Monday, July 25, 2016

What Happens When Humans Need Not Apply?

What Happens When Humans Need Not Apply
Last week we wrote about our trip to Africa which, given the state of affairs, was perhaps a little too upbeat and cheery. Life isn't all vacations and new experiences. Life is work, and disappointment, and setbacks, and sometimes dealing with hardship. Especially the kind of upcoming hardships that CGP Grey talks about in depressing-but-fast-paced-and-somehow-still-entertaining videos.

In case you hadn't heard, automation is coming for us...and probably not the kind of automation you're used to, either.

This post won't make a lot of sense unless you watch the excellent video from CGP Grey first so, here it is.

Depressing, right?

For anyone who can't watch the video right now, here are the highlights:
  • Automation of labor to this point had two important factors: 1) automation used mechanical muscles to replace human and animal muscles and, 2) it usually involved fairly dumb, specialized robots, like machines that are very good at building cars in an assembly line, but fairly terrible at doing anything else.
  • Two new aspects to modern automation may have drastic impacts. 1) There are now general purpose robots that can do most tasks they are shown and, 2) instead of mechanical muscles, engineers and programmers have created mechanical minds that can do the work of human minds, except they are faster, make fewer mistakes, and operate at lower costs.
  • A whole mess of our workforce is potentially at risk. People who drive vehicles of any sort are the obvious candidates, but people in the professions (like doctors or lawyers), white collar workers who type on computers for a living, and creative types are all candidates for automation. As a result, the newly unemployed may not be able to lean on the old advice to just get better skills or education and find new jobs.
  • New technology does not necessarily bring more jobs. It can mean fewer, and possibly lower paying ones, to boot. Just as horses used to have a ton of work 100 years ago and, through no fault of their own, were replaced by mechanical muscles, we humans may be replaced in large numbers without it really being the workers' fault, per se.
  • Automation is inevitable. It's the natural next step in innovating business processes and making them more efficient.
  • The devastating conclusion is that we need to prepare for a future in which drastically high unemployment is simply part of our new reality.
Of course, we have heard before that "it's different this time" when it comes to technology and, usually, the thing that was once novel and exciting just ends up being par for the course. Planes, trains, and trucks have changed the way we travel and get goods forever, but they generally have made life better for us and have provided more jobs as well. PCs and the internet have certainly been game changers but, again, they're generally good for humans and have created more jobs. Plus, in a lot of material ways, life isn't that dramatically different than it was in 1992 thanks to computers: I just get my information and entertainment from different screens now. (But as I am writing that notion on a blog, the irony has to be noted.)

Still, with any revolutionary change, stating that things will turn out the same way as they always have is not all that insightful. And in those rare cases of an actual once in a lifetime change, it's also probably incorrect. This time, with mechanical minds entering the workforce, it really might be different.

Let's assume that CGP Grey is right on both main ideas: that automation is inevitable, and that human employment is going to be dramatically and negatively impacted.

What should we do to prepare for and solve for unprecedented levels of unemployment?

Idea One: Basic Income
Basic Income is one of my favorite topics to discuss, even if I'm not all that much of a proponent. Basic income is, as far as I understand it, a system where individuals are paid a fairly decent income, just enough to live, regardless of how much or how little they already make, and with no strings attached or requirement to earn that income. Everyone just gets a living wage on top of what they already make.

I love the boldness and scope of the idea. You need brass balls to propose something like this in a serious way to governments, businesses, or voters.

While Basic Income is perhaps seen as a cute-but-not-serious proposal today, it provides an interesting solution to the problems posed by this new brand of automation. Grey posits that humans will not be able to compete with their automated rivals, in the same way that, no how hard they worked, horses could not compete with their new motor-driven counterparts. It's not a matter of laziness or needing new solutions: horses lose to their motor powered competitors because the animals are outmatched.

If we extend the metaphor, human taxi and Uber drivers just aren't going to out-compete automated vehicles, no matter how hard the drivers try. And it will be like that in a lot of other fields, too. Thanks to cleverly designed, superior technology, a lot of people will become permanently unemployable. Basic Income solves the "we have a mob of hungry citizens who want to break down your door and steal your stuff" problem.

