire before the transition is complete, they will be replaced by another generation who are also unlikely to be well educated — not just because of the genetic component of intelligence, but also because kids who grow up in an educated family that puts a high value on education are more likely to care about becoming educated themselves and tend to do better in school.

So what will we do with all of those extra workers? Do we have the government create artificial work projects that could be done for far less money through automation just so that they will have something to do, or do we pay them money to do nothing and hope that they use their free time for something productive or creative? Because the only real alternative I can see to doing one of those things would be to let them starve, and historically speaking, that usually leads to guillotines and heads on pikes.

Re:Only?

[tsqr]

Question then becomes - what other jobs are there that exist throughout the country - not just in places like San Francisco or DC?

Plumber. Auto mechanic. Carpet layer. Tile setter. House painter. Carpenter. Steamfitter. The list goes on and on.

Re:Only?

[dgatwood]

Whoops. Screwed up the formatting there. That third paragraph was supposed to be part of the outer quote. Mea culpa.

Re:Only?

[dgatwood]

Only a modern idiot believes that it requires more skill to sit in an office doing what is effectively busywork than to raise crops.

You can teach people how to pick fruit or cotton or do other basic repetitive tasks in minutes. This is not true for office work. You need, at a bare minimum, basic literacy to do even the simplest tasks like filing, and you need computer competence to do anything significant. For people who have those skills, they aren't a big deal. For people who don't, you're talking about a significant learning curve.

A mechanic or anyone who works on engines, process systems or construction typically has many more skills that most of the office drones that spend their lives pushing around electrons for a living, despite the fact they are not compensated for it to the same level.

Probably true. However, it should be noted that skill level has very little to do with whether a job will be automated, and even intellectual requirements are not necessarily predictive. It takes a fair amount of intellect, skill, and learning to read bad doctor handwriting and type it into a medical records system properly. However, computers are getting better at handling these tasks automatically, and more medical professionals are doing the data entry directly as they work. Eventually, these tasks will become entirely unnecessary.

The primary dividing line between things that are readily automated and things that aren't is creativity. If your job doesn't involve creativity, you can assume that it is just a matter of time before it goes away.

For example, architects will exist for a long time. People building houses will continue exist for a while, because lots of stuff has to be figured out on the job site based on interpreting the architectural diagrams. But as prefab house construction becomes more automated, at some point, you'll see more and more prefab houses being built by machine and delivered, because the cost savings will eventually make these more common than traditionally built houses. It is unclear how quickly this will occur, of course. And there will still be a need for carpenters to do renovations and repairs until robots get good enough to do work on-site, which could be a very long time because of the need to creatively tweak the plans based on real-world conditions.

Heinlein had it right. Specialization is for insects. People are generalists. And a generalist can always move on to another field. A lot of the up and coming generation see it. A person is not their job and the purpose of a job or jobs is to generate the income needed to do what you want to do. Once you make more than you need you're just keeping score and more and more people are losing interest in just keeping score.

Well, yes, but only to a point. People are generalists, but each individual can function usefully only within a narrow ability band. Give that person a job that is beneath him or her, and the person will get bored and do a bad job or leave. Give that person a job that exceeds his or her capabilities (mentally or physically) and the person will fail. Two keys to maintaining stability are ensuring that there are always enough jobs in each of those ability bands to accommodate the workers and finding ways to improve people's abilities so that they can work in ever-higher bands over time.

Re:Only?

[_Sharp'r_]

You're analysis is completely missing the effect of the massive increases to overall wealth which would be a necessary result of automation improvements being worth replacing all those people's efforts. That increase in wealth increases the demand for consumption, moving the quantity supplied of consumption in goods and services currently too expensive at the margin into the realm of actually being provisioned.

i.e. we currently purchase so much fast food because we're wealthy enough to afford to buy it and it makes sense due to the value of our time being higher than a less assembly-line version of meal preparation at home.

Re:Only?

[_Sharp'r_]

People who used to do invoicing mostly didn't transition to programming. They transitioned to another job with comparable human capital requirements which wasn't previously getting done, in part because the wealth increase of the efficiency gains from computers doing invoicing allows those participating in the economy to spend that new wealth on new stuff the humans are supplying. As a result of technological advances (in many areas), you can buy food for $1 in less than 3 minutes which was previously unavailable except in exchange for multiple days wages in the past.

If there is an end state of automation+ai can do everything a human can do just as good and humans aren't needed to work on anything anymore, that end state looks like a paradise of massively increased wealth, where no one would need to work who didn't want to, exactly the opposite of the dystopian predictions of the modern Luddites.

Re:Only?

[_Sharp'r_]

Yep. Fortunately, increased efficiency from automation also results in increased overall wealth and an increase in wealth is an increase in the supply of "people willing to pay them to do it."

We currently pay lots of people to do jobs which would have been unthinkably ridiculous 100 years ago. Really, you need someone to choose what to wear for you? Yet the job now exists and is currently growing, based on automation via online clothing stores enabling people to supply their expertise to the masses.

Re:Only?

[_Sharp'r_]

Yes, I've heard of entropy. I anticipate employment in the project to access alternate universes and creating new universes to be booming... someday.

Re:Only?

[_Sharp'r_]

Right. The argument basically boils down to the economists saying, "This fallacy has been seen every time big gains from automation were in the works...." and the the other side saying, "But this time it will be different!"

