Any thoughts on the "robots displacing human workers" issue?

[Beric01]

But I think that's where one of the disconnects occurs. You keep saying the outcome is the same. Our point is that it is not. The robot doesn't make mistakes. While there would in theory always be a market for niche items or services made by humans it wouldn't be a large enough share of the greater market to make a difference.

I don't think it's likely that consumers will regard robots as hassle-free.  They will break down.  They will have bugs.  They will do exactly what you say but not what you actually had in mind, etc etc.  But even assuming you're correct that robots never make a mistake, why shouldn't humans just occupy the very many industries where the quality of the product can be inspected before the purchase?  (To say nothing of offering a quality guarantee like many companies do today.)

Think of it another way:  if something goes wrong, do you want there to be a human being that you can hold responsible for the error?  Or do you want to have to debug a complex machine that you own yourself?

All of these supposed issues will be solved by another robot. Robots will fix and debug other robots. They may even fix themselves! All without human intervention, and all without mistakes. I'd take the robot-made product any day.

[PloddingInsight]

But I think that's where one of the disconnects occurs. You keep saying the outcome is the same. Our point is that it is not. The robot doesn't make mistakes. While there would in theory always be a market for niche items or services made by humans it wouldn't be a large enough share of the greater market to make a difference.

I don't think it's likely that consumers will regard robots as hassle-free.  They will break down.  They will have bugs.  They will do exactly what you say but not what you actually had in mind, etc etc.  But even assuming you're correct that robots never make a mistake, why shouldn't humans just occupy the very many industries where the quality of the product can be inspected before the purchase?  (To say nothing of offering a quality guarantee like many companies do today.)

Think of it another way:  if something goes wrong, do you want there to be a human being that you can hold responsible for the error?  Or do you want to have to debug a complex machine that you own yourself?

All of these supposed issues will be solved by another robot. Robots will fix and debug other robots. They may even fix themselves! All without human intervention, and all without mistakes. I'd take the robot-made product any day.

It's turtles all the way down, huh?

[matchewed]

But I think that's where one of the disconnects occurs. You keep saying the outcome is the same. Our point is that it is not. The robot doesn't make mistakes. While there would in theory always be a market for niche items or services made by humans it wouldn't be a large enough share of the greater market to make a difference.

I don't think it's likely that consumers will regard robots as hassle-free.  They will break down.  They will have bugs.  They will do exactly what you say but not what you actually had in mind, etc etc.  But even assuming you're correct that robots never make a mistake, why shouldn't humans just occupy the very many industries where the quality of the product can be inspected before the purchase?  (To say nothing of offering a quality guarantee like many companies do today.)

Think of it another way:  if something goes wrong, do you want there to be a human being that you can hold responsible for the error?  Or do you want to have to debug a complex machine that you own yourself?

You do know that manufacturing and service delivering companies abhor inspection. It is non-value added to the product. Creating a machine which makes the product or puts it together repeatedly in a manner which obviates the need for inspection is ideal. You don't need people to check the machine work if you design the machine appropriately. There are several examples of this out there already.

If something goes wrong it won't matter if it was a person or a complex machine, something still went wrong. The machine in your scenario is actually less likely to produce an error. I get the impression that you believe they are prone to failure or consistently make mistakes. As if we're somewhere here.

But the reality is we're here. And the impact of that is not that people will feel all fuzzy and happy with self driving vehicles. But that these people and these people's jobs are going to be pretty much obsolete in a relatively short time frame. Nearly two million people in the US without jobs in a very short time line. That's just looking at transportation. The fact that something as complexly human as operating a vehicle can be done by machine safely is a big change. And can't be handwaved away with "well people will want people." Or "people will want to be able to hold someone responsible." Inevitably in our history efficiency wins out over the rebuttals you're posing. What makes it different this time?

As for who fixes the robots you've already determined through your own reasoning that there would be an employable group still. Programers, engineers...etc. These guys who build robots can and will build software and robots to fix robots. Why is that so hard to believe when it already happens. You think someone runs around playing with code when a server has a problem. No they use software, or reset remotely. The future is here already. We're just only seeing the tip of it.

[PloddingInsight]

You do know that manufacturing and service delivering companies abhor inspection. It is non-value added to the product. Creating a machine which makes the product or puts it together repeatedly in a manner which obviates the need for inspection is ideal. You don't need people to check the machine work if you design the machine appropriately. There are several examples of this out there already.

If something goes wrong it won't matter if it was a person or a complex machine, something still went wrong. The machine in your scenario is actually less likely to produce an error. I get the impression that you believe they are prone to failure or consistently make mistakes. As if we're somewhere here.

But the reality is we're here. And the impact of that is not that people will feel all fuzzy and happy with self driving vehicles. But that these people and these people's jobs are going to be pretty much obsolete in a relatively short time frame. Nearly two million people in the US without jobs in a very short time line. That's just looking at transportation.

Oh, you're right of course.  Millions of people's current job description will soon be obsolete.  And if we look at one profession in a vacuum, all other things being equal, it's obvious that robots who replace a given profession will work for cheaper and the human can't compete.  But all other things will not be equal.  What I'm trying to point out is that it isn't valid to infer from there that the robots will take over all the jobs and everybody will be out of work.  People who drive trucks and taxis today will find other work.  People who would have grown up to be tomorrow's taxi drivers will find other work.  The way to look at it is:  Today we need taxi drivers, so these people are stuck driving taxis and unable to create value elsewhere.  But tomorrow when the taxis drive themselves, all those people's productive potential will be released to do something new -- something that we can't afford to have them do today.

The fact that something as complexly human as operating a vehicle can be done by machine safely is a big change. And can't be handwaved away with "well people will want people."...

People today pay more for ethical products and I expect more of that in the future, but that is not where my argument rests.  It's a trivial point.

... Or "people will want to be able to hold someone responsible."

On the other hand, this is huge.  You will never be able to pay a robot to feel responsible for a project being a success.  In fact you can't pay most people to feel responsible for a project.  That's why there are managers who make a lot more money than the people under them and don't even contribute to the actual work.  At the end of the day, being able to delegate responsibility to another human being who you somewhat trust (for instance a VP handing a project to a manager, or a homeowner hiring a contractor that is part of his social network) is an enormous value.  Robots may earn our trust over time with routine tasks, but ultimately when large amounts of money are on the line, there is real and serious value in being able to offload the burden of sweating the details to someone else. 

This is the single one thing that people will always want to pay for and robots and technology will never be able to do.

Inevitably in our history efficiency wins out over the rebuttals you're posing. What makes it different this time?

Wait, history shows that robots create mass unemployement?  I thought you guys were the ones arguing that this time it's different.

As for who fixes the robots you've already determined through your own reasoning that there would be an employable group still. Programers, engineers...etc. These guys who build robots can and will build software and robots to fix robots. Why is that so hard to believe when it already happens. You think someone runs around playing with code when a server has a problem. No they use software, or reset remotely. The future is here already. We're just only seeing the tip of it.

We will never reach a point where technology has no glitches.  It will always malfunction and break down and cause frustration.  These things don't happen because the engineering problems involved are unsolvable.  They happen because the technology is always so new.  The only way to fix all the glitches is to outlaw innovation.

[genselecus]

Yeah, I'm in the "robots will severely disrupt the current economy, but the world won't end" camp. As ploddinginsight says, particular industries will be devastated just like has occurred in the past. But those people will be free to do something else. Now the challenge is to enable them to do something else. So let's look at taxi drivers for instance. Let's assume that 100% of taxi drivers will be out of the workforce in 20 years due to automated cars (although we need to remember that the taxi union is a very strong union that is putting up a huge fight already to companies like Uber). In this future, you'll have ~250k people that are currently, or may be, working in that industry that are no longer able to compete (thanks matchewed for the link and number). What will those people do? Well right now those people are predominately less-educated (high school or less). So what will they do? The answer is not much, at least not the way our education system currently works. Their option is either to bargain for protection (I can imagine the "Saving Taxi Drivers Displaced by Automation Bill of 2031" creating a new welfare program, because let's be honest the strongest groups in democracies are the special interest groups that create strong voting blocks) or to become educated/trained to perform a new job. I think governments need to look for ways to innovate the training and education programs so that people can empower themselves to innovate as technology advances.

