Disruptive Technology: In the next 20 years will there be widespread permanent unemployment?

Yes. Your assertions and reasoning are fairly accurate. There are two probable end game societal structures: extreme socialism or extreme kleptocracy. If the world keeps heading towards what the Nordic countries have done, it'll be great for everyone. That's the Star Trek outcome. If the world heads in the reverse direction, like China, India, Brazil, Russia and America, we'll end up in a poor position. This is the French Revolution outcome where a lot of people will end up dead. I hope for Star Trek but I do fear that we'll end up like France did during the French revolution. Yes I know about all the preceding economic arguments people use that are adequately detailed in Ryan's answer. For the most part they're true. They worked well while we automated repetitive and strenuous physical labor over the past 2 centuries. But they stopped being true when AI become good enough to automate low-mid level intelligence. Not everyone is equipped to be a doctor or engineer. For those people, things will get bad over the next two decades. For the rest, it'll eventually get worse. No job is safe and no industry is secure.

Yes, permanent unemployment is already upon us.  Right now the official US unemployment rate is about 7.8%.  Yet economists admit that the actual rate is more like 20%.  Thus 12% of the workforce has given up on finding a job.  After over 3 years of jobhunting, they fell off the rolls of the unemployed.  They are now permanently out of a job.  The jobs they used to have have departed these shores, bound mostly for China and increasingly replaced (even in China) by industrial optimization, remote telecommunication, and automation.  The bulk of unfilled US jobs are either subsistance level (less than $20k/year, AKA minimum wage with no benefits) or require advanced specialization like an engineering degree.  No amount of economic rebound is going to create high school level jobs that pay well AND can't be offshored or automated.  A most revealing statistic is this: in the past decade 50% of the US jobs that were lost paid well enough for a  decent middle class lifestyle.  During that same time frame, only 2% of new  jobs have paid as well.  Without a doubt, the american worker of tomorrow is in serious trouble. But the problem isn't limited to yesterday's semi-skilled workers.  Today's freshly minted college graduates are also faring poorly.  In 2013, 53.7% of college grads age 25-and-under are unemployed.  That's in an economy that's four years out of 'The Great Recession", which ended in 2009.  The weak economic rebound that we've seen appears to be structural; the world has changed.  In all likelihood, the US economy isn't going to get significantly better, especially not for the new underclasses. We don't have to wait 20 years.  Permanent unemployment is here today.

Im kinda with you on your points. We have been told for years emerging technology just increases our need for a greater education base. From what I can tell this fear is not warranted in a historical context. In the past we have had greater technological improvements than the computer, (this is an arguable point which I do not feel like arguing). However, technology does seem to impose chronic growing pains. Machines have replaced many field workers, and laborers over the last 100 years. Subsequently we have had significant spikes in unemployment to follow. I would also argue more people are employed today than 100 years ago as well. So this technology does seem to foster overall growth in the need for labor. There are other aspects to consider. In the US we have wars that may be winding down which will cause an influx of soldiers needing civilian jobs. Also hopefully the baby boomers will be able to retire soon removing some workers from the market. Perhaps these factors will negate each other. There was recently a discussion asking, what would happen if the US opened there boarders to everyone? Its a scary thought for a minute, but for some reason  I get the feeling that could create more jobs than any other policy. Ultimately, I am concerned about the points you bring up. However, reason tells me there is nothing to worry about.

They all sound like reasonable assertions and the scenario seems very possible. I also think that, if this scenario occurred, almost all of the fighting would be between different factions of 'have nots' for dwindling resources while the increasingly wealthy would become more insulated. I would love to be convinced otherwise.

Disruptive innovations always generate innovation gaps: periods during which productivity gains destroy more jobs than it creates. Eventually, innovation opens up new fields that allow the job market to catch up, and to resorb this innovation gap. A lot of universities (including mine) have been studying the history of innovation gaps, and most of the research concludes that the innovation gap that the digital revolution opened more than 15 years ago is the longest ever observed, and that it show no sign of recession. Actually the innovation gap is expanding. We may not know all the mechanisms, all the feedback loops that underlie the economy, but we can't afford not to consider the eventuality of massive long lasting unemployment. I have read in some answers that the only two possible outcomes are generalized kleptocracy or generalized socialism. I think there's only one possible outcome, and that it is neither of the aforementioned. Social unrest is endemic to generalized kleptocracy. It is way to unstable to be durably implemented. Generalized socialism is as unlikely. It's based on a centralized redistribution of wealth, which is equally unstable because centralized systems have a really poor capacity to adjust their state of homeostasis. In my opinion, what we're headed at is a collaborative distributed economy based on activities that can't be automated. An activity that can't be automated is something that generates a different value every time you perform it. That's the case with knowledge sharing, artistic creation, social bonding, and projects like Quora, Couchsurfing, Tsū, StumbleUpon, and any open source software you can name are exemple that this can work.