I'd have more confidence in the second side if they ever demonstrated they actually understood why economists believe what they do about an issue in the field they spend their lives studying and made an argument which was actually on point for that discussion, rather than continuing to argue a point which has been debunked by empirical reality over and over again.

So if you are of the second side, please answer why "there won't be new jobs" is:
1. Likely (taking into account the increase in wealth from the efficiency gains of the expected level of automation to replace that many humans)
and
2. A bad thing on net.

Then we can have a discussion at least on the real empirical questions.

Re:Only?

[unixisc]

Why do you assume that those jobs won't be automated? Particularly auto mechanic - when so much of cars are now computerized and have to go through diagnostics?

Also, from your list above, only physically fit, strong and active men would be employed?

Re:Only?

[tsqr]

I'd like to see a robot replace a tie rod or pull a transmission, or remove and replace an electroc motor in an EV. And yeah, those occupations are for fit, strong, and active persons - the unfit, weak, and inactive ones will have to remain in IT.

Re: Only?

[Monster_user]

I think you may be confusing being subsidized with needing subsidies.

Have you considered that it is the mainland that requires the subsidies for Alaska, not Alaska's die hard citizens and "natives"? Would you refuse subsidies for any reason?

What is the true purpose of those subsidies? What vision got them approved?

Re:Only?

[_Sharp'r_]

Not sure if you're disagreeing with me or not on the first part, but to clarify, whatever the difference between the cost of the automation and the cost of the person to do the same job is the increase in wealth created by the automation. As in the vast majority of cases, the automation wouldn't be put into place without some sort of substantial savings, the automation trend creates a substantial amount of wealth.

As a tiny example, if it used to cost a manufacturer $10 to make a set of dinnerware using human labor, but using an automated machine only costs $8 to produce the same dinnerware, then $2 of new wealth per dinnerware set is created. If previously the dinnerware set sold for $13 (to the store, or whoever is next in the supply chain), then that extra $2 will end up split between the manufacturer and the consumer. If there is competition in the dinnerware market, the first manufacturer to automate will capture some of that $2 for a while (paying off investors), but eventually all of the manufacturers will have automated in a similar fashion and their price will fall to $11 with almost the entire $2 extra ending up as a consumer surplus. The previous $10/set which went to a human is now $2 going to consumers and $8 going to the plate-making-machine company, their employees, their suppliers, those suppliers employees, investors, etc... (i.e it doesn't vanish, either). The extra $2 doesn't vanish, it ends up in many people's pockets, who now can either use it for consumption (increasing demand now) or save/invest it for future consumption (partially increasing demand now based on what was invested in needing to purchase something and partially increasing demand later based on the interest returned from the new wealth created by the investment).

(Disclaimer: This example uses a 1:1 result, which is unrealistic as a result, but in reality you have to figure way more people and consumer surpluses involved and sorted out by the marketplace involving both.) At the same time, the human involved now wants a new job. In the past, consumers weren't wealthy enough to hire people to arrange their dinnerware, but now with their newfound $2 per dinnerware set of consumer surplus, investment proceeds, etc... they can afford to hire someone to do that in addition to all the other stuff they're consuming, so the human involved becomes a dinnerware arranger, a job which never existed before. The human involved also gets that fraction of $2 in consumer surplus if they buy a dinnerware set for them-self, as well as another smaller fraction if their retirement account is invested in the dinnerware manufacturer, and of course the more automation happens, the more of those types of returns on it they receive. Or if enough new wealth is created by automation, they may choose to use their new surplus to be able to work less and maintain a similar living standard, because they value leisure more than the additional consumption they could have from doing more work (i.e. new wealth is sometimes translated to more demand for leisure vs. more demand for work, based on people's choices).

If you push that out to the conclusion where everything is 100% automated and we don't need anyone to work (every possible job is filled by automation like intelligent nano-machines consuming all to produce stuff), then at that stage we are so wealthy people can just consume the results of the automation and live a life of massive enjoyment with no need to work on anything except creative stuff for fun. That's called a good thing....

Re:Only?

[Aighearach]

Mergers aren't the same as "going away." The same type of mistake could lead a person to think that cars were an unsuccessful product in the same time frame, because look how many companies there were before highways were built, and how many fewer there were afterwards!

All the industries were experiencing consolidation, "number of companies" doesn't tell you anything. It is not a valid proxy for number of workers. And which company's name was kept is often strategic, so the companies that used to be something different are less likely to have the name survive. But it doesn't mean they went away.

And you might be surprised by the amount of wood that was used for decades in trailers. Or perhaps even by the amount of metal in a carriage.

My advice, visit a "Pioneer Museum" and take a close look at the covered wagons that were used to settle the west coast! Lots of metal, very very similar construction to a flatbed trailer. A person building trailers in the 1830s could easily still build a trailer in the 1930s! That is how far away from being different they are. ;)

25% of the Country Rioting

[mentil]

The Great Depression never had the unemployment rate go much over 10%. Imagine what would happen to a society with a permanent 25% unemployment rate. Most companies are already one down quarter away from massive layoffs, so don't expect them to absorb the losses. The unemployment would also hit all consumer-oriented businesses, leading to a vicious cycle. We won't cleanly have full employment one day, then a SOTU address saying "Good news, you don't have to come in to work tomorrow. ALL jobs have been fully automated." There'll be a loooong transition period where people are either twiddling their thumbs with no idea what to do with the structurally unemployed, or saying 'fuck you I got mine.' And that transition is going to be 'interesting times' as the Chinese would say.