A flaw that I'm seeing with the "Robots are taking over all jobs" camp is that people seem to be overlooking the fact that while robots may end up taking over 80% of jobs that exist today, that process may take more than 50 or 100 years to occur, which gives everyone time to retrain, and innovate to create new jobs. Yes there will be disruption, and yes there will be change, but you can't leap 50 or 100 years into the future to argue that there will be no improvement. There will be new industries that don't exist today, new problems that need to be solved, and new areas that need to be explored. Yes, eventually robots may be smart enough to take over all jobs (I mean once we've programmed robots how to learn and think for themselves, we don't need to pay people to come up with new ways to save the environment, the robots will do it for us). When that occurs, I think we need to worry about the robots taking more than just our jobs...

A couple of additional, random thoughts:
1. I thought the first video that was posted that was called "Humans need not apply" was interesting, but it was fear mongering. Particularly funny was that it kept bringing up the horse analogy and that horses back in the early 1900's weren't expecting their jobs to all go down the drain. No of course not! They are freaking horses that can't innovate! They don't have thumbs, they don't have aspirations, they don't want more! Humans will never be like that because we don't settle. That said, there are those humans that act as the horses quoted in the video, and they won't innovate.
2. A scary thing that I think could come out of this is a major political reset. Let's imagine that the inequality grows in the US and that at some point, the 1% owns 99% of the wealth. History has shown that the 99% tend to get pretty angry and violent, and that they tend to kill the 1%. As mustachians, our goal is to get to full financial independence which involves most of us getting towards the top 10% (or higher, given that to be in the top 10% for net worth in the US you need assets of ~$1M). This get's kind of scary when you imagine a popular uprising, although I guess if we live in small houses and drive older cars people won't come storming into our houses...

[maizefolk]

Fascinating thread!

There seems to be an argument between people saying automation could be very disruptive to society and people arguing that history and economics suggest people will still be able to find jobs even after their existing jobs are automated. I don't think the two are mutually contradictory. While acknowledging that many people who lose their jobs will be able to find new ones (often in new industries), there are two problems:

1. In cases of rapid technological change, the process of readjustment can take a generation. A fifty-year-old factory worker whose job is replaced by automation generally won't go back to school to become a web designer or nurse (obviously there will be exceptions!). And studies (focused more on outsourcing than automation) tend to show that when a company closed or moved out of the country, their former workers tend to end up in much lower paying positions once they finally find work again. However, that factory worker's son or daughter almost certainly won't go into manufacturing and may very well end up on one of the above fields not yet directly threatened by automation.

2. With each iteration of technological advance, the level of skill/knowledge/creativity needed to get a good paying job increases. Consider three classes of jobs, from easiest to replace with automation to hardest:

• What percent of the overall population could learn to work in a fast food restaurant (assuming no alternative jobs were available)?
• Now what percent could learn to write C++ well enough to get paid for it (same conditions)?
• Now what percent could create music/art of high enough quality people would pay for it (same conditions)?*

*In the world were currently live in where music and art from the best creators around the world is available at the click of a button

I'd argue the percentages are somewhere around 90%, 30%, and 5% but by all means make your own decisions about the numbers. Automation is good for the quality of life for people who can still do things robots/computers can't do. We get rid of tasks/jobs we don't want to do, and our standard of living goes up as technology pushes down the cost of goods and services. At early stages of technological advance, almost everyone who enters the workforce receives those benefits. But as technology continues to advance, a larger fraction of the population ends up competing with robots on cost rather than working in jobs that complement the robots.

So while the laws of economics likely ensure that in the long term most people will still have jobs, that doesn't mean coming advances won't be quite painful for people currently trained to perform the jobs being replaced, and it doesn't mean standards of living won't go down for large portions of the population who will be competing on price with robots to sell their labor.

I'm not sure what the take-away is. Obviously stopping technological progress is not the answer (and wouldn't be possible even if it was). Not being able to envision any job I could count on still existing when I am fifty is one of the things pushing me to grow my stash while the 20+ years I spent in school continues to make my labor valuable.

[mozar]

That's why I brought up government policy. The 1% knows* that if the rest don't have jobs (and I don't think that people only have jobs to buy more, it also creates meaning in life for many) they will riot, as such will take steps through government policy to make sure they have options. A guaranteed wage has been kicked around of late. I've read about the possibility of every adult in the US getting 10k, to raise living standards. I think of Japan often, some might say the most advanced country on the planet. Robots are more common there, the stock market has been flat for decades...and 3.8% unemployment.
What gets me is what happens when humans fully become robots. When you can upload your conscious on to software and your robot body never makes mistakes. That changes everything and will be happening sooner than we think.

*Old money knows this, but not new money. New money has to figure this out, this is a cycle through the past several hundred years.

[matchewed]

You do know that manufacturing and service delivering companies abhor inspection. It is non-value added to the product. Creating a machine which makes the product or puts it together repeatedly in a manner which obviates the need for inspection is ideal. You don't need people to check the machine work if you design the machine appropriately. There are several examples of this out there already.

If something goes wrong it won't matter if it was a person or a complex machine, something still went wrong. The machine in your scenario is actually less likely to produce an error. I get the impression that you believe they are prone to failure or consistently make mistakes. As if we're somewhere here.

But the reality is we're here. And the impact of that is not that people will feel all fuzzy and happy with self driving vehicles. But that these people and these people's jobs are going to be pretty much obsolete in a relatively short time frame. Nearly two million people in the US without jobs in a very short time line. That's just looking at transportation.

Oh, you're right of course.  Millions of people's current job description will soon be obsolete.  And if we look at one profession in a vacuum, all other things being equal, it's obvious that robots who replace a given profession will work for cheaper and the human can't compete.  But all other things will not be equal.  What I'm trying to point out is that it isn't valid to infer from there that the robots will take over all the jobs and everybody will be out of work.  People who drive trucks and taxis today will find other work.  People who would have grown up to be tomorrow's taxi drivers will find other work.  The way to look at it is:  Today we need taxi drivers, so these people are stuck driving taxis and unable to create value elsewhere.  But tomorrow when the taxis drive themselves, all those people's productive potential will be released to do something new -- something that we can't afford to have them do today.

What other work? Where do you see the opportunity? Even during the industrial revolution you could see a clear connection between farm automation and people moving to work in factories. Where is the outlet?

The fact that something as complexly human as operating a vehicle can be done by machine safely is a big change. And can't be handwaved away with "well people will want people."...

People today pay more for ethical products and I expect more of that in the future, but that is not where my argument rests.  It's a trivial point.

Fair enough if you want to drop that point.

... Or "people will want to be able to hold someone responsible."

On the other hand, this is huge.  You will never be able to pay a robot to feel responsible for a project being a success.  In fact you can't pay most people to feel responsible for a project.  That's why there are managers who make a lot more money than the people under them and don't even contribute to the actual work.  At the end of the day, being able to delegate responsibility to another human being who you somewhat trust (for instance a VP handing a project to a manager, or a homeowner hiring a contractor that is part of his social network) is an enormous value.  Robots may earn our trust over time with routine tasks, but ultimately when large amounts of money are on the line, there is real and serious value in being able to offload the burden of sweating the details to someone else. 

This is the single one thing that people will always want to pay for and robots and technology will never be able to do.

So let me see if I'm understanding this right. Robots aren't a worry because people can be taught or managed to feel responsible for their work but robots can't. If the robot makes the product or service quicker, with less errors, less variability then what is the need for the responsibility? Does the automation experienced in the automotive industry not fall under your proposition? Where are the swaths of car companies hiring people just so that they can feel like they're delegating responsibilities? Nowhere. They jumped on the automation bandwagon first, drove it even.