All those suggesting there will be more jobs created later on to absorb the displaced workers have not been paying attention.  There's a definite trend and it's been in place for a long, long time.When a new technology is created which saves some human labor, not one, but two changes happen.  First is the obvious one that people lose their jobs.  This happens in proportion to the labor saved.  An example is cars, which directly displaced some large percentage of humans in the horse & buggy industry and their supporting industries.  If 90% of jobs are lost, then this means 90% of those people are out of work.  Straight forward.When this new technology creates the new jobs, there are another proportional ratio of new jobs created.  For the 'cars' example above, new jobs were directly created in car manufacturing but also in many new supporting industries.  This happens in proportion to the complexity of the new technology.  Lots of car parts and lots of infrastructure to support cars and lots of regulations on them = lots of new jobs in each of lots of new industries.  Still sounds easy, right?And this sounds like all new technology eventually creates more jobs than it destroys, which actually has been the case throughout history.  The reason is that higher tech needs more basic level support to make it work.  Just picture how many entire industries are 100% dependent on the auto industry.  Everything from manufacturing to sales to marketing to delivery, maintenance, washing, bling stuff, insurance, roads, road maintenance, building and its supporting industries, regulation and it's sub industries, and all the finance related to all the above.The problem occurs when we look at those trends.  Historically, we've lost a tiny number of jobs in comparison to those gained and this is "proven" to be accurate by citing civilization's advancement and growth.  On closer inspection though, the numbers of jobs lost has had a growth rate that's exponential above 1:1.  In other words, we lose 101 jobs at some point and the next losses are 102, then 104, then 108, 116, 132 and so on.  The exact numbers aren't important (the 100 portion and the 1/2/4/8 portion) because so far, this total has been so low compared to the jobs created number.  The jobs created, however, is a linear ratio.  This has followed a progression similar to 10000, 20000, 30000, 40000, 50000.  As such, the creations has been way higher than the losses for a long time.  However, playing this out for a longer time we see that it doesn't take long for the first number to overtake the second.  When this happens, where each number started and how big they were are all unimportant factors.  The only important part is to recognize that losses have overtaken gains and will continue to do so even faster as time goes on.  All this mathematical 'theory' is just conjecture on my part but with it in mind, we can look at history to see how accurate it is.  From what I've seen, it's rather close even though it isn't just two smooth curves.  It's more like two jagged, spikey curves with the spikes often called innovation gaps.It was the 70's when unemployment first reared it's ugly head in the US (due to technology).  Most people don't recognize this but when CNC manufacturing arrived, it was quickly adopted and it put millions out of work.  Then those people displaced other jobs as they changed careers.  Of course, since this happened during a period of high inflation (or maybe it caused it), it's been overshadowed.  And then, later decades have covered the effect up because we had a 'recovery' and then a tech boom and then a war and another recession.  However, that recession of '08 changed the game again.  The distracted decades of 1980-2008 promoted many fluff jobs to help reduce unemployment.  These are jobs that really aren't needed but which get created to keep people working.  Movements like management training and just-in-time production ensured lots of people counting the beans two or more times.  When the recession hit, companies were forced to downsize and quickly learned just how few people they could get buy with.  Boom, all the fluff was gone.  And since then, they've built on that optimization by automating more tasks.  The other side of this is that as infrastructure and supporting industries begin supporting multiple new industries (the computer and internet industry is the infrastructure for millions of new business now and incurs no new costs on either end), the multiplier factor for job creation falls.So, jobs are being automated and companies are optimizing labor out of their processes, i.e. looking for automation.  This is the stage of acceptance of the very automation which is killing jobs and I see it only growing from here out.  The reason is that it's self supporting since technology has caught up to the demand.  If an automated method to do something is desired and well known enough, it now takes very little time until we see it happen.  Often, mere months.I love the story about the company that promoted their robots by a future competition at ping-pong.  They took too long before releasing the actual accomplishment that others actually beat them at it.  Even though this campaign only took about 3 months until they showed that their bot could "play" the game, the other guys that did it in their garage, showed that theirs could "beat a champion".  It kind of shows how a large, multi-national corporation no longer has such an edge over the average person.  Oh, and that technology is advancing faster than anyone would think.So, technological unemployment is definitely here to stay and speeding up, despite what old-school economists and business majors think.  But that doesn't mean we only have the two popular choices of how things will play out - socialism or civil unrest.  Rather, the economics will still work to balance things out because as jobs are lost, buying decreases and prices deflate and purchasing power increases ... which solves the problem... but then... quality competition increases and resource use decreases and money deflates and eventually money loses all its value.  At this point, we'll transition to a barter system based on personal characteristics (generosity, friendship, moral standing, heroism, honesty and maybe more) to use as a more casual system of account.And lastly, this may take a while or not depending on how we collectively fight it.  Without any safety net or with opposition to automation, we will find many people falling through the cracks of survival before it's complete.  Our current system simply can't handle permanent rising unemployment so it's probably best to transition our tax system to a transaction tax (to catch all labor-less production) and a basic income (to replace all means-tested welfare).

I suggest that the underlying problem is that labour rates in order to achieve any half way reasonable standard of living and pay all legitimate taxes are more, in total, than the average person can pay. In other words when disposable income falls to such a level, subtracting mortgage, commuting, car, food, child care, that the ordinary citizen cannot afford a gardener, a cleaner or a plumber, those jobs become in short supply, and those in them overworked. Meanwhile those who can only aspire to menial labour become unemployed.

During last two months  a landscaper guy could not find one day to fix retaining wall at my house. Last Friday he had a spare hour to take out several pieces of wood to be replaced, hopefully in the next month. My wife contacted a guy who is doing windows cleaning. He was able to come only a month later. The handyman who is helping us is always busy and, if it is not an emergency, requires at least one month notice. There is a shortage of people who are doing good job using their hands and I do not expect oversupply in the next 20 years.