Re:25% of the Country Rioting

[plopez]

More like 25%. 13 % under Reagan. Both U3 numbers. Obama hit about 10%. U3 unemployment. So think about Regan and Obama's unemployment rates and then double them if 25% of the workforce is displaced.

Re:25% of the Country Rioting

[mentil]

Guess I misremembered. Swear I read a Slashdot headline several years ago where economists found the unemployment rate would never go above X% because at that point people will be willing to take any job in exchange for food. The Great Depression was used as an example of this. Of course I can't find it now.

Re:25% of the Country Rioting

[Riceballsan]

It's worse than that... "Much of the labor will be "enhanced rather than replaced by machines". Does no one think what that means. When a job is enhanced, that means it can be done in less time, and with less people. So the position isn't struck from the roster of the company, but instead of 10 full time educated workers that sometimes need to work overtime, it can be done with 2 minimum wage workers.

Re:25% of the Country Rioting

[drinkypoo]

More like 25%. 13 % under Reagan. Both U3 numbers. Obama hit about 10%. U3 unemployment.

Taking the U3 seriously is a sign of mental illness. At least use U6.

Guess

[Impy the Impiuos Imp]

This includes sex workers, if we can keep the neo-sex-luddite SJWs from dictating sex bots shall be illegal. They will be worse than the remants of the religious.

A prediction. Write it down.

Re:Guess

[Aighearach]

Wait, what? You thought SJWs were a type of puritan?!

LMFAO When you make it to the surface, a word of caution: Those aren't mutants, tattoos, colored hair, and body piercings are simply normal fashion now. Also, those are the SJWs.

and of those 25%

[FudRucker]

they will need to double the numbers of employees to do maintenance and repair on robotics and AI for those automated tasks

Re:and of those 25%

[fluffernutter]

If that were truly the case, then the financials wouldn't work for getting the machines in the first place; they might as well have kept the person and saved money. In most cases, automation is used because it takes 1/50th the people to maintain it than were there in the first place.

actually 100%

[plopez]

If you lose your job it is 100% job loss. Never forget that.

Re:actually 100%

[h33t l4x0r]

Says the lucky guy with only one job.

Now the hard question.

[Gravis Zero]

When the time comes and an entire sector is laid off in a period of a year or so, are we going to stick to our "free market" guns and leave these people to fend for themselves (hello crime wave) or will we do the right thing and implement something like UBI?

Re:Now the hard question.

[liquid_schwartz]

When the time comes and an entire sector is laid off in a period of a year or so, are we going to stick to our "free market" guns and leave these people to fend for themselves (hello crime wave) or will we do the right thing and implement something like UBI?

There are other options. We could limit legal immigration, stop illegal immigration, and retrain the workers. If only we had more politicians who were actually on the citizens side. I hope the US gets a working class party before things get too out of hand.

Re:Now the hard question.

[Patent Lover]

Sounds like a libertarian's dream to me. Not the dream they were hoping for, but a dream nonetheless.

Re:Now the hard question.

[h33t l4x0r]

We could limit legal immigration, stop illegal immigration..

Can you explain why you're worried about immigrants taking your jobs, but fine with robots taking them?
What's your position on immigrant robots?

Re:Now the hard question.

[Gravis Zero]

There are other options. We could limit legal immigration, stop illegal immigration, and retrain the workers.

Sorry but that wouldn't even help help because of the very short time period on which automation will wipe out a sector. As for retraining workers, what do you think training millions of people to do various trades is going to do to the wages of each trade?

Also, if we're supplying people with state-funded training then why aren't we supplying state-funded college education?

Any way it goes, they are the "welfare queen"s that they despise... they just don't know it.

Re:Now the hard question.

[djinn6]

Immigrants cost society money. Robots cost their owners money.

One gets needs subsidized medicine, the other doesn't. One has children that need to be taught by teachers. The other only needs a "dd" command. One goes on welfare when they're no longer able to work. The other goes into the recycling bin.

If you want decent UBI, or any kind of social program really, those poor illegal immigrants will have to stop coming.

Re:Now the hard question.

[HiThere]

The problem with UBI, depending on implementation, is that if 25% of jobs become redundant, that other 75% will be upset if they have to pay for those 25%.

A real UBI should be granted to ALL citizens, or at least all adult citizens. If you earn more, it should not be undercut. The problem here is that lots of things, say housing, will suddenly experiences price rises. So just about nobody benefits, except those in restricted provider positions. How this should be solved is questionable. Wealth concentrations enable things like SpaceX, but they also enable lots of socially destructive things, like land speculation. (I'm less worried about those who spend it foolishly than about those who use to to extend control.)

Re:Now the hard question.

[AHuxley]

With the UBI who in the state/federally is going to pay for 100% more people to get full support all the time?
Better to just give the 25% some income while they are all looking for work. A few might find a new job.
Then move them to some new early retirement benefit.
Paying for 25% of working people for a short time is a lot more tax friendly than 100% of every generation of working population getting an instant UBI all the time.

Re:Now the hard question.

[AHuxley]

The robots dont need years of a tax payer funded education and free city health care.
The robots dont request chain migration for generations of more robots.

Re:Now the hard question.

[The Evil Atheist]

That's not an argument against immigrants. That's just arguments against people in general.

Re:Now the hard question.

[h33t l4x0r]

Most immigrants are hard-working folks who just want a chance to provide for their families. And you didn't exactly come over on the Mayflower, pal.

Re:Now the hard question.