Where is this value for responsibility in all the other forms of automation that have happened and are happening today? Frankly I think your last line in the larger paragraph that hits the nail on the head but hit the wrong nail. There is true value in offloading burden. And people are a burden to employers in light of technology that can do it better.

Inevitably in our history efficiency wins out over the rebuttals you're posing. What makes it different this time?

Wait, history shows that robots create mass unemployement?  I thought you guys were the ones arguing that this time it's different.

A bit out of context or you're missing the point. My point in saying that is that efficiency wins out over your proposed concept that consumers and employers will want people because responsibility is valuable. I think that point has been refuted handily evidenced by the fact that very few westerners wear hand woven products and other obvious reasons.

As for who fixes the robots you've already determined through your own reasoning that there would be an employable group still. Programers, engineers...etc. These guys who build robots can and will build software and robots to fix robots. Why is that so hard to believe when it already happens. You think someone runs around playing with code when a server has a problem. No they use software, or reset remotely. The future is here already. We're just only seeing the tip of it.

We will never reach a point where technology has no glitches.  It will always malfunction and break down and cause frustration.  These things don't happen because the engineering problems involved are unsolvable.  They happen because the technology is always so new.  The only way to fix all the glitches is to outlaw innovation.

True we will never reach a point where technology has no glitches but I don't think I ever said we would. So what does that mean to what I said?

[Mega]

OK, a couple of thoughts :

1 - 40%, or even 20% unemployment pretty much ensures there will be massive unrest, unless massive changes are made to income /wealth distribution.
2 - people that are displaced by machines suffer a permanent decrease in their standard of living. Retraining for a new career is a myth (or even more cynically, a lie told to people to make them less angry about being replaced).
3 -  there was no improvement in the overall standard of living until the industrial revolution. What has happened before may not continue.
4 -  the pace of change continues to accelerate. Think about this 10 years ago everyone was using flip phones. 20 years ago Windows 95 didn't even exist, and no one except academics were using the "Internet", and cell phones were the size of a brick. 30 years ago cell phones had a battery the size of a brief case. 40 years ago cell phones were Star Trek science fiction.

Where do you think we will be in another 10 years?
Self driving cars, forcing millions of taxi, bus,  transport , and public transportation drivers out of work. And the associated police, ambulance, fire fighter job losses due to a lack of accident and funds(no more speeding tickets). The end of buying cars, car dealerships are the source of like 40% of state sales tax revenue.

3d printed food, leading to the end of the restaurant and grocery store industries.

The destruction of information barriers, leading to the destruction of knowledge based jobs (like the current situation for lawyers). Maybe Watson will replace your family doctor, and a machine will replace the surgeon. What could they retrain for?

End rant.

[PloddingInsight]

I want to underline a couple key points.

1)  Yes, the optimistic argument relies on the continuing appearance of new kinds of jobs that weren't contemplated previously, and by definition this is difficult to wrap your head around.  I understand the reasons some people might think "this time is different."  However a proper (imo) contemplation of the economics provides a strong argument in the abstract, which I have tried to explain with rather inadequate examples.  They are only illustrations of an abstract point.  If there's nothing else you take away from my comments, I hope you will remember this statement:  Income is fundamentally related to absolute productivity (what can you produce), not relative productivity (are you better than the competition).  This is hard to express properly because if you restrict yourself to a single profession and suppose that's your only option, it's not true.

2)  There's a recurring argument that middle class incomes have already stagnated and won't start rising again.  This is an unintentional straw-man, because I think the increase in income due to automation will be rather modest at the median and below.  It might even be zero.  I understand how you might take the opposite reading from some of my previous comments.  I'm saying things will turn out pretty ok and there's no need to worry, and in our current society that's easy to mistake for "incomes will continue to rise in real terms".  A better statement would be "We don't need incomes to rise in real terms.  We're fine where we are!  Yes automation will vastly enrich the wealthy and the very capable, but not at the expense of the ordinary man.  Today you can live a pre-industrial life just by trash-picking.  In the world of automation, you can probably trash-pick your way to a 2014 standard of living.  I will express this in poetry, with apologies to Kobayashi Issa.

A rising tide
lifts all boats,
but slowly, slowly.

[LibrarIan]

Librarian - You say that you wouldn't hire a person for a job that a "robot" could do. Does this mean that you never pay to attend live theater or concerts (an HDTV and a Bose Wave Radio are a lot cheaper in the long run)? You don't patronize any bakeries or restaurants for handmade pastries or dinners, since you can get a commoditized version of the same thing from the robots at ConAgra or Kraft and whoever makes Stouffers meals?

What I'm trying to do with this approach is depict what I would be inclined to believe is the mindset of most businesses. Yes, I understand that there are businesses with more favorable values and ideals, etc. However, I still think most businesses, out of pure business interests, would "hire" a robot to do [insert task] over a human any day because of cost. Robots cost very little compared to humans, no matter how little a human is willing to work for. Again, the eat, sleep, benefits, breaks, etc. stuff comes into play here. Would I, personally, refuse a human in favor of a robot? That all depends on what I'm doing. I'd rather see a live, human-performed concert over an automated orchestra, but I would rather buy a car built by robots at an auto plant over a human-built car because this process has been shown to be more reliable, safer, etc. And if I'm a business owner, humans reduce my potential profit when robots maximize it. Easy choice there.

Think of it another way:  if something goes wrong, do you want there to be a human being that you can hold responsible for the error?  Or do you want to have to debug a complex machine that you own yourself?

I build software for a living that only exists to fix other software. Quite often I don't have to debug something myself since I taught something else how to do that for me. So, to answer this question, I don't care if a human is around to take blame.

People today pay more for ethical products and I expect more of that in the future, but that is not where my argument rests.  It's a trivial point.

Not always. Just look at how successful companies like Walmart are. Pretty much the largest retail corporation in the world with the most horrible ethics. Obviously someone is buying into their practices. I also point to the prevalence of factory farming and all the nightmarish things that go along with that. If you eat meat, unless you hunted it yourself or bought it from someone you know raised that cow in a nice field and such, you can pretty much guarantee what you're eating suffered unimaginably for your meal. My point is not to debate retail or diet, but to point out that playing the ethics card in today's world is laughable.

To address the example above of automated cars putting people out of work and the whole "look at the bigger picture - they'll find work elsewhere" thing: I think this is a big part of the debate here. Historically, yeah, people obviously found work. But I'd argue that since advancement technologically was painfully slow way back when compared to now (we're talking decades or centuries between big advances that disrupted society), arguing that drivers (or whoever) will just find other work is becoming a weaker argument. Since the dawn of modern computing technology, we've already seen whole swaths of job sectors diminish at an incredibly fast rate and the rate is ever-increasing. Speed is key. It's hard to switch professions or catch up, especially if you're older. Again,  I suggest watching Will Work For Free (it's free on YouTube) for an in-depth look at every sector of work we know and how they can and will easily be automated away.

[gillamnstr]

So there is a lot of debate on what is going to happen, but I am curious on what anyone thinks we should try to DO to prepare for this...

I'm thinking the obvious:

-save save save $
-keep educating yourself to stay current in a job market, though I'm not sure what would be good areas to focus on...
-try to pay off any mortgages in case of job loss

Dunno. Tell me what you think.

[PloddingInsight]

So there is a lot of debate on what is going to happen, but I am curious on what anyone thinks we should try to DO to prepare for this...

I'm thinking the obvious:

-save save save $
-keep educating yourself to stay current in a job market, though I'm not sure what would be good areas to focus on...
-try to pay off any mortgages in case of job loss

Dunno. Tell me what you think.

Invest in companies that do robotics and automation, of course!

