Uber self-driving car in Pittsburgh

[Simon_Jester]

You could equally well turn that around and say that Japan has a collapsing working-age population which is partly alleviated by automation...

[Guardsman Bass]

The big difference is just that it's not obvious what kind of jobs could exist; there is no big growth sector to replace the current set of jobs.

It's never that obvious in advance what's going to replace the jobs lost, which is why you always see predictions of technology destroying jobs and causing mass unemployment over and over again (particularly in downturns, like the recent recession, the early 1990s recession, the minor downturn in the late 1950s/early 1960s, and so forth). Try telling someone rendered unemployed in the late 19th century about the jobs in the early 20th century, for example.

This is why it's mostly pointless to debate this topic. All I can do is point out that despite repeated predictions, it hasn't happened. Stas provided some interesting stuff on Japan, but Japan's also got extremely low unemployment regardless of that and a rapidly aging population that's quickly shrinking the work-force.

[K. A. Pital]

Try telling someone rendered unemployed in the late 19th century about the jobs in the early 20th century, for example.

indeed a pointless exercise. Millions of weavers in India starved to death for the glory of empire before the century even begun. They couldn't care less what kind of jobs would exist in the 20th century.

[Broomstick]

Sometimes you can make a prediction - when the automobile displaced the horse and buggy you could at least predict jobs in and servicing the auto industry. Sometimes you can't. Also, just because in the past new areas of industry opened up does not mean that will always apply in the future. That uncertainty is part of what feeds anxiety about the future to the masses.

[Alferd Packer]

To be a bit of a contrarian, I think the inevitability is what causes more anxiety than anything else. We know the mass automation is coming--we just don't know when, and how fast. But we do know that there's absolutely nothing we can do to stop it. You can't outlaw the research, and once the automation is in place in an industry, it will always be cheaper to replace damaged or vandalized equipment than it will be to hire humans to replace them.

But really, that boils down to uncertainty about the future, so I guess I'm just splitting hairs.

Anyway, it could be my generation that winds up staring into the abyss, or my infant niece's generation, or even her children's generation. I don't know enough about how close we are to any of the major milestones in developing AGI. Like how many years are we away from passing the Turing Test?

[Guardsman Bass]

Try telling someone rendered unemployed in the late 19th century about the jobs in the early 20th century, for example.

indeed a pointless exercise. Millions of weavers in India starved to death for the glory of empire before the century even begun. They couldn't care less what kind of jobs would exist in the 20th century.

I hear mustard would be a very tasty condiment on that red herring of yours.

Sometimes you can make a prediction - when the automobile displaced the horse and buggy you could at least predict jobs in and servicing the auto industry. Sometimes you can't. Also, just because in the past new areas of industry opened up does not mean that will always apply in the future. That uncertainty is part of what feeds anxiety about the future to the masses.

It doesn't, but when the same argument cycles up with downturns for the past 200 years, I think you have a strong obligation to at least be skeptical of it just happening to be right this time.

[Broomstick]

Skeptical is not the same as dismissing it out of hand, which you seem to be on the edge of doing. Even in a best case scenario there will be losers, and under capitalism "losing" in this context can be extremely painful.

[K. A. Pital]

The problem with crude unemployment rates is that we never evaluate the level of work-time-reduction necessary to maintain low unemployment over the decades, and the underlying shift of full-time employed to part-time employed is thereby hidden.



Technology-induced unemployment could be actually seen as a time-conserving benefit if it would allow to shift to a 4-day or 3-day working week instead of today's 5-day WW.

The problem is that hourly wages are not rising as fast as they used to in 1945-1970, so a reduction in time worked would also mean a reduction in income for workers. This is also why, I guess, the trend towards working hours reduction has stopped and the fall from 70+ and 60 hour work week in the late XIX century to a 40-hour work week in the mid-XX century has not continued until now.

[Broomstick]

Here's the thing:

When I started out in office jobs manual typewriters were still the norm, you used carbon paper to make copies, dictation was taken with a pen and steno pad, and accounting was done on paper with pencils using a technique called "double entry bookkeeping" (which is NOT the same as keeping two sets of books). There were adding machines or, if you were really lucky, an electronic calculator. If you were really lucky you didn't need to learn reverse Polish notation in order to use it.