[Anonymous Coward]

Immigration is one of the biggest drivers of economic growth. Every capitalist knows this. Here's a short writeup at the American Enterprise Institute, a conservative thinktank: http://www.aei.org/publication/how-immigration-boosts-american-economic-growth-and-innovation/

Re:Now the hard question.

[djinn6]

There's nothing wrong with them wanting that, but if you let them come and immediately give them handouts, that's what they'll come for and we'll go bankrupt trying to take care of them all.

Your choices are:
1. Don't let them come
2. Let them come, but don't have social services for anyone
3. Let them come, but keep them as 2nd class citizens who don't receive social services
4. Let in only the best people, those who don't need social services

The early parts of US history was dominated by #2. Then as we started providing social services, it turned into #4. Both are viable options.

But what the liberals are proposing now simply won't work. You can't have everyone come and then also take really good care of them. We can't afford to give the entire world a first world living standard, at least not for long.

Re:Now the hard question.

[mapkinase]

Robots do not join MS13, they only join Skynet.

Re:Now the hard question.

[h33t l4x0r]

The robots dont need years of a tax payer funded education and free city health care.

The tech cos that make them took a bunch of federal dollars, hate to break that to ya.

The robots dont request chain migration for generations of more robots.

A robot can make thousands more robots. Better put back on that white hood, buddy. You got work to do.

Re:Now the hard question.

[Gravis Zero]

The problem with UBI, depending on implementation, is that if 25% of jobs become redundant, that other 75% will be upset if they have to pay for those 25%.

Oh please, we already have most of the blue states carrying most of the red states. This will just be more pronounced.

How this should be solved is questionable.

it's why I said, "something like UBI". Universal Basic Assets is similar idea but it doesn't have the recognition UBI does.

Re:Now the hard question.

[Gravis Zero]

Better to just give the 25% some income while they are all looking for work.

Yeah, we have that, it's called collecting an unemployment check.

A few might find a new job.

And the rest can die in a ditch, right? -_-

Then move them to some new early retirement benefit.

"early retirement" is a euphemism for murdering people.

Paying for 25% of working people for a short time is a lot more tax friendly than 100% of every generation of working population getting an instant UBI all the time.

Pff! "tax friendly". You care more for your money than your fellow citizens.

Re:Now the hard question.

[liquid_schwartz]

We could limit legal immigration, stop illegal immigration..

Can you explain why you're worried about immigrants taking your jobs, but fine with robots taking them? What's your position on immigrant robots?

I didn't say I was fine with the robots, however I see that as being more beyond our control. Moreover robots could be used for the greater good, immigration can't. To help you understand the scale of mass immigration I'd suggest this video: https://www.youtube.com/watch?... In terms of why I suggested limiting immigration - it's simply that if robots create a shortage of jobs then limiting immigrants helps the citizens get the jobs that remain. Limited skilled immigration is a good thing. Mass immigration most certainly is not.

Re:Now the hard question.

[liquid_schwartz]

There's nothing wrong with them wanting that, but if you let them come and immediately give them handouts, that's what they'll come for and we'll go bankrupt trying to take care of them all.

Where's you're evidence for this? Immigrants pay taxes, just like everyone else. Immigrants are also less likely to take advantage of public services. If you look at something like social security, illegal immigrants are a boon for the system, because they are contributing to something that they will never be able to take advantage of. There are many, many economic benefits from immigration, and if we didn't have lots of immigration, we would be facing the same problems Japan is having. If you look at the GDP projections over the next few decades, the US is going to struggle to maintain economic dominance, just because population is exploding in many other countries, and GDP growth is tightly related to population. When you think about it, there are only two ways to increase GDP: add more people, or increase your per-capita GDP. China, India, and many countries in Africa have massive population growth, and are simultaneously improving per-capita GDP rapidly. Clamping down on immigration, which has made the US what it is, is going to slow our economic growth at a time when many other places are accelerating and becoming far more competitive.

But what the liberals are proposing now simply won't work. You can't have everyone come and then also take really good care of them.

Liberals aren't generally proposing this - they've voted regularly in favor of increased funding for border security.

You're conflating immigrants with illegal immigrants. Legal immigrants as a group are less likely to take advantage of us. Illegal immigrants, and anchor babies, are far more likely to take advantage. By mixing the positive stats of legal immigrants with the negative stats of illegal immigrants you're painting a rosy picture that isn't true. Indeed it's this type of shady use of language that keeps many Americans from understanding the true impacts of illegal immigration. Citations: https://cis.org/Report/63-NonC... https://www.washingtonexaminer...

Re:Now the hard question.

[drinkypoo]

You're conflating immigrants with illegal immigrants. Legal immigrants as a group are less likely to take advantage of us. Illegal immigrants, and anchor babies, are far more likely to take advantage.

It's been shown time and again that illegal immigrants 1) contribute more than they consume and 2) commit less crimes than the background rate, because they are trying to keep a low profile. So even if they are more likely to take advantage, they're still a net positive. We also take actions that cause these people to come here, like when we interfere with their democratic processes, or with our War On Some Drugs. If you look at where people are coming from when they come here, it's overwhelmingly from countries we've shit upon in a well-documented fashion. If you don't want refugees here, then stop supporting politicians who will create them.

Re:Now the hard question.

[liquid_schwartz]

There are other options. We could limit legal immigration, stop illegal immigration, and retrain the workers.