[Posthumane]

I can see that automation will be able to replace certain tasks which fall into the realm of "observe-react" which encompasses many of the jobs on the low end of the pay spectrum, but I don't think they will be able to replace creative thinking in the foreseeable future, if ever. While this may put a lot of people out of work, being able to think creatively without having to do labour will open up a lot of new industries. This has already happened with things like the manufacture of small plastic products. In the past if someone wanted to create a new product, they had to know how to design it, draft it, and then machine, cast, or otherwise build that item. Nowadays people can imagine a product, draft it in 3D on the computer with much less hassle than before, and print it on their 3d printer, or mill it on the CNC mill. This probably is displacing many jobs of people who specialize in machining and manufacture, but it has allowed a lot of people to get into manufacturing than was otherwise possible.

One industry which has been mentioned that can be easily replaced with automation is the food industry. The technology has existed for automated manufacture of food products for a long time, and yet it hasn't taken over the food sector for several reasons:

1. People seem to prefer products made by humans. This is a counter to the argument that robots make products that are perfect and people won't buy products made by fault-prone humans.

For many years we've been able to make an automated coffee stand which freshly grinds the perfect amount of coffee, heats the water to a consistent temperature, froths the milk, and pours it into a cup, all for less than 50c/cup. After all, making coffee is a purely mechanical process requiring no creativity. And yet, some of the most successful businesses are coffee shops selling drinks for $2-$5/cup made by a "barista". There are other examples, though they are less prevalent, such as the existence of bakeries even though factory made Twinkies and other factory made desserts are cheaper and more consistent.

2. The cost of labour is only one of the inputs to making a product, and in many cases is actually one of the smaller ones.

Driving is a good example of this. The cost of gasoline, maintenance, the initial purchase price, and other inputs to run a vehicle is a large percentage of what you pay for when you get a taxi or bus ride. Replacing the driver with an automated solution will only save a percentage of the cost. This means that some people will still be willing to pay a little extra for "human service" (see the barista effect in point 1), and human drivers will only be replaced in areas where labour is expensive and machinery is cheap. This is true for industries like fast food as well, where the cashier making $10/hr is a tiny portion of the operating cost when they are selling hundreds of dollars of goods per hour.

One thing I noticed when travelling is that europe has the most automation for menial tasks, most places I've been to in Africa have almost none (north america seems to have a bit less than europe). Places I expected to see automated gates and security cameras had gate attendants and security guards in Africa because the cost of adding automation and maintaining it was more than the cost of labour.

Anyway, generally I agree with PloddingInsight on the macro points and he's able to express them quite well, and I just wanted to throw in a couple of specifics of why technological advancement won't suddenly render humans obsolete.

[TreeTired]

1952,    Player Piano,   Kurt Vonnegut

the chilling tale of engineer Paul Proteus, who must find a way to live in a world dominated by a supercomputer and run completely by machines.

http://www.amazon.com/Player-Piano-Novel-Kurt-Vonnegut/dp/0385333781

[PloddingInsight]

Let's try looking at this from a different angle.  Today there are people who are long-term unemployed and don't seem to be good for much in the modern economy.  Let's pull a number out of the air and say that's the bottom 10%  (might be wildly off -- the number is not the point.)  Now, according to the arguments I've made, these people should be employable, right?  So does the fact that they are not suited to any kind of employment mean that we're already moving into a robotic dystopia?  I would suggest not.

First, these people are not employable at the current minimum wage.  We don't know that they're not employable at lower wages.  Second, they are often supported by the public, up to some minimum standard of living.  This means that a job that pays less than government benefits is relatively unattractive to them.

Removing both government benefits and the minimum wage would no doubt make life extremely hard for this class of individuals.  (I'm not suggesting we do this, I just want to examine the probable outcome for the sake of an economic argument.)  Even once they got over the shock and emotional impact of the change in their situation, they would be working harder and consuming considerably less than they are today.  Today, working for a minimum wage gets you about $15,000 a year.  Even then you are eligible for a lot of government benefits.  The combination of all laws and programs probably sets a minimum consumption level somewhere in $20,000-$30,000 per year.  It's probably no surprise that this is around half the GDP per capita.  We're not a completely egalitarian society but we're uncomfortable with a standard of living that is much less than, say, 1/3 of the GDP per capita.

I said in a previous comment that wages are fundamentally attached to your absolute level of productivity.  So from that perspective the problem for that bottom 10% is that their productivity is too small in comparison to the GDP per capita.  If they did work for a market wage, we would be uncomfortable with the result.  The result is that they're out of work.  In 1850 the same individuals probably would have been employed, earned a small wage, and supported themselves at an 1850 standard of living.  The only thing that has changed is that we have essentially abolished that kind of life because it is uncomfortably low in comparison to our standard of "normal".  Technology has made most (probably all) people more productive than they were in the past.  However, at the low end of the ability scale, technology hasn't increased their productivity enough for them to keep up with the rest of us.

As robotics and automation progress, real GDP will skyrocket and this effect will become more pronounced.  The most capable human beings will be astronomically more productive, while ordinary folks will see their productivity only rise by a lesser extent.  The question is, at some point will we become so rich in aggregate that we will abolish the kind of living standard that is achievable for a "normal" human being in the automated economy?  If we, as a society, decide that such a lifestyle (and it will be better than the 2014 median) is unacceptable, we may choose to level out  consumption to a greater degree than the free market would allow.  In this scenario, we are forgoing even higher production, (just as today we're forgoing productivity that would occur if there were no minimum wage or welfare) in order to level out living standards and bring everyone up to an agreed-upon minimum.  A very high minimum.

So to me the question isn't whether we'll all have jobs, or else be destitute.  It's whether we'll be allowed to remain at the low level of consumption that our jobs would provide for us.  Will the 2014 American standard of living become regarded as an unacceptably impoverished way to live?  If so, I think it may become true that the machines have taken all our jobs!  (In that case, we will continue this conversation over gin and tonics on the cruise ship.)

[Mega]

Short summary.

It sucks to have your skills replaced by a machine. It sucks so much you suffer a permanent loss of income. BUT it is great for your children as they will benefit from the increase in productivity.

[maizefolk]

@PloddingInsight

In the absence of a minimum wage, I agree most people would still be employed doing _something_ although I think you are not giving enough weight to wage stickiness when it comes to existing workers vs new people entering the workforce without existing expectations of what their labor is worth.

But this is a bit of a strawman argument as well. When a robot displaces a human from a job that person has trained for for years or decades, their productivity drops a lot because outside of their now non-existent field of specialization they don't have the skills/knowledge/ability to contribute to society at the level they were previously. A second-tier lawyer who previously made made 80k a year wading through piles of documents in the discovery phase of lawsuits who is replaced by a computer algorithm for finding relevant documentation* might be able to make 20k a year answering ads to organize people's closets on TaskRabbit. Just because that person was able to find other employment doesn't mean they have been any less displaced by automation or their chances of saving enough for FIRE haven't been severely disrupted as a result.

*Actually happening and has been for years: http://www.nytimes.com/2011/03/05/science/05legal.html?pagewanted=all

[DarinC]

In general the pace of technology tends to be proportional to the number of jobs that are displaced. That's true of any transition, and I don't think this one will be any different. Ultimately, people will find other things to do, although it can take a while and both the private and public sector have large parts to play in retraining displaced workers.

IMO the real issue isn't that automation is displacing entire industries, but that the profits from said automation are going to a very small group of people and aren't being reinvested into economic sectors that benefit people as a whole. How much we reinvest in training/transitioning workers, research, higher education, and in general projects that are good for humans as a whole, will dictate how much this automation improves everyone's life.

If most of these profits go to a select few, then sure, we're increasing productivity, but we're also maintaining or maybe even reducing the standard of living via income inequality. If a proportional amount is reinvested back into the world, even if it takes a bit (eg the Gates foundation), then we're enhancing both our productivity and our standard of living.

[PloddingInsight]

@PloddingInsight

In the absence of a minimum wage, I agree most people would still be employed doing _something_ although I think you are not giving enough weight to wage stickiness when it comes to existing workers vs new people entering the workforce without existing expectations of what their labor is worth.