Still, there's the pesky issue of how to pay for Basic Income. Since our government struggles with paying Social Security to old people who've paid into the system for four or five decades, this seems like a tough program to fund. The simple answer is to target the entities doing the automation: the businesses who formerly hired human employees, but now hire bots. And there's a nice symmetry of having businesses somehow pay for the living expenses of the labor force they're putting out of work.

But this taxation has some serious drawbacks, such as disincentivizing innovation, and potentially forcing companies to simply do their now-automated work outside the reach of our government (i.e. - abroad).

Idea Two: Widespread Frugality and Financial Independence
As naive as this idea is (most of us Americans are neither frugal nor are on the path to any sort of financial independence) spending less and untethering ourselves from employment certainly mitigates the problem of having too few jobs to go around.

If we all spend much less than we currently do then, by definition, we are less reliant on our current employment income. Assuming society had a sudden and large exodus of people from the workforce, and assuming the companies we invest in continue to do well in an age of widespread automation and reduced average incomes, then, hey, we might not need that many jobs anyway.

There's a tiny part of me that hopes that this niche financial revolution that Early Retirement Extreme and Mr. Money Mustache cultivated will reach a tipping point, and become somewhat mainstream. But this is America, home of the second car loan, the jumbo mortgage, and the negative savings rate. Widespread financial independence is such a long shot that it's hard to see it becoming a reality, even with my rose tinted glasses.

Idea Three: The Sarah Connor Solution
If we are ever to avoid this machine-centric future, we first need to convince the corrupt guards at our psychiatric hospital that we don't actually believe in Judgment Day any longer, so we can get to John and keep him safe.

Step two is kidnapping computer researcher Myles Dyson, taking his ID card, and sneaking into Cyberdyne headquarters. Then, we destroy the last remnants of the first T-800 and Skynet, before AI is realized and all of civilization is incinerated via nuclear war.

I'm obviously an advocate for financial independence and for living simply. That's a big part of why I write this blog, instead of just working towards these financial goals in private. My wife and I find a modicum of peace from living well while still within our means, and trying to build some financial security for our future. This kind of life stands on its own merits, and we need not be driven to good financial goals by fear.

But after seeing that kind of video, that depicts a future that seems both terrifying and strangely inevitable, we American workers with our cushy white collar jobs, sipping coffee and reading blogs on a Monday morning instead of doing our actual work, maybe we ought to get our financial house in order sooner rather than later.

*Photo of Baxter, the bot coming for your job, is from jurvetson at Flickr Creative Commons.


  1. It's an interesting thought that the financial independence movement might be a saving grace for certain people whose jobs are automated out of existence.

    What seems certain is that the return on capital is far outstripping the return on labor in the market. If this continues to widen, the owners of capital are going to do better and better vs the people who rely on their labor to generate income.

    But even if you don't believe the doom and gloom, wouldn't it better to have both capital AND labor at your disposal than just labor?

    1. Hey there, Biglaw Investor.

      It's an ongoing problem for me that some of the comments are more insightful than the post, and this is an example of that. I really wish I'd framed this issue as you had, and considered the value of labor vs. capital. That's really at the heart of this, and another example of how I'm probably not an economist at heart.

      Like you said, having both at your disposal is pretty great.

      I was arguing with a communist at a party a few weeks back, and tried to convince him of something similar: that working people really needed to invest, not only for their own financial well being, but because stock ownership is the means (at least in our society) by which labor ends up owning the means of production.

  2. Very Interesting! I agree completely with your comment about living simply, and DEFINITELY within your means and trying to max your savings rate. Thanks for the article

    1. Thanks, Passive Income Dude.

      The title of your blog probably gives a good fourth option, or at least a more actionable step than simply advocating for financial independence. Passive income as a goal is particularly cool, since it jives with our current mindset about earning an active income.