Sorry but that wouldn't even help help because of the very short time period on which automation will wipe out a sector. As for retraining workers, what do you think training millions of people to do various trades is going to do to the wages of each trade?

Also, if we're supplying people with state-funded training then why aren't we supplying state-funded college education?

Any way it goes, they are the "welfare queen"s that they despise... they just don't know it.

It most definitely would 'help help'. It just wouldn't mitigate things fully. Retraining is cheaper than welfare. Retraining could be weighed against how much they have paid in taxes if that makes you feel better. Free college is a bad idea as many people just want to avoid joining the working world - college is fun. Having the meter running helps keep that impulse in check.

Re:Now the hard question.

[Gravis Zero]

Free college is a bad idea as many people just want to avoid joining the working world - college is fun.

I don't know where you get the idea that lots of people want to avoid joining the workforce. People don't get PhDs because college is fun. If you think college is fun then you weren't focused on education. Clearly the sports and frats need to go in order to get people like you to focus.

Re:Now the hard question.

[liquid_schwartz]

You're conflating immigrants with illegal immigrants. Legal immigrants as a group are less likely to take advantage of us. Illegal immigrants, and anchor babies, are far more likely to take advantage.

It's been shown time and again that illegal immigrants 1) contribute more than they consume and 2) commit less crimes than the background rate, because they are trying to keep a low profile. So even if they are more likely to take advantage, they're still a net positive. We also take actions that cause these people to come here, like when we interfere with their democratic processes, or with our War On Some Drugs. If you look at where people are coming from when they come here, it's overwhelmingly from countries we've shit upon in a well-documented fashion. If you don't want refugees here, then stop supporting politicians who will create them.

Since it's been shown time and time again how about some citations? I usually include some, and you haven't refuted the ones I listed.

We also take actions that cause these people to come here, like when we interfere with their democratic processes, or with our War On Some Drugs. If you look at where people are coming from when they come here, it's overwhelmingly from countries we've shit upon in a well-documented fashion.

If you feel bad about this, as some do, I propose that *you* take them in. That's an immigration process I could get behind. If you have skin in the game I suspect that you won't be quite so cavalier. Seriously though if you think that letting in immigrants even makes a dent then you need to see the scale o what we're talking about. Here's a quick video that will give you that perspective https://www.youtube.com/watch?...

As far as not messing up other countries, I'd be perfectly happy to see us pull back everything world wide. Doing it over a well defined time table would help prevent a power vacuum but if we started this year and finished in say 10 years I'd be thoroughly happy.

Re:Now the hard question.

[liquid_schwartz]

Free college is a bad idea as many people just want to avoid joining the working world - college is fun.

I don't know where you get the idea that lots of people want to avoid joining the workforce. People don't get PhDs because college is fun. If you think college is fun then you weren't focused on education. Clearly the sports and frats need to go in order to get people like you to focus.

"People like me". That's funny. I didn't play sports, I didn't join a frat. I did enjoy the experience though. I guess that somehow makes me and "people like me" bad people.

Re:Now the hard question.

[djinn6]

I guess you're not aware of California trying to give medicare to illegal immigrants.

Re:Now the hard question.

[djinn6]

You might have a reasonable opinion and you might consider yourself a liberal, as I once did, but you don't represent the Democrats, especially not those in power. "Abolish ICE" is not a Republican campaign and Texas is not home to 40+ sanctuary cities. The elected Democrats have to appease 38% of California that are Latino, even if what they want will eventually bankrupt the state.

Re:Now the hard question.

[werepants]

"Abolish ICE" is not a Republican campaign and Texas is not home to 40+ sanctuary cities. The elected Democrats have to appease 38% of California that are Latino, even if what they want will eventually bankrupt the state.

ICE is a pretty recent phenomenon, to be honest - only created in 2003. Somehow we managed to deal with immigration for 200+ years without it, and I don't think it's unreasonable to discuss if it is in fact performing a critical role. That said, it's a minority of liberals that want to abolish it completely.

All states should hope to have California's budget "problems". They consistently contribute far more to the federal budget than they receive, and all on their own have a larger economy than most countries in the world, outranking the entire UK. Somehow, all those illegal immigrants and sanctuary cities haven't managed to bankrupt them.

I think you make reasonable contentions, but they seem to be based on assumptions that don't match the actual numbers.

Re:Now the hard question.

[djinn6]

So you don't want open borders, except you oppose the very agency prevents it from becoming one? Also, ICE's job was previously done by U.S. Customs Service and the Immigration and Naturalization Service. It's not created out of thin air.

As for the budget, a fool can balance it when their revenue is growing by billions a year. The problems only appear when you hit a recession. What are you going to do then? Tell all the illegal immigrants to GTFO?

Great Depression and revolution

[sphealey]

At the peak of the Great Depression the US came close to a red revolution. Unemployment at that point was 25%. "Only" 25%.

Billionaires and Investment Capitalists

[WillAffleckUW]

Unfortunately, 100 percent of occupations held by billionaires and investment capitalists are subject to automation, but they could always retrain as guillotine blade sharpeners and other useful occupations, if need be.

Re:Billionaires and Investment Capitalists

[LynnwoodRooster]

So Bill Gates, Jeff Bezos, Warren Buffet could be automated away with no impact?

Re:Billionaires and Investment Capitalists

[drinkypoo]

So Bill Gates, Jeff Bezos, Warren Buffet could be automated away with no impact?

If you replaced Bill Gates with a small shell script, it would be a substantial improvement.