But this is a bit of a strawman argument as well. When a robot displaces a human from a job that person has trained for for years or decades, their productivity drops a lot because outside of their now non-existent field of specialization they don't have the skills/knowledge/ability to contribute to society at the level they were previously. A second-tier lawyer who previously made made 80k a year wading through piles of documents in the discovery phase of lawsuits who is replaced by a computer algorithm for finding relevant documentation* might be able to make 20k a year answering ads to organize people's closets on TaskRabbit. Just because that person was able to find other employment doesn't mean they have been any less displaced by automation or their chances of saving enough for FIRE haven't been severely disrupted as a result.

*Actually happening and has been for years: http://www.nytimes.com/2011/03/05/science/05legal.html?pagewanted=all

This has always been happening and will continue to happen.  I don't think it has much to do with the discussion we're having.  We're talking about the long term effects of automation, not the disruption it causes over the first 10 years or so.  So for a lawyer who is replaced by a big computer like Watson, the question isn't what impact it has on him, the question is, will young kids with similar abilities be able to find work when they grow up in the age of automation, or is old-fashioned second-tier lawyer work really the only thing they would be good for?

We should still care about people who are laid off.  It's just not what this thread is about.

[CCCA]

I think I maybe, sort of of understand the argument that PloddingInsight is making though I'm not an economist so I don't have the inherent faith in the answers without specifically thinking of concrete examples and that is the challenge with this type of idea.

I do think there are many humans who are not particularly intelligent, who are not particularly creative, and who do not have exceptional physical (motor skill, construction, etc) abilities, where it's only a matter of time when a machine can replace any and all of their potential "value" to society (and I mean that in an economic, i.e. productive, income producing way) at a potentially much lower cost.  I don't think that is a controversial statement.  It may take a different amount of time for each one of these abilities to be surpassed by humans, but it seems to me to be inevitable.  Technology is improving exponentially and presumably will continue to do so for decades. 

When viewed only through this lens (which is the point of the "Humans Need Not Apply" video), it is scary to think of how our kids and grandkids will find jobs and earn an income.   

People with greater amounts of intelligence, creativity and physical abilities can expect that their skills will take longer to be automated but again, it seems like it is a matter of time. 

The question then is what do people do?  What do people have as income if they cannot trade their labor for money? 
I think this is an important question moving forward.  If everyone is given a stipend by the government and the cost of goods and services are super cheap because of all the automation cost savings then maybe the system works.  I just worry about a future in which people don't have jobs AND don't have the income to live at a reasonable quality of life.  In this world, it's not like you can tell someone to work harder or get a vocational education because there is no point. 

[genselecus]

2 - people that are displaced by machines suffer a permanent decrease in their standard of living. Retraining for a new career is a myth (or even more cynically, a lie told to people to make them less angry about being replaced).

I just wanted to address this issue because I wholeheartedly disagree with you. There are hundreds of thousands of jobs that require little to no higher education (college/university) in the United States that you could get with some skills training. I work in the energy space and my clients constantly worry about whether they will have labor available to do the construction work that their projects require. People can get new jobs, because they exist. Do they take time? Yes, you will have to be an apprentice welder, electrician, mason, boilermaker, etc. (although recognize that the demand is so high right now that you could probably do that at an annual salary of $30 - $50k a year) for 1-4 years. But then once you become a journeyman, you would be making anywhere from $60k to $120k a year. And I don't see those jobs going anywhere for 50+ years. For some reason, a lot of people are unwilling to take craft labor jobs even though there are tons to be had.

If you are graduating from high school sometime soon, and are looking to FIRE asap, don't go to college, join a craft labor union (I would recommend welders or boilermakers), and be FIRE by the time you are 30.

And this isn't an issue that is unique to the US. In most advanced countries, there exists a significant skills gap in many of the craft industries. Unemployment is huge in Spain right now, but I just read an article about a severe shortage of skilled labor.

[Bakari]

Librarian - You say that you wouldn't hire a person for a job that a "robot" could do. Does this mean that you never pay to attend live theater or concerts (an HDTV and a Bose Wave Radio are a lot cheaper in the long run)? You don't patronize any bakeries or restaurants for handmade pastries or dinners, since you can get a commoditized version of the same thing from the robots at ConAgra or Kraft and whoever makes Stouffers meals?You think this sounds ridiculous, and five people are ready to shout "Not the same thing at all, bad example, preservatives, LOCAL!, farm-to-table" and I'll simply respond "Exactly."Maybe it's subjective evaluation, maybe it's snob appeal, often it's a little bit of both, but new needs always arise. See Water, bottled.

This is the first comment that has actually addressed the question at hand.

The point of my adding "solar powered" and 3D printed is that eventually the marginal cost of robot labor becomes ZERO.
Not a sandwich, not a penny, zero.
So nobody will be "willing to work for as little as it costs to hire a robot to do the same job" - you don't have to "hire" a robot.  You just buy one, and then it works forever for free.

So, as Alex321 points out, some people are willing to pay extra for the sole sake that a human does the labor instead of a robot - even if the robot produces equal or even superior quality - whether its food or art or massage.
So employement won't drop to zero.

However, the question remains whether that will be enough jobs to provide something to do for 10 billion laborers.
Especially as long as we maintain the worldwide standard of 30-50 hours a week.

 It's not clear to me why a self-contained hydroponics system with a built in 3D printer couldn't make sandwiches just as cheaply as the solar powered roomba could mow your lawn.

ok, so sandwich makers just lost their jobs too!

But I'm not claiming that the lower and middle classes will receive a standard of living increase proportional to their numbers.  I'm just saying that all other things being equal, the worst that will happen is that their wages will stay the same.  They won't be thrown into the poor house just because robots are creating a lot of goods very cheaply.  If anything they will be better off.

We don't have to speculate: lower and middle class wages really have declined relative to inflation over the past few decades as technology has expanded

Under a free market, increases in efficiency are likely to benefit everyone.  But under capitalism, they only benefit investors, at the expense of labor.

Not sure what definitions you are using here.

The technical ones: http://biodieselhauling.blogspot.com/2014/04/free-market-vs-capitalism.html

In theory, having robots do almost everything, and having humans do just those few things that robots couldn't possibly do (not sure what those things are, but lets assume there are some), could mean that every human has a 1 hour work week and a 10 years working career, and they earn an inflation adjusted $1000 per hour for the work that they do.  The economy would be able to support it.

This is an interesting claim.  Aren't you assuming there's no more keeping up with the Joneses?  If the economy ever resembled what you described, most people would spend their $1000 and then spend a lot of their free time doing labor to increase their standard of living in order to show off or attract a mate, or whatever.  They'd either create something for themselves (like doing an addition to convert your house to a McMansion) or find some kind of freelance income to earn money for a nicer car, or nicer vacation, etc.  It doesn't matter that robots could do things more efficiently because once they spend their $1000 for the week, all they have left is their capacity to labor.  So they would labor.  They'd either try to sell their labor to other workers or else create things that they inherently value. 

There are plenty of people today who want to sell their labor, and can't find a buyer.  About 10 million of them, 1/30th of the population.


If people of the future could find anyone interested in buying their labor then that's great! 

I'm not objecting to some people working harder to get more stuff.  I'm objecting to a system in which all gains made from technical progress go to the investor class and none goes to laborers.  I'm also objecting to a system in which instead of distributing the available need for labor equally we have some people work for 50 hours a week and others who are unemployed.

Another way to look at it is that 2014 technology could probably provide everyone with an 1850's level of consumption with only an hour or two of work a week.  But people would never just settle for that.  They work more in order to consume more.  It doesn't matter that their most basic needs are already met at very low costs.