Then technology advanced. The Nimble Worker adapted - electric typewriter definitely improved productivity, xerox machines made copying much easier (when it didn't jam or otherwise malfunction - troubleshooting skills needed to be acquired as well), little tape recorders gave an option for dictation and improved things by allowing the dictator to record any time, not just when a transcriber was available, and the transcriber to have an original oral record to review, thereby eliminating errors in writing things down. Some couldn't adapt, but most could.

Technology advanced again. There was this interesting new box that was sort of like a calculator, but more so, call a microcomputer and one of the new and exciting features was VisiCalc, which very quickly supplanted the old pen-and-pencil double-entry bookkeeping (much easier to use, more accurate, easier to make corrections). Also, the first word processors appeared, where you could type, proofread, and make corrections all before you actually produced a document so you have perfect originals every time (well, at least in theory). Those microcomputers with the VisiCalc and word crunchers also eliminated the need for that reverse Polish notation. Again, the Nimble Worker adapted. With the efficiencies gained a few jobs were shed, but there were plenty of other industries to take up the slack, including things like copier repairman, computer servicing, and the like.

And technology advanced again... only now any need for any form of dictation has been supplanted by speech-to-text software. The Nimble Worker can't adapt, there is simply no more role for N.W. here. No more need for speedy touch-typing, either. Another skill rendered obsolete, with no obvious replacement. Less need for mail delivery, as information travels electronically and the only time anything physical is actually sent is when an actual object needs to be transported - fewer delivery people. Less need for copiers when so much exists as data rather than sheets of paper.

That's the difference here - before, technology was giving workers better and better tools. Now, in at least some cases, the technology has replaced the worker in part or in whole. And that's a potential problem.

[Guardsman Bass]

Yet it hasn't led to mass unemployment - unemployment still follows the business cycle, and is down below 5% again.

That's always been there, that individual sectors of the economy can disappear or drastically shrink in size because of technological change. Hell, that's happened - and is happening - to manufacturing employment. But the "robots take our jobs" argument is an argument that technology is causing overall unemployment in the economy, and that's not happening (at least not in the US).

Skeptical is not the same as dismissing it out of hand, which you seem to be on the edge of doing. Even in a best case scenario there will be losers, and under capitalism "losing" in this context can be extremely painful.

I never argued otherwise. Parts of the Midwest are proof of that.

[Broomstick]

Yet it hasn't led to mass unemployment - unemployment still follows the business cycle, and is down below 5% again.

It did in 2007.

And while many of those people did find new jobs, they were at signficiantly reduced pay. For example, I went from making $50k+ a year with great benefits, perks, and bonuses to a bare minimum $11k - and I got that only after several years of looking for permanent work.

It's not just complete unemployment to be feared - falling into grinding poverty sucks, too. I went from solid middle class to under the poverty line. Why? Because my entire skill set that earned me the higher income has been rendered obsolete by automation or supplanted by out-sourcing. So sure, I wound up somewhere else and now I'm living in drastically reduced circumstances and just barely hanging on by my teeth.

So, OK, tens of millions of people got another job - that does not make everything OK, which is one of the pitfalls of relying solely on statistics.

Great, I have a job. Do you realize that most re-training and apprenticeship programs have age limits? I've been deemed to old for re-training yet I am expected to get a new job/career on my own between now and retirement. Or be homeless. Then there's the high cost of education in this country, the private college scam/scandal, and the difficulty and expense of acquiring a skill for a higher paying career.

The system is set up for a certain number of unemployed. When you get substantially more, like during the Great Recession, it chokes which is why even in Republican states like mine you had a quiet expansion of public aid programs - all of which were cut back as soon politically feasible. The unemployed are coerced in the first job that comes along, whether well suited to the person or not, whether it pays a living wage or not, whether it's viable long term or not. Then everyone pats themselves on the back and says problem solved.

No, no it isn't - taking the former middle class and slamming them into long-term poverty is not a good strategy. Telling people to get college degrees in a system that saddles them with crushing debt then only having low-wage jobs available is not a good strategy. It's not the same as being outright unemployed, but to some extent it's just papering over the problem.