Actually a big number

[DaMattster]

25% is actually a very large number when you look at the population. That's 1/4 jobs are at risk of automation. This article is full of shit. If 1/4 jobs were to disappear the economic fallout would be staggering.

Good News!

[fuzzyfuzzyfungus]

Luckily, the other 75% can be kept in fear of losing their jobs to members of the unemployed 25% who will work for not-starvation; so we needn't let this disappointment interfere with reducing the cost of human resources. Progress!

Not too bad!

[danbuter]

It's only ~37,500,000 people who will lose their jobs! /s

Re:I lost my job yesterday for a different reason

[danbuter]

Don't. Just leave. Or better yet, "train" him to royally screw up.

Over what time frame? Bureacracies.

[aberglas]

A critical detail.

Over the next ten years, just normal automation. Factory jobs, some agricultural work. A few percent each year.

In about ten years time self driving vehicles will become mainstream. And many other easily automated tasks. At that point 25% may become reasonable over the next few decades.

In some 50..100 years after that computers will be able to program themselves. They will then no longer need humans at all.

But the good news is that bureaucracies will continue to grow regardless of any attempt at automation. 50 years ago there were typing pools and clerks balancing ledgers by hand. All those jobs gone but bureaucracies just grow and grow. So soon, everyone will be a bureaucrat whose job it is to regulate everybody else. As predicted by Parkinson long ago.

https://www.economist.com/news...

(A classic, well worth reading.)

Parkinson's law

[aberglas]

Un paywalled version
http://www.berglas.org/Article...

How It's Made

[perpetualinput]

If you watch it you would have a realistic idea of how automated the various industries already are... Failing to know vs wanting to know, makes a difference. Kinda like with pi on a different level, yeah?

Re:How It's Made

[turp182]

That show is both terrifying and fascinating.

It's terrifying that in some plants the only thing people do is box shit up or pour raw ingredients into the machines. Not a glamorous thing to talk about when trying to meet people at the bar (bad pick up line: "Then I pour 200 pounds of sugar into a hopper and press a button")...

Fascinating due to the mechanical engineering. And with 3D part manufacturing (https://www.makepartsfast.com) even machine shop jobs for one-off parts are on the line. CAD jobs are safe I would contend.

I love the episodes where people actually make things, specifically instruments (the Gibson episode was awesome, but I play a hand-made Seagull from Canada...).

just do it another way

[RhettLivingston]

I think the number is too low because they considered getting machines to do the human's job instead of adapting the job to the machine or at least meet halfway. Often, you can't get the machine to do the same thing, but you can get the machine to do something else that can accomplish the same end in a way that a human couldn't.

The machines don't have to take over our jobs to eliminate them. For example, in some fields the numbers of scientists needed are decreasing rapidly because a few scientists are able to leverage AI to more quickly find a solution to a problem by testing numerous combinations of something instead of trying to think their way to a solution - much like the way the first Chess computers beat people at Chess. The AI is not doing the same thing, but remarkable results are being achieved.

I've said this before

[rsilvergun]

but I wish I could stop saying it: the Luddites were real people put out of work by technology in a society where if you don't work you don't eat.

It took a _long_ time for technology to catch up. The Luddites died jobless. Their kids did too. There were decades of unemployment and social strife following the industrial revolution until the two greatest employment programs in history restored full employment.

The tech that came out of the war (and the mad rush to rebuild the world) kept us going for this long. But the rich don't want another World War. It's their stuff, why would they let us break it? I remember when Pakistan looked the other way when some of their folks attacked India's capital. I was ready for WWIII. Didn't happen. India sucked it down because it would have been bad for business.

This means we're not gonna have wars save us this time. The next industrial revolution is upon us in the form of computerized automation. I've never _once_ heard a credible explanation of exactly what jobs we're all gonna retrain for. That's because there isn't one. You say that it's just that I can't imagine those jobs? That's because by the time they're available I'll be dead and I'll have died dirt poor.

It doesn't have to be that way. We can learn from history for a change. We know technology unemployment is coming. Now's the time to do something about it while we still have some economic power.

Re:I've said this before

[_Sharp'r_]

The actual Luddites went on a spree of violence and were either killed, or arrested and transported to Australia. The Luddite resistance only existed for a couple of years.

In terms of the general population, England's employment issues of the time were more related to lots of recently discharged soldiers and sailors (300K all at once) at the same time as a big population boom (50% increase between 1801 and 1831). By 1834, many factories in England were complaining about a shortage of available labor, although some of that was because it was illegal for the poor to move to a different parish where there might be work.

Agriculture used to employ 70-80% of the population, way more than the 25% being described here.. Then automation and technological progress came and people, including those with no skills but manual labor at the time, moved out of most rural areas to more beneficial jobs.

Ironically, you're complaining about capital investment and at the same time extolling the benefits of capital investment in manufacturing and automation to produce more during the world wars.

"We can learn from history for a change. We know technology unemployment is coming." is mistaken on several grounds, not the least of which is that historically, the best remedy for dealing with technological change has been to do the exact opposite of what most modern Luddites suggest, namely allow markets to adjust as quickly as possible to minimize any transitional problems. Instead, the typical suggestion is to come up with ways to prevent, delay or resist the needed adjustments to different production patterns.

Re:I've said this before

[hazardPPP]

This means we're not gonna have wars save us this time. The next industrial revolution is upon us in the form of computerized automation. I've never _once_ heard a credible explanation of exactly what jobs we're all gonna retrain for. That's because there isn't one. You say that it's just that I can't imagine those jobs? That's because by the time they're available I'll be dead and I'll have died dirt poor.