I agree!  We could.  We should!
The reason we don't isn't because "people would never settle" for it.  Lots and lots of people would be willing to accept a 20 hour work week (at the same total monthly take home pay) without taking a second job, if their employer would agree to it.  The reason we have a 40 hour work week is because its the law.  In the 1850s, and up until the 1920s or so, 60-80 hour work weeks were fairly commonplace.
That changed because workers demanded it (and technology facilitated it).
Because gains of productivity go to investors, not to workers, workers work more just to consume the same amount.  If the choices are full time or unemployment, having basic needs met isn't something anyone can take for granted.

[Bakari]

There seems to be an idea that robots could only replace menial labor, that those who are intelligent and/or educated and/or creative will be exempt from unemployment.

AI threatens to displace all of that too:
http://www.motherjones.com/media/2013/05/robots-artificial-intelligence-jobs-automation

2)  There's a recurring argument that middle class incomes have already stagnated and won't start rising again.  This is an unintentional straw-man, because I think the increase in income due to automation will be rather modest at the median and below.  It might even be zero.  I understand how you might take the opposite reading from some of my previous comments.  I'm saying things will turn out pretty ok and there's no need to worry, and in our current society that's easy to mistake for "incomes will continue to rise in real terms".  A better statement would be "We don't need incomes to rise in real terms.  We're fine where we are!  Yes automation will vastly enrich the wealthy and the very capable, but not at the expense of the ordinary man.  Today you can live a pre-industrial life just by trash-picking.  In the world of automation, you can probably trash-pick your way to a 2014 standard of living.  I will express this in poetry, with apologies to Kobayashi Issa.

A rising tide
lifts all boats,
but slowly, slowly.

Now that is actually a pretty good point.


The question I still have is will the 99.8% of humans who end up trash pickers be content with that in a world where a few people inherit all their wealth and live forever in android bodies and make peons fight to death for their entertainment?
Or rather, is their any legitimate reason to have most people living by bartering food and trash-picking tech while others do no work and live in unimaginable luxury? 
At what point do we start to question to assumptions about property and what it means to "earn" that we have under capitalism?
http://biodieselhauling.blogspot.com/2014/04/free-market-vs-capitalism.html

[maizefolk]

So for a lawyer who is replaced by a big computer like Watson, the question isn't what impact it has on him, the question is, will young kids with similar abilities be able to find work when they grow up in the age of automation, or is old-fashioned second-tier lawyer work really the only thing they would be good for?

No, the question is whether young kids with similar abilities will be able to find work which provides quality of life equal to or better than the previous generation. As I said, assuming no minimum wage and no better options, many people will probably be able to find something to do with their time that will earn some quantity of money. However if automation continues displacing human beings from fields where it is possible for a person to be more productive, median standards of living could still fall radically (and at earlier points in the automation of the economy the standards of living for the 5th, 10th and 25th percentiles).

[zolotiyeruki]

We don't have to speculate: lower and middle class wages really have declined relative to inflation over the past few decades as technology has expanded

Actually, that graph proves exactly the opposite. Note the "Real 2010 Dollars"--that means it's adjusted for inflation. So the bottom 90% really have seen no significant change in their buying power.

[sol]

So the bottom 90% really have seen no significant change in their buying power.

You think negative 7% isn't significant?

Especially looking at the large positive increases for the top 10% or 1%, I think the fact that the bottom 90% is negative while the top 10% keeps going up is pretty on point.  Lower and middle class wages have declined, both in absolute and (more so) in relative terms.

[PloddingInsight]

I don't think that graph allows much in the way of a firm conclusion.

It's difficult to see how much fluctuation there is in the bottom 90%.  So it's not clear how crucial the selection of beginning and ending points is.

The decline may be swamped by some increase in government services and benefits that doesn't count as "income".

[PloddingInsight]

The point of my adding "solar powered" and 3D printed is that eventually the marginal cost of robot labor becomes ZERO.
Not a sandwich, not a penny, zero.
So nobody will be "willing to work for as little as it costs to hire a robot to do the same job" - you don't have to "hire" a robot.  You just buy one, and then it works forever for free.

I'm not buying what you're selling here.  The machine will eventually break down or need service.  It will have a lifespan.  This is like saying there's zero marginal cost for driving somewhere, because you already paid for the gar and filled up the gas tank.

And you certainly have to hire it if you don't already own one.

 It's not clear to me why a self-contained hydroponics system with a built in 3D printer couldn't make sandwiches just as cheaply as the solar powered roomba could mow your lawn.

ok, so sandwich makers just lost their jobs too!

You have completely missed the point, because my original argument rested only on the assumption that the mowed lawn and the sandwich were worth the same amount, regardless of how little that was in dollars.

[PloddingInsight]

[/font]

But I'm not claiming that the lower and middle classes will receive a standard of living increase proportional to their numbers.  I'm just saying that all other things being equal, the worst that will happen is that their wages will stay the same.  They won't be thrown into the poor house just because robots are creating a lot of goods very cheaply.  If anything they will be better off.

We don't have to speculate: lower and middle class wages really have declined relative to inflation over the past few decades as technology has expanded

I don't know where you found this chart, but the following pdf that I found through google roundly contradicts it.

http://www.russellsage.org/sites/all/files/chartbook/Income%20and%20Earnings.pdf

[PloddingInsight]

The technical ones: http://biodieselhauling.blogspot.com/2014/04/free-market-vs-capitalism.html

What is this, your personal blog?  Still not buying it.  If you go to the Capitalism page on wikipedia, the first thing you read is "free enterprise redirect here".

There are plenty of people today who want to sell their labor, and can't find a buyer.  About 10 million of them, 1/30th of the population.

I'm pretty sure I already addressed the causes of unemployment today.

I'm not objecting to some people working harder to get more stuff.  I'm objecting to a system in which all gains made from technical progress go to the investor class and none goes to laborers. 

I'm a laborer and I have a smart phone so I'm not clear on why you think I'm not benefiting from technical progress.  The profits earned from selling technology goes to the technology companies of course, who are mostly owned by the public via stock.  Including me.  So I'm a laborer and an investor, which is true of most Americans.

I'm also objecting to a system in which instead of distributing the available need for labor equally we have some people work for 50 hours a week and others who are unemployed.

Go apply for a job at a construction company.  But tell them you only want to work 10 hours a week.  It's simply not worth their time and effort to go through the motions of hiring you.  "It's the system, mannn!"  No, it's reality.

The reason we don't isn't because "people would never settle" for it.  Lots and lots of people would be willing to accept a 20 hour work week (at the same total monthly take home pay) without taking a second job, if their employer would agree to it.

Well, yeah.  The same people would all work 40 hours at double that amount, too.

The reason we have a 40 hour work week is because its the law.  In the 1850s, and up until the 1920s or so, 60-80 hour work weeks were fairly commonplace.
That changed because workers demanded it (and technology facilitated it).
Because gains of productivity go to investors, not to workers, workers work more just to consume the same amount.  If the choices are full time or unemployment, having basic needs met isn't something anyone can take for granted.

At less than 40 hours a week, a large number of people become willing to put in extra hours to increase their pay.  That tells us that 40 hours is the sweet spot, or close to it.  There's a reason that the work week didn't keep shrinking (as some predicted early on).  The reason is keeping up with the Joneses.

I'm not going to really comment about the gains going to investors rather than the laborers.  Socialism has been tried.  It don't work.

[Bakari]

I don't know where you found this chart, but the following pdf that I found through google roundly contradicts it.

http://www.russellsage.org/sites/all/files/chartbook/Income%20and%20Earnings.pdf

No it doesn't!  It doesn't divide income levels, it just shows an average.  The two are completely compatible.

[PloddingInsight]

I don't know where you found this chart, but the following pdf that I found through google roundly contradicts it.

http://www.russellsage.org/sites/all/files/chartbook/Income%20and%20Earnings.pdf

No it doesn't!  It doesn't divide income levels, it just shows an average.  The two are completely compatible.

Ok, rising incomes at various percentiles across the board isn't really consistent with a falling average, but I believe I can explain the discrepancy now.  Your chart is probably income rather than household income.  As two-income households have become more prevalent, the household incomes can rise even as individual incomes fall.