[Esquire]

This. The problem isn't 'unemployment numbers are high;' that's a symptom of the actual problem, which is 'people don't have enough money for a decent quality of life. What addresses the first may not necessarily fix the second.

[Simon_Jester]

Yeah.

The problem is that we're running out of things a person can learn to do quickly, that they can also do well enough to 'earn' all the resources and services that we collectively call a First World standard of living.

Historically, people shifted between categories of labor in response to changing technology, but the categories at least existed and were comprehensible to people prior to the economic change. If you told a rural farmer of 1780 that his great-grandson would be working in a factory making chairs, he'd at least comprehend what that was. If you further told him that his great-to-the-eighth grandson would be working in an office as a clerk, or cutting people's hair, or helping a large corporation sell products... these are all things that existed in some form. The specific details of the jobs have changed enormously as technology advanced, but you could sort of imagine someone reasonably imagining "well, right now everyone works to grow food. If growing food becomes easy, people will start making other things. If making things becomes easy, people will work on organizing, selling, and preparing the things, and on aiding and serving each other in various ways."

Automation in the modern software-oriented sense is different because it replaces judgment. Even when industrial machinery replaced jobs in the past, it couldn't replace judgment- it could swing a hammer repeatedly with designated force, but it couldn't tell you where to swing the hammer to accomplish a job that hadn't been designed into the machine from advance.

Over the last thirty years or so, many jobs that used to require judgment and knowledge (e.g. travel agents) have simply been automated out of existence. And we fully expect this process to continue (driving is an excellent example of a job that requires good judgment). If the average worker's judgment is no longer good enough judgment to outperform a machine at comparable tasks, then who's going to pay them more than a pittance? There will be a few jobs for which a machine to do the job would be 'too expensive' for a long time, and people of merely average judgment may be able to retain those jobs.

But for a lot of people, the choices will be "spend the better part of a decade training to do something that hopefully won't be automated in the next thirty years, honing expert judgment that it may be hard or impractical to replace with a machine...

...Or just plain giving up.

And the first plan requires investment capital. People usually don't have that if they''ve just lost their job because a machine or network of computers can do it for a quarter of the money at an acceptably low failure rate.

So as Esquire alludes to, the basic problem is that while automation will surely keep the total productivity of the economy high, there's no guarantee that it will enable everyone to participate in that economy in a way that a capitalist system deems "useful enough" to justify a living wage. It may well be that by 2050 the fair market value of labor from the average high school graduate is, oh, the equivalent of a few dollars an hour in today's money... Because at any higher rate than that it's cheaper to buy a drone that can do the same thing.

What do we do then?

[K. A. Pital]

Usually in normal countries the first plan does not require any capital expenditure as education is free.

But, of course, the US champions such "agreements" like TTIP and TISA that might wipe out public pensions and public education totally, and just about anything public that is still not privatized.

So the double-whammy, as usual, comes not from the machines alone, but from a capitalist application of the machine. As my bearded old bro once said.

Remove capitalists from power, and you pretty much remove the threat from the machines.

[Simon_Jester]

Usually in normal countries the first plan does not require any capital expenditure as education is free.

Eh, you're right, my expectations have been lowered by living in the country most desperate to avoid the faintest tinges of socialism.

So the double-whammy, as usual, comes not from the machines alone, but from a capitalist application of the machine. As my bearded old bro once said.

Remove capitalists from power, and you pretty much remove the threat from the machines.

Would this be the guy from Trier?

I don't disagree with the thesis that the problem is capitalist application of automation. As I have said before on this site...

Automation affects the relationship between labor, capitalism, and our attitude towards public funding of the needs of the poor and unemployed. Specifically, within the next fifty years it will break at least one of those things, or alter it beyond all recognition.

[FaxModem1]

Recent news shows another benefit for the company. Less chance of litigation due to employee misconduct: Uber driver sexually assaulted a female customer.

Though I'm sure that this individual is the exception, not the rule, it's certainly going to affect how business owners look at human vs machine employees.

[K. A. Pital]

Recent news shows another benefit for the company. Less chance of litigation due to employee misconduct: Uber driver sexually assaulted a female customer.