We were saved by two humongous wars a result of which about 100 million people died (directly during the war...so we can make this number bigger)? Guess who did the vast majority of dying? Rich folks? No. The poor and the middle classes...

Yes, I'm saying that you just can't imagine those jobs - and no, not because you'll be dead by the time they're available (unless you are already an octogenerian), but simply because you can't imagine them. Some of the jobs which are widespread today (social media coordinator for example) were unimaginable by most people just 30 years ago.

Re:I've said this before

[apoc.famine]

Some of the jobs which are widespread today (social media coordinator for example) were unimaginable by most people just 30 years ago.

You just don't get it, do you?

It doesn't matter what the jobs of the future are, or whether or not we can imagine them. That's immaterial to the problem.

The problem is that between AI and automation, robots will do them better, faster, and cheaper. It doesn't matter what jobs we come up with when it doesn't make any fiscal sense to hire humans to do them.

Re:I've said this before

[hazardPPP]

Some of the jobs which are widespread today (social media coordinator for example) were unimaginable by most people just 30 years ago.

You just don't get it, do you?

It doesn't matter what the jobs of the future are, or whether or not we can imagine them. That's immaterial to the problem.

The problem is that between AI and automation, robots will do them better, faster, and cheaper. It doesn't matter what jobs we come up with when it doesn't make any fiscal sense to hire humans to do them.

I'm afraid I get it a lot better than you do.

Of course it matters what the jobs of the future are - it's far from immaterial. If this is not clear to you, then you are the one who truly does not get it.

Robots and other machines and "AI" have been putting people out of work and automating away human jobs for about a century (at least) now. Yet, we do not have mass unemployment. The reason you think it will be vastly different in the future is beacause 1) you are believing the AI hype and overextrapolating present abilities of AI and robotics into the future (things will not improve as much and as fast as you obviously seem to think) and 2) you have a lack of imagination. So do I - but at least I admit it.

Now even if we assume you are correct - which I strongly believe you are not, but let's put that aside for the moment - and that some point in the medium term future, AI and robots are going to put something like 95% of people out of work. I say - great! That's a precondition for utopia...that kind of massive disassociation of wealth (and the ability to gain wealth) from labour will foster a social revolution. Remember, the industrial revolution was not only technological - it brought with it significant social change as people dealt with the consequences of technological advancement. This went hand in hand with radical economic change: remember, there was a time when most people didn't survive by earning a wage, when very little money circulated, and when most people - even most of the very rich - did not count their wealth primarily in how much currency they had in their possession. If 95% of the people cannot work for a living, the economy will collapse (who will spend the money and buy the automatically produced things and automatically produced services?). The means of production will then be in some way socialized (either nationalized or heavily taxed) leading to a radical social re-alignment. If we can produce almost everything we need for a high standard of living without the need for a large input of human labour, the economy becomes mostly about resource management. Now, after having had such a realignment or revolution, with all basic needs cheaply provided for by automation, imagine all the things we could get - indeed pay - people to do....things that today do not make "fiscal sense". Like we could pick up all the pieces of trash on this planet. Or do really cool sci-fi type experiements and projects under the sea or in outer space...raise the educational level of everyone...produce a lot more art...etc. Which gets us back to "there will be ne more jobs" fallacy. Welcome to Start Trek, Ensign Crusher.

There's no need to panic!

[cas2000]

Soylent Corporation is always "hiring".

How is Family Physician not top of the list

[FeelGood314]

This is a job that is 99% matching symptoms to diseases and then looking up the cure. And the good thing about AI would be that they would actually learn from their mistakes instead of burying them. (I'm only half joking, the feed back mechanism for many doctors is limited or too slow to result in significant learning)

cascading effect of automation

[SethJohnson]

The thing to consider about displacing swaths of the workforce is that they'll shift to other occupations. These will often be the jobs that are too complicated to automate relative to the payoff opportunity. Sex worker is something of a safety-net occupation and when people can't find other employment after they have been displaced by automation, many will turn to this type of gig economy work. The customers (employers) will benefit from an abundance of willing workers, which will push prices down. Overall, this dimension of capitalism (i.e. automation) will result in an increase in worker exploitation to the benefit of the wealthy.

Disaster

[Sqreater]

Only 25%?? That is huge. And what nobody seems to understand is that AI and machines start at the low end of jobs held by the lower end of the IQ curve and move up the IQ curve. Those jobs to be automated will displace people who are not going to be able to "educate up" to higher level jobs because they have not the capacity. Those lower level jobs have always provided income and daily work for those who are not capable of doing higher work. Thus, automation is a disaster for society as it will create an underclass of floating workers who will become resentful outsiders prone to violence.

Food service jobs are not at-risk

[cordovaCon83]

Cooking is not an exact science and never will be. Even the lowliest fast food worker is still visually inspecting everything he/she so autonomously cooks to make sure the meat is cooked all the way through, that the lettuce hasn't turned, etc. Those food service workers with a bit more freedom are playing with recipes and checking for variance in the cooking process. We'll still have servers delivering food to tables, visually inspecting the food one last time, then checking back regularly that the dining experience. We already pay those in the food service very low rates, and if a food service worker gets sick then they are easily replaced on the line by a cross-trained co-worker. If a robot breaks in an assembly-line fashion kitchen, it's like to bring the whole system down until a very costly technician comes out and repairs the system. Is that going to cost hours, days, or weeks worth of sales? Until robots can experiment with food recipes, tell the difference between good and bad lettuce at a glance, smell expired ingredients, self-repair, and provide friendly interactive experience with humans, don't experience food service jobs to take a major hit from robots any time soon.