  3. I know in China, they've already order billions of money on Robots to the work for big companies such as Apple. so yeah it's coming

    1. Right: the scale of 'old' innovation, like the kind that assembles iPhones and iPads, is a real risk, too.

      What happens when bots start creating apps that replace Pokemon Go? Then we're really screwed. :)

  4. Or you could be someone who is really good at inventing robots to do other people's jobs so you have job security? :) I least I hope there will still be the need for humans on jobs. Maybe this is one reason I should be glad I'm 40-something? It probably won't happen in my lifetime. Anyway, I voted early retirement!

    1. I think that's definitely a good answer for some enterprising and techy individuals: become the engineer or coder whose job it is to create the automation bots. The rub is that, like so many of the new fields, there simply aren't that many jobs in the new and upcoming fields: certainly nothing to compete with the numbers of people in transportation, retail, nursing, etc.

      Anyway, probably something for the kiddos to worry about. :)

      Always glad to see a comment from you, Tonya, and thanks for tweeting out the post!

  5. While I agree that machines will replace many of the manual jobs in the future, I believe that many more tech jobs will be created. When you have more machines, you need people to debug them, maintain them, program them, etc. That's why the tech industry has been booming in the past decade, and the demand for software developers has gone through the room! I remember trying to get a co-op job in 2001, after the dom com crash, and it was BRUTAL. Now there are so many tech jobs, and not enough engineers. The irony is that I don't need them anymore :)

    I do believe that becoming FI will be a way to hedge out of the loss of manual jobs. Hopefully, as more and more FI blogs come online, and more and more regular working stiffs take the leap and PROVE that it can be done, the masses will see it's not that difficult.

    1. Hey there, FireCracker.

      I'm hopeful that you're right, and that needs for jobs like Computer Programmer (currently only #33 on list of most popular jobs) end up growing a lot faster than they currently are. Per the end of the video, new, techy jobs are certainly be created all the time, but the gross number of jobs available in those new fields pales in comparison to those old jobs ripe for automation, like transportation, retail, cashiers, etc.

      My assumption is that tech jobs may be unlikely to replace the jobs automation destroys on a 1:1 ratio. If the 3 million jobs in transportation are halved over the next decade due to automation, do we have the capacity to create 1.5 million tech jobs? I suppose we'll see, but my guess is that this sort of tech will have a net reduction in overall human jobs required.

      That's why I'm looking outside of pure job replacement as a solution, which ultimately might not be a sustainable approach. But if enough people approach FI or otherwise are able to live on some meager, basic income due to frugality, maybe high levels of unemployment will be acceptable...or even desirable.

    2. That may be true in the near future but what about in 10-15 years time when computers are smart enough to program themselves? :O)

    3. Exactly. The idea that there will necessarily be more jobs in technology, to make up for the jobs that are being automated by said technology, isn't necessarily true. There may be far, far fewer jobs available overall.

      More technology does not necessarily mean more better jobs for horses, or people.

  6. Come with me if you want to live! Great terminator it weird that the only reason I don't want automation progress and a.i. Is because of that movie?

    1. I'm so glad someone commented on the T2 reference. That whole section nearly didn't make the cut, to be replaced by some other boring-but-real idea for a solution.

      I do think the technoptimists are a bit too naive when it comes to risks of AI. I don't necessarily think "Judgment Day", but I wonder if Baxter will be all that pumped to basically be our robot slave.

  7. I am trying to understand what happens when there are machines doing all of the work, but very few poepol can afford the products and services they are producing. Seems that it will have to reach some balance like a natural ecosystem. Too many machines doing too much, so they have to turn off the machines or hire some humans, who can buy the stuff. Other side is when there are too many humans, automation increases.

    1. That's a good point, Cactus. I think the rub is that, while we won't ever reach a point of 0% human employment, the equilibrium you're talking about might have a very high level of unemployment. CGP Grey posits that it could exceed the levels of the Great Depression, where we have something higher than 25% US unemployment, but the other 75% of people are still clearly buying the goods and services from automated enterprises.

      Another (difficult) reality might be that automation hits US workers first and harder, because we make so much more. So companies here may simply sell their products/services worldwide to humans still working, while a higher number of US workers have trouble finding work. That is to say, the equilibrium we're inevitably going to find may still be very rough for us in America.

  8. Interesting read! I'm one of those optimists that believes that faced with adversity, the market can and will adapt in unique ways that we cannot predict.