Though I'm sure that this individual is the exception, not the rule, it's certainly going to affect how business owners look at human vs machine employees.

I think the problem is that instead of creating a humane environment where skilled professionals can get a decent wage, Uber was conceptually created as a race-to-the-bottom, "cheap licenseless taxi" thing. This type of transport already exists in the Third World - it is called a self-employed taxi man, and usually it comes with a multitude of risks from scamming to rape and robbery.

The latest switch from human drivers to automated cars makes Uber's model sustainable by kicking out the human component.

But frankly, the question which must be asked is why such an unethical model of doing business deserves to become sustainable in the first place?

[Simon_Jester]

Recent news shows another benefit for the company. Less chance of litigation due to employee misconduct: Uber driver sexually assaulted a female customer.

Though I'm sure that this individual is the exception, not the rule, it's certainly going to affect how business owners look at human vs machine employees.

Well, as Stas notes, Uber, specifically, has been going out of their way not to run serious background checks on their drivers as far as I can tell. Basically they operate by being a taxi company that ignores taxi company rules (allowing them to lower their fares) and gets away with it by claiming to be a "shmaxi shmompany" or whatever. The convenience of their cell phone apps, and the lower fares they have by ignoring issues like maintenance of 'their' vehicle fleet, insurance, and licensing, mean that people looking to save a quick buck go for them.

[Alyrium Denryle]

Recent news shows another benefit for the company. Less chance of litigation due to employee misconduct: Uber driver sexually assaulted a female customer.

Though I'm sure that this individual is the exception, not the rule, it's certainly going to affect how business owners look at human vs machine employees.

I think the problem is that instead of creating a humane environment where skilled professionals can get a decent wage, Uber was conceptually created as a race-to-the-bottom, "cheap licenseless taxi" thing. This type of transport already exists in the Third World - it is called a self-employed taxi man, and usually it comes with a multitude of risks from scamming to rape and robbery.

The latest switch from human drivers to automated cars makes Uber's model sustainable by kicking out the human component.

But frankly, the question which must be asked is why such an unethical model of doing business deserves to become sustainable in the first place?

Deserves? It does not. However, I am also pretty sure that computer driven cars are an inevitability across the board, and they will probably show up and proceed to dominate in commercial fleets first (probably using an app to summon the cars. Uber's business model is unethical trash with human drivers, but with robots it is pretty much optimal). And they should. In the US alone, 30k people die in automotive accidents every year. 1.3 million per year globally. Yes, you read that correctly. Car crashes kill two to three times more people than fucking Malaria. An additional 20-50 million people are seriously injured. And as much as I adore Raw Shark, if I have the choice between a million people surviving and Raw Shark keeping his job...

The point of failure the vast majority of the time is one or more human drivers. Our reaction times are slower than transistors, we cannot communicate by radio with other drivers to determine intent and position like a computer can, and we dont follow road-optimization rules (like maintaining equal distance between the car in front of us and the car behind, we also like to drive slowly in the fast lane yadayada).

[K. A. Pital]

I know the number of people who die in car accidents.

But I also know the number of people who die in the mines and the number of people who die in other industries.

Should machines take over everything? Were it socialism, my answer would be yes. Under guaranteed full employment, society would actively support the people whose jobs are gone and put them in new positions in a planned fashion

Under capitalism, I say... Yes, but only to create unbearable horrible conditions for a social explosion that would wipe out this ugly plutocratic social order once and for all.

[Esquire]

So automation as socioeconomic chemotherapy, then? Extremely negative short-term effects for, one hopes, a better long-term prognosis? I'd want to be extremely sure of the post-revolution results before signing on.

[U.P. Cinnabar]

Under capitalism, I say... Yes, but only to create unbearable horrible conditions for a social explosion that would wipe out this ugly plutocratic social order once and for all.

Assuming said glorious people's revolution runs on rails, and doesn't involve the burgeoisie co-opting it for their own ends, or a new class of burgeois overlords taking up from where the old bosses left off.

But, you're right. The only way it would work in a capitalist society is to effect a fundamental(and radical) change in that society, razing it to its foundations, and re-building. I'm just not sure(as I once was in angrier days)that violent revolution will accomplish that, or play right into the hands of the empowered minority, and their somewhat-less empowered foot soldiers to continue the status quo under another name.