Only 25%

[GLMDesigns]

We're going to have a real problem on our hands soon. We don't have a need for uneducated or rote work. And, here's the kicker some people (5% of the population) cannot be trained for higher skilled jobs. As automation continues it will not simply be warehouse jobs, truck driving jobs, taxi drivers, lawn care, street repair, farming positions but also a lot of low level office jobs. Notice how few professional typers there are anymore? That trend will continue.

Unlike 100 years ago we do not simply need a hard workers with a strong back.

This affects everything - from immigration policy, to social security (as we live longer and no longer have back breaking jobs) to everything. And "living wage" arguments don't cut it anymore. Take a look at the food service industry in NYC and other places that increased minimum wage. And UBI (Univeral Basic Income) is not a panacea either. People want to feel needed, useful and most are not self-driven to find satisfaction in the arts.

Re:Only 25%

[drinkypoo]

UBI (Univeral Basic Income) is not a panacea either. People want to feel needed, useful and most are not self-driven to find satisfaction in the arts.

People have been sold a line of bullshit that states that if they're not working a job, they're worthless because they're not serving anyone. It's some puritanical garbage that has been embedded in our culture. If you want to see the logical end of that game, check out Japan. They've got whole jobs, not just job duties, that are pure make-work. Responsibility to the culture leads corporations to inefficiently employ people doing things that don't need to be done. But there is a real energy cost to putting asses in seats, to say nothing of traffic congestion, and that harms everyone else.

Pretty much the whole world works on the same principle, but it's crap. Absent work requirements, we'd take on our own projects. Sphincters would relax long enough for people to even conceive of what they would do with their time, in a way that most cannot even imagine today.

YOU ARE NOT YOUR JOB. YOUR JOB DOES NOT GIVE YOUR LIFE MEANING. Many jobs are actually counterproductive, and the world would be better off if they were not done at all.

Re:Only 25%

[GLMDesigns]

I don't think it's bullshit. And I'm not suggesting make work either. Make work is not the solution.

"What is my purpose in life?" Is a key question people ask themselves; go through periods of angst trying to understand.

We see people who have things simply given to them. (Welfare kids and trust-fund babies) How are they doing? We two groups one with all the money in the world, the other with just enough to survive. Are these people doing well? Are they producing art, literature? Are they happy? Or are they drugging themselves and frittering away their time?

I wish UBI was the answer. It's not. (IMHO)

Job replacement

[Anarchduke]

You know the difference between the AI revolution and the industrial revolution? The AI revolution isn't replacing jobs, its replacing workers. Essentially, you are manufacturing an artificial person who can be instantly taught to replace a worker in any job. We could do the same thing without automation by allowing immigrants to come into the country and immediately firing all currently employed people and replacing them with the hypothetical army of immigrants. Let me repeat. We aren't creating new jobs, we are creating new workers.

Re:Job replacement

[Anarchduke]

So unless you have a doctorate, you're going to be out of a job. I realize someone is going to say, well what about the technicians who watch over the robots? Well, they're just techs, they can be replaced by robots too.

There's a saying among cynics

[rsilvergun]

"The survivors will thank us".

If I can't imagine the job then it's not going to exist on a time frame that matters to me or even my kids. The resulting technology unemployment, OTOH, will exist.

Re:There's a saying among cynics

[hazardPPP]

If I can't imagine the job then it's not going to exist on a time frame that matters to me or even my kids. The resulting technology unemployment, OTOH, will exist.

Are you asserting that you are the most imaginitive person on the planet? While this is possible of course, it's highly unlikely.

As I said, there are lots of jobs today that people couldn't imagine 30 years ago. I don't know how old you are, but I'm sure there are still more jobs that your parents could not imagine at some point in the past, so making that assertion about your kids seems kind of foolish to me.

As for resulting technology unemployment: there will be some, of course. As there has been in the past. Keep in mind that in developed countries, the working age population is set to shrink...

25% HAH!

[messymerry]

Come on you guys, how can you present a static number (25%) for a dynamic process? Perhaps a better representation: 25% within 5 years, 50% within 10 years, 75% within 15 years, and everybody within 20 years... Stupid college profs think they are indispensable. Teaching is a boring repetitive job and AI will quickly come to the conclusion that human teachers are inferior and EXPENSIVE. Just sayin'

You're straw manning

[rsilvergun]

the fact that Luddites were violent has nothing to do with their job losses.

Luddite resistance only existed for a few years because they were brutally oppressed.

The discharged soldiers just show that their economy couldn't absorb an uptick in workers. What do you think will happen when we put 10-25% of a population of 350 million+ out of work.

And employers are _always_ complaining about a shortage of employees. Nothing new there.

I'm not saying we should resist these patterns (you're putting words in my mouth). If you want to know what I'm driving at I'm saying we need a society where you can eat even if there's no useful work for you to do. i.e. socialism, UBI or whatever you want to call it.

"Only" 25%?

[whitroth]

Unemployment during the Great Depression was 25%-32%. And, of course, all those folks put out of work won't be buying much, so everyone *depending* on them is going under.

Y'know, sort of like what we're seeing right now, with the government shutdown.