    As I was driving through Oregon on a road trip over the last weekend, I found myself so frustrated with this ridiculous (IMO) law that doesn't allow competent motorists like myself to pump our own gas. At one station, desperately needing the restroom, I had to wait a solid 10 minutes before the minimum wage employee with a broken arm was able to get to me and start the flow of fuel. This is not an activity that requires assistance - and hasn't been for decades. Personally, I find that these attempts to preserve jobs and the status quo are naive and do more damage over the long term than good.

    I think the saddest part of our current society is the belief that government should prevent painful outcomes for all of us.

    Obviously there are pros and cons to this, which is why we are in the situation we are in - it's very compelling to help people who need it. But our direction over the past few decades seems to have led to a prevailing mentality that the government is the best resource for "saving" people and even that it is essentially the only go-to resource (despite the thousands of nonprofits, churches, and community groups who are getting very little credit for often doing a much better job across many metrics at delivering similar services).

    It seems we would rather pay more in taxes, and accept a lousy return on that investment in humanity, than give to nonprofits and churches who are turning those funds around to deliver direct services, often with a management cost of only 5-15%. You will never find that return in any government program.

    So, I vote for innovation, and pain. Pain is inevitable and the source of growth and new innovation. When we try to prevent pain from occurring, we interrupt the evolution of our society that needs to take place. Just as when we pop Prozac instead of dealing with our own emotional difficulties, we simply numb the pain rather than growing and fully resolving it.

    I'm sure we're all happy that the industrial revolution occurred - but that was a painful process for labor as well. One that had to occur to result in the modern society we now take for granted. My two cents :)

    1. Thanks for the awesome comment, Emily! I'll try to unpack some of it...and some I'm just going to leave alone. :)

      -For the example of the Oregon gas station attendant, I have to assume he personally is going to fare worse in a world with more automation than he currently is. Which is kind of stark, since this is the employment he's currently able to obtain, and it's not that great.

      -I hear what you're saying about government, but I personally think there are plenty of examples where government is way more efficient than the private counterpart (take the library, or public transportation). I've written about that here, but I don't think it's fair to paint with a broad brush and say government is not efficient:

      -Innovation has tremendous upside: in the case of automation, we can produce abundance for less cost. But as CGP Grey notes, there is nothing at all that says that more and better technology means more and better jobs for humans (or, as he notes, horses). Innovation can, and very well might, lead to much higher rates of human unemployment.

      -The scale of that unemployment very well could exceed the capacity of private charities and government programs combined. The dichotomy between the two might not be that meaningful in such an economic crisis. I'm sure, in that situation, we'll have all the pain we can handle as a country (and, looking for a silver lining, the motivation required to come up with innovative solutions).

  9. So....

    Bizarrely, I am what could only be considered a fiscally conservative Republican but I have come to the believe that in the long term a 'basic income' (though I appreciate the term Heritage Check as coined by Heinlein I believe) is the most likely solution to a variety of issues as we move forward.

    I spent several hours writing what could not possibly be considered a reply to a blog, but possibly some sort of presidential platform statement, and then realized we'd all be happier if I just went and had some M&Ms.

    Done and done.

    1. Hey Morgen! If you saved that presidential platform statement, I'd love to read it if you want to email it to

      I love that the idea of basic income would have appeal on the right, too.

  10. I love this topic! Thanks for bringing it up :)

    I've been reading loads about it recently including the mega scary AI post from waitbutwhy blog (check it out if you haven't read it yet) plus a book called Post Capitalism which reckons that automation is all of our golden tickets to emancipation from work, if we (i.e. our leaders) do the right things to take it of course.

    My thoughts are that a Universal Basic Income (UBI) is the only solution that will work on a grand scale.

    The issue with solution 2 is that people are not that forward looking and by the time people have lost their jobs the horse has already bolted. MMM has become absolutely massive already but does his advice really apply to taxi drivers? I'm not so sure. There is a limited scope for it and I don't personally think it would work if "everyone" did it, despite me reading the post he did on it (something about a frugal utopia I think - all sounded very nice but I just can't see that happening, not in my lifetime anyway).