[K. A. Pital]

So automation as socioeconomic chemotherapy, then? Extremely negative short-term effects for, one hopes, a better long-term prognosis? I'd want to be extremely sure of the post-revolution results before signing on.

It's either that with, of course, unpredictable consequences - how can one know in advance the consequences of social upheaval? - or a future that evolves from the current situation steadily and most resembles cyberpunk.

The Dozois criterion which was used to describe the genre is already becoming a self-fulfilling prophecy.

"High tech - low life". With the latter meaning not just poverty or joblessness, but also lack of protection (fewer and fewer long-term job contracts with solid social guarantees, and fewer guarantees in general, fewer "socialist" elements like universal education), lack of rights (cooption of governments by ruthless corporations in the form of TISA, TTIP and other such agreements, a general globalist trend to disempower people and elected national governments; creation of a true global plutocracy) and lack of future perspectives (society's future depicted as "more of the same with new technologies", children are projected to live worse than parents, face a greater social support burden and greater hurdles in getting a decent job, own home, etc).

What people don't see is that this future can have similar technical economic indicators (e.g. unemployment rate, inflation rate etc) and yet be much worse than the "spring" of capitalism when full-time jobs and full social protection were the norm. For example, because 60% or more of all jobs are part-time. Not quite there yet, but we are moving there fast.

[cosmicalstorm]

Robotic sewing may or may not be coming. I'm not nearly qualified to make a decent assessment but progress is being made on automating the process:

September 23, 2016
Robot will finally automate sewing in the garment industry
automation, commercialization, future, jobs, robotics


A new robot by Sewbo could automate the feeding of fabric into sewing machines. Zornow has created a process by which a robotic arm guides chemically stiffened pieces of fabric through a commercial sewing machine.

Sewbo has used an industrial robot to sew together a T-shirt, achieving the long-sought goal of automation for garment production. Sewbo’s technology will allow manufacturers to create higher-quality clothing at lower costs. It will shorten supply chains and lessen the long lead times that hamper the fashion and apparel industries, helping to reduce the complexity of today’s intricate global supply network.

Despite widespread use in other industries, automation has made little progress in clothing manufacturing due to the difficulties robots face when trying to manipulate limp, flexible fabrics.

Sewbo avoids these hurdles by temporarily stiffening fabrics, allowing off-the-shelf industrial robots to easily build garments from rigid cloth, just as if they were working with sheet metal. The fabric panels can be easily molded and welded before being permanently sewn together.

The water-soluble stiffener is removed at the end of the manufacturing process with a simple rinse in hot water, leaving a soft, fully assembled piece of clothing. The stiffener can then be recovered for reuse.

Machines already play a large part in clothing manufacturing. Fabrics can be woven by machines, and then cut into pieces by computer-controlled cutting machines. There are also a few small items like dress shirt collars and cuffs that can be machine-sewn, according to North Carolina State University textiles and apparel researcher Cynthia Istook. But humans still have to put all of the pieces of fabric together, guide them through a sewing machine, and then pass them onto the next assembly line station.


The process requires an off-the-shelf sewing machine and a robotic arm, which is built by Universal Robots and costs about $35,000. The UR5 can be trained to continuously repeat a task; just move the arm to teach it a new sequence of moves. Zornow demonstrated the system by sewing a T-shirt, but it could be retrained to work with other patterns.

Apparel companies often move their manufacturing to countries where wages are lower in a perpetual quest to cut costs. The Center for American Progress found that in 2011, 15 of the top apparel exporters to the U.S. paid their Chinese garment workers an average monthly real wage of $324.90. Bangladeshi workers earned just $91.45. Meanwhile, U.S. sewing machine operators earn an average monthly wage of $1,922, according to the Bureau of Labor Statistics.

SOURCES- Technology Review, Sewbo, Vimeo

http://www.nextbigfuture.com/2016/09/ro ... ng-in.html

[Broomstick]

Great. And that's how many more people unemployed?

It wouldn't be so bad if there was an abundance of other jobs for these people to step into, but there aren't.