Look at the second-to-last page of the pdf, which uses individual incomes instead.  Now we can see the detail that is obscured by the scale of your chart, as well as the context of what happened before 1970.  (Probably it's best to ignore the women's income line because women's workforce participation and educational achievement was dramatically changing during the course of time.)  It's pretty clear that there's a slight decrease in men's income if you measure specifically from 1970-2010, which is consistent with your chart.  But that is because all the truly massive increases happened in the 1950-1970 timeframe, and the great recession caused a drop in incomes just before the final datapoint in your chart.

This isn't a story of a decline.  The take-away is clearly that men's individual incomes have fluctuated in the range of $32k-$37k since 1965 and don't show any indication of leaving that range.

[Bakari]

Really, what wikipedia redirects to is how to define economic terms?  You don't even have to read past the introduction to see even the wikipedia page acknowledge there is a distinction, and that capitalist markets are not necessarily free.

"Capitalism involves the further abstraction of money into other exchangeableassetsand the accumulation of money through ownership, exchange, interest and various otherfinancial instruments.The accumulation of capital refers to the process of "making money", or growing an initial sum of money through investment in production. Capitalism is based around the accumulation of capital, whereby financial capital is invested in order to realize a profit and then reinvested into further production in a continuous process of accumulation.
The defining feature of capitalist markets, in contrast to markets and exchange in pre-capitalist societies like feudalism, is the existence of a market for capital goods (the means of production), meaning exchange-relations (business relationships) exist within the production process. Additionally, capitalism features a market for labor. This distinguishes the capitalist market from pre-capitalist societies which generally only contained market exchange for final goods and secondary goods."
Some requirements for a perfect free market, according to the father of Economics, Adam Smith:
A large number buyers and sellers

• No barriers of entry and exit
• Perfect factor mobility
• Perfect information
• Rational buyers
• No externalities

The corporation is the perfect example of capitalism undermining the free market - the government grants a corporate charter, giving the company a competitive advantage, due to its access to capital, over an independant business.

Go apply for a job at a construction company.  But tell them you only want to work 10 hours a week.  It's simply not worth their time and effort to go through the motions of hiring you.  "It's the system, mannn!"  No, it's reality.

That was the argument buisness made against the 40 hour work week too - which apparently didn't destroy the economy as predicted.

At less than 40 hours a week, a large number of people become willing to put in extra hours to increase their pay.  That tells us that 40 hours is the sweet spot, or close to it.

At more than 40 lots of people are willing to put in extra hours too.  The reason more don't is employers don't want to have to pay overtime.
If overtime kicked in at 20 hours, employers would would still not want to pay overtime, and lots of new jobs would become available, and there would be minimal unemployment.   In fact, there would be a labor shortage, and since labor has a supply and demand relationship just like goods do, it would push wages up to where they should be

Socialism has been tried.  It don't work.

I think you meant to say "communism"?
Socialism is still alive and well.
Denmark, Finland, Netherlands, Canada, Sweden, Norway, Ireland, New Zealand, Belgium, all socialist economies, and all, incidentally, with about as high or higher standards of living as the US.
A few examples doesn't "prove" what system is superior, but its enough to prove that it can work.

[PloddingInsight]

Whatev

I didn't come here to argue free markets vs socialism.

[Daley]

If robots can create everything for 10% of what human workers cost, then all our consumables will be 90% cheaper and workers will be able to purchase everything they need on a salary that is only 10% of what they used to earn... which means they'll be competitive with the robots again.

It isn't magic, it's economics.  You're looking too hard at the dollar bills.  The money is just a system that allows us to trade our labor for things.  As long as our labor doesn't get less productive and the things don't get harder to make, we'll be all right.  And in fact robots will make humans more productive and things less costly.

Whatev

I didn't come here to argue free markets vs socialism.

Kinda goes hand-in-hand with your defense of the technology, chum.

[PloddingInsight]

If robots can create everything for 10% of what human workers cost, then all our consumables will be 90% cheaper and workers will be able to purchase everything they need on a salary that is only 10% of what they used to earn... which means they'll be competitive with the robots again.

It isn't magic, it's economics.  You're looking too hard at the dollar bills.  The money is just a system that allows us to trade our labor for things.  As long as our labor doesn't get less productive and the things don't get harder to make, we'll be all right.  And in fact robots will make humans more productive and things less costly.

Whatev

I didn't come here to argue free markets vs socialism.

Kinda goes hand-in-hand with your defense of the technology, chum.

I find the robotics & automation thing very interesting, but as soon as I realize the person I'm talking to believes in socialism, I get bored.  It's like discussing the merits of a rule in American football with somebody whose whole view is based on their preference for soccer.  There's not enough common ground.

[AlanStache]

It seems everyone agrees on a few things
>> automation will be highly disruptive in the near future
>> some new jobs will be created
>> price of things will drop with more automation

I am somewhat swayed by PloddingInsight, but not 100% convinced.  I still believe that in the longer term very large percents of the population will be unemployable, that automation will be to cheap (costs only sunlight) and to good, at way to many things that a great many humans just wont be able to compete at any price.  I see your point that with cheaper goods humans can work for less compensation, but I just am not convinced that humans will be worth the effort for the employer/customer.

As far as AI/machines breaking down or not being good - WTF?  I have NEVER had a problem with any ATM-ever, I booked over a dozen plane tickets last year on delta.com with nearly zero problems, my '04 Toyota gets serviced twice(?) a year, rarely do I look past the first page on google, go into BestBuy and buy anything I will pay you 100$ if it does not work 100% out of the box for years and when it does brake the new one will be half the price and twice as good.

re Africa using cheap human labor vs current automation: Perhaps this is an equilibrium point in economics but could we in the 'west' be able to get there?  Society is a path dependent function - where you can go depends on where you are and where you have been. 

re 80k lawyer: Ok so today's lawyer is S.O.L. when he is replaced by AI and ends up with a job paying 25% of the old purchasing power but his similarly competent kid will be better off with some as yet unimagined job.  No.  How long till the kid is replaced by AI as well?  It may not be turtles all the way down but there are a heck of a lot of turtles towering up.

Strongly recommend The Lights in the Tunnel by Martin Ford, is a free pdf version on the authors site (google it).  Makes a strong case for those at the higher wadge end to be worried too, basically it is worth the effort to automate a doctor/lawyer.

How many baristas and local artisans do we need?  How many service sector jobs will be around just because we like human contact?  Keep in mind that the unemployed could supply that human contact for free, no 5$ cup of Joe required.

Basically we are fucked in lots of ways.  Two hopes I see
1) Our leaders recognize the situation and help steer us to a happy utopia.
2) Humans evolve to compete more directly with AI.  We become the Borg.  If you are laughing at this remember driverless cars are now legal in two(?) states.

[Bakari]

I find the robotics & automation thing very interesting, but as soon as I realize the person I'm talking to believes in socialism, I get bored.  It's like discussing the merits of a rule in American football with somebody whose whole view is based on their preference for soccer.  There's not enough common ground.

I was never advocating socialism.  Not unless you want to call Adam Smith a socialist.  You brought up the word, I just added a brief note that you probably meant "communism".

What I'm advocating is only this:


1- reduce the government interventions which enhance capitalism at the expense of a truly efficient market (most notably by not issuing corporate charters - corporations are by design anti-competitive)

2- tie working hours to total average productivity.  If output per worker doubles, overtime laws should reflect that by kicking in at half what they used to.  This way as the need for labor decreases as automation gets better, it affects all employed persons equally, instead of having some work long hours and others get unemployment checks.

[jordanread]

Ah, I do love this topic, and am glad to see a good discussion happening. May have gotten a touch derailed, but it all goes hand in hand. We live in a system where everything is connected. From politics, to economy, to education, all facets of life would be affected by this. All right, now let's get to the content.