    Going back to the UBI, once businesses are so efficient at making stuff without human help the cost of goods will start to be so low that the UBI can also be a very low amount, therefore the actual outlay from government should be affordable without hiking taxes up too much. They will also save a lot of money by slashing the bureaucracy of the existing benefits system which has many arms and is overly complicated (although again this will put loads more civil servants out of work, ironic huh)

    Also businesses must surely agree to be taxed to help for the UBI because otherwise no one will have any money to buy the goods they are producing. It's a catch 22 but in a good sense, so hopefully they will see the basic logic in that and not want to shoot themselves in the foot here and end up with the inevitable baying mob at their doors.

    Another point made in Post Capitalism is that when 3D printing and other related technologies really take off practically anyone can make practically anything in their own home so large businesses may slowly become replaced by a massively distributed production network. Similar to how you used to buy CDs that were manufactured in a large factory, but now just download an MP3, you will in future just download the instructions to make widget X to your 3D printer for pennies and there you have it. It effectively drives the cost of buying or making any extra widget (marginal cost) to near enough zero, in which case everyone is already so wealthy we won't have any problems any more.

    It all sounds very pie in the sky but also I can't help but believe something like that is possible in the very near future, the rate that tech is changing nowadays (again, see the wait but why post I mentioned, which actually thinks the biggest issue is not AI replacing jobs but getting so super intelligent and just destroying us - a bit like your scenario 3 but probably lot worse or at least in ways that are totally unimaginable to us)

    As I say I find this all very interesting - not got many (any) answers but I do enjoy reading and chatting about it :)

    1. I'll definitely give butwaitwhy a read -- I'm actually adding to the blog roll now (trying to get back up to 40) so maybe they'll make the cut. :)

      UBI is a very interesting solution. I think the concept of how to pay for it (and who should pay) is the big problem. Even if one country figures out a viable solution, the capital & owners are likely to move to another country. As with everything, you've got to follow the money.

      The more I get into personal finance, the less I see frugality is a cure-all. I think it's one tool in the bag, but there certainly seems to be a upper-middle class, well educated pattern that emerges when you see who's retiring early. Frugality alone probably isn't going to work for everyone in the US, and certainly not everyone worldwide.

      I like the optimism with decentralizing production, for sure. I'm probably a bit more cynical on that front but, hey, 3D printing is very new. Who knows.

      Like you said, not many easy solutions to these big honking problems. I think part of the allure of these discussions is that they're large enough and complex enough that only a moonshot sort of innovation is going to possibly have a shot.

    2. I was very cynical on a lot of things I read about in Post Capitalism but having read around for the last few months on things that are actually already out there such as self driving cars and what is very shortly going to be here with nanotechnology, AI and so on, I have kinda changed my mind. This isn't coming from crackpots with pie in the sky ideas, these are world class scientists telling us they are already doing it, look at some of the stuff coming out on TED talks for example. It's mind blowing! What's also great is a lot of these new ideas are coming from people who genuinely seem to just want to improve our lot in the world and are open sourcing all the ideas and tech. That is why I have stopped being (as) cynical on this stuff and so pessimistic about where we are heading as a species. Hopefully my newly found optimism is well placed. (of course all this new technology will surely bring it's own issues, such as T2000's etc... ack!)

      What a time to be alive :)

    3. Oh, I hear you on the techno-optimism. I realize this stuff is happening now.

      But like the CGP Grey video notes, self-driving cars, while awesome technology that assuredly will have tons of upside (no DUI with robot driver) can still have devastating downsides (no human drivers needed).

      I know this sounds cynical, but any sort of optimism is, by its nature, downplaying the downsides of things. Techno-optimism, while very cool, is likely to be somewhat optimistic in its prognostications, too.

    4. Yes of course, I'm with on that one. I'm sure there will be a whole heap of short term pain (as someone pointed out above).

      But the prize is practically free solar powered autonomous personal transport units instead of smelly self driven ICE cars, in about 15-20 years time though? How cool does that sound! :OD

    5. For me, I think it sounds great. For the taxi, Uber, and long haul drivers, it sounds like inevitable doom. :)