Up thread, Knaak put in the link for "Humans Need Not Apply", and shortly afterwards, LibrarIan put in a link to "The Lights in the Tunnel". Both are pretty good, and do a good job of laying out what is happening, and what can be expected to happen over the next years. The fun part comes in as to how these things happening is going to shape our lives. Even though it wasn't the purpose of the book, The Lights in The Tunnel did an excellent job of outlining a fundamental issue that causes so much concern. The question everyone asks is "How can our economy survive?" at which point the discussion goes the route of wages, prices for goods, and how the two will look. Will it really even out? Will the wage gap get worse? How can we sustain this? Martin Ford (author of Lights in the tunnel) addressed this in an interesting way. But once again, he's asking the same question as everyone else seems to be. It is sometimes hard to remember that we build fundamental assumptions, and we sometimes forget to challenge those. I get more surprised by this in a community like this one, but it always exists.

So there are lots of ideas and proposed solutions on how to make it so the economy survives. My question is this: Should it survive? My personal stance on this is that it shouldn't. With everything that has happened with the crash, unemployment has some understandably bad connotations. PloddingInsight started to touch on it, as did a few others. We see our current consumption based economy, and know that it's going to be disrupted, and that scares us. But lets take this all the way. Goods become cheap, people aren't necessary for the "economy" to run. Why would we worry about jobs? Base needs are handled, and innovation will increase, as the only reason to work on something would be passion. The mindset right now is that our jobs define us. What about everything else? This starts moving into post-scarcity territory, but the underlying assumption that we have to consume to maintain the economy is only valid when you consider the economy in it's current form. There is a huge shift in thinking that would be required to make this work in society as a whole, but I think we could do it.

What do you think? It sure seems like we are trying to stay in the box with our thinking, and ignoring the paradigm shift that will happen.

[Posthumane]

I don't see a post scarcity world being created with increased use of robotics and automation. Any product which we consume has a number of inputs required to produce it, and human labour is just one of them. In many industries it is a very small percentage of the inputs, with farming being a good example. A typical North American farmer is both the owner and operator (labour) on his/her farm, and yet only brings in a few tens of thousands of dollars per year despite having cash flow that is an order of magnitude larger than that. For every 10k he pays himself he probably spends over $100k on fuel, chemicals, seed, equipment maintenance and replacement, etc. Having a completely robotic farming infrastructure would not necessarily drop the price of food by a significant margin.

[Bakari]

I don't see a post scarcity world being created with increased use of robotics and automation. Any product which we consume has a number of inputs required to produce it, and human labour is just one of them.

I don't think anyone here has been arguing that there will be unlimited everything.  The question is how are resources allocated in a post-labor society. 
Although its also true that with good recycling a lot of inputs may become cheaper and easier to acquire - I can't think of much that would provide a hard limit that couldn't potentially be recycled or produced synthetically or have an alternative source, can you?

farming being a good example. A typical North American farmer is both the owner and operator (labour) on his/her farm,

If you look at farms by number of farms, yes, the greatest total number is small farms.
However, if you look at production - percentage of food that actually comes from farms - the majority is from extremely large farms.



http://www.washingtonpost.com/blogs/wonkblog/wp/2013/08/11/farms-are-gigantic-now-even-the-family-owned-ones/


The dark blue is just totaling farms, which makes it look like the "average farm" is small, but in terms of where your food actually came from, the light blue is the relevant data.  Small farms produce a fairly negligible percentage of food in the US.

Even farms that are technically "family" owned are almost all gigantic



http://www.epa.gov/agriculture/ag101/demographics.html


The term "family" farm just means its owned by a private individual, as opposed to a corporation.
Thats like calling Dell Computer a "family business".  The 3rd largest computer supplier is still owned by the Dell family. 

And the majority of farms that are genuinely owned and operated by an individual are working land sold off to them by a corporation,  with contracts to supply that same corporation
http://www.bigpictureagriculture.com/2010/09/us-statistics-on-farming-by-contract.html
 in what amounts to almost indentured servitude, in the manner of the old mining town company stores

Having a completely robotic farming infrastructure would not necessarily drop the price of food by a significant margin.

You are correct, it doesn't.  Just like all other automation hasn't significantly dropped prices.  The reason isn't because it isn't genuinely cheaper, the reason is the extra value goes into profit margin.
Completely robotic farming infrastructure already exists:

http://farmindustrynews.com/precision-farming/20-technologies-changing-agriculture#slide-0-field_images-45641

If it wasn't more cost effective overall, people wouldn't be investing the capital into these systems.
"Farmers, ranchers and other agricultural managers will see the steepest decline of any employment category by 2020, losing a projected 96,000 jobs this decade out of 1.2 million positions, part of a broader trend toward less labor in the sector, according to the Bureau of Labor Statistics.... Farm owners like the Liefers are able to manage larger tracts of land without hiring overseers... Agriculture is an expanding sector with rising profits, even as overall employment, including laborers, is projected to drop 2.3 percent over this decade. "

http://www.bloomberg.com/news/2013-01-30/record-profits-no-job-creator-on-farms-as-owners-automate.html

[theadvicist]

Invest in companies that do robotics and automation, of course!

I'm loving this thread. Anyone got any ideas of companies to invest in? I've got some capital burning a hole in my retirement account!

[Guses]

I can't believe that nobody mentionned the singularity yet!

If we establish that we are talking about a point in the future where robots are so advanced that they can do anything that a human does better, it would stand to reason that human conciousness could likely / would have already been integrated in the machines.

We would all be machines at this point.

[GuitarStv]

We already are machines.  Our hardware is biological rather than silicone and aluminum though.  (Indeed, many have already begun to incorporate silicone into their design).

[more4less]

Well, days of free market economy will be over for sure. Robots create supply, but not demand, so market will implode. Where do we go from here? It depends on how humane ruling elites will be
1. Society either turn in to communist state with mandated equal distribution
2. Dystrophy where top "1%" will live behind the wall surrounded with robot servants enjoying unprecedented abundance, while the rest of humanity will get by. And I don't know if they will be able to. While production capabilities of 1%'s robots is infinite, resources and energy are. Guess who wins in competition for those? Remember indians trading gold for glass necklaces? Or simply exterminated by more technologically advanced civilization? The only hope is that 1% wont consume too much because of its size.
PS: And we're not talking about machines becoming self aware. Imagine robots (no, not rising and slaughtering all humans), but demanding equal pay for equal work? If robot is smart enough to do complex work, why it can't realize that it's in the position of slave exploited by lazy humans? :)

[MoneyCat]

Don't forget that when the robots take over there will always be the option to live a life of crime (as long as you don't do something stupid like work a job where you have to submit your fingerprints.)  Unless Robocop is invented, you can easily get away with a lot if you are smart.

[gentmach]

I, for one, welcome our new robot overlords.

You beat me to it.

Do not fear the coming automation. It is inevitable. Do not fight it for it is for the best.

[jordanread]

CNN article/video: See Amazon's new robot army

CNN, super timely information. :P They've only been using these for about 1.5 years.  Although it does look like they greatly increased the number of them. Just a quick modification, and they can be used to bring beverages!!

Got me all excited to see, but then they are the same thing they've had for some time now.

[forummm]

It must be robots month at NPR:

http://www.npr.org/sections/alltechconsidered/2015/05/18/407648886/attention-white-collar-workers-the-robots-are-coming-for-your-jobs
http://www.npr.org/sections/money/2015/05/06/404701816/episode-621-when-luddites-attack
http://www.npr.org/sections/money/2015/05/08/405270046/episode-622-humans-vs-robots
http://www.npr.org/sections/money/2015/05/13/406461675/episode-623-the-machine-comes-to-town
http://www.npr.org/sections/money/2015/05/15/407086723/episode-624-i-waiter
http://www.npr.org/sections/money/2015/05/20/408292388/episode-625-the-last-job
http://www.npr.org/sections/money/2015/05/22/408834372/episode-626-this-is-the-end