You Are Obsolete. (But That Might Be A Good Thing)

You Are Obsolete. (But That Might Be A Good Thing)

Like what you see? Get it straight to your inbox weekly! Sign up now.

A couple of years ago, I took notice when BART workers in San Francisco went on strike for higher pay. BART is short for Bay Area Rapid Transit, the commuter rail / subway service that serves most of San Francisco and the outlying areas. It has historically been a huge boon for its workers- it’s one of the only places left in SF where a worker can advance up pay grades and make a good living without a college education. One tech CEO, in response to the strike, famously said this:

“My solution would be to pay whatever the hell they want, get them back to work, and then go figure out how to automate all their jobs.”

This understandably did not go over well with the general population, with many people criticizing him for wanting to get rid of workers who might be older and unable to transition to another job- and others simply criticizing him for putting profit ahead of human livelihood in general.

Now, my first thought when I heard that was to agree with him. it would be wrong to put people out of a job simply for the sake of automating and cutting costs. But there is a flip side to that- not automating just because people need jobs seems like a ludicrous, almost cartoonish strategy that inhibits progress. Scott Alexander over at Slate Star Codex sums up the issue really, really well:

“if your job can be done more cheaply without you, and the only reason you have it is because people would feel sorry for you if you didn’t, so the government forces your company to keep you on – well then, it’s not a job. It’s a welfare program that requires you to work 9 to 5 before seeing your welfare check.”

if your job can be done more cheaply without you, and the only reason you have it is because people would feel sorry for you if you didn’t… it’s not a job. It’s a welfare program that requires you to work 9 to 5 before seeing your welfare check.

Unfortunately, he’s right. And that’s applicable to more and more jobs- already in the infancy of the 21st century, we have the capability to automate huge numbers of jobs. Major metro lines in cities like Copenhagen, Paris, Turin, Rome, Barcelona, and more are completely automated- meaning that not a single person is required to operate the train safely. Many, many others are capable of being automated, and essentially have conductors solely to press the buttons to open and close the doors- like BART and many lines of the London and NYC subways.

And that’s just the beginning. Right now we are capable of automating far more jobs than we do. If we had the political will, at this very moment we could be rid of most factory line workers, fast food clerks, bank tellers, gas station attendants, grocery store clerks, and more.

I’m not talking about next year, or in 2020. I’m talking right now, tomorrow. That technology already exists. The only reason companies aren’t doing it is because of the bad press they’d get or the upfront cost in automation. But the future gets even more roboticized if you look further out: many people estimate that self-driving cars will hit the streets by 2025 and be the majority of cars on the road by 2050. That means truckers, taxi drivers, bus operators, and limo chauffeurs are all gone.

But the future gets even more roboticized if you look further out: many people estimate that self-driving cars will hit the streets by 2025… that means truckers, taxi drivers, bus operators, and limo chauffeurs are all gone.

It might not even take that long. Some scientists are predicting that over half of US jobs will be lost to automation in just 2 decades- by 2035.  And that’s not even bringing into account how automation will affect other developed countries similarly- we can expect the same kind of trend to happen to EU countries as well and even in developing countries, as automation becomes cheaper than outsourcing (Foxconn, the company that manufactures most of Apple’s products in China, has made no secret of the fact that it wants to deploy over a million robots in its factories in the coming years).

Labor has always been a pretty easy (relatively) question of supply and demand- you want to have more jobs (demand) than you have workers (supply). But that whole process becomes fundamentally broken when automation comes into play.

Or, to put it a different way: in a supply and demand economy, what happens when supply is infinite?

Or, to put it a different way: in a supply and demand economy, what happens when supply is infinite?

This has been a famous problem in economics for decades, and it has a fancy name for itself: it’s called a “post-scarcity society”, and the branch of economics that deals with it is called “post-scarcity economics”, since such a society requires a radically difficult approach to its economics. It cropped up as early as 1997 with Aronowitz’s seminal “Post-Work: The Wages Of Cybernation”, a collection of essays that seem quite prescient now, even though they’re decades old, as well as bleaker looks at it, like Bookchin’s “Post-Scarcity Anarchism”.

A number of different economists have approached the question with varying levels of skepticism and optimism. In our current system, of course, it would be a disaster- as jobs disappeared people would be out of work, and more and more wealth would become concentrated in the hands of people who were lucky enough to have jobs that are hard to automate– for the time being. Almost anything can be automated, given time, and even doctors and web designers may find themselves obsolete soon.

You can already see this happening- wages have remained relatively flat for low-skill jobs, even as productivity and profits per person have soared. Individuals are seemingly more productive than ever before- and yet they’re making less. That’s because it’s not individuals that are getting more productive- it’s that they’re using the machines that are starting to replace them.

Individuals are seemingly more productive than ever before- and yet they’re making less. That’s because it’s not individuals that are getting more productive- it’s that they’re using the machines that are starting to replace them.

An easy, attractive answer that some people have clung to is to try and simply stop the progress of automation- essentially to do what BART is doing above, and keeping their workers on even if they don’t need them. This has its own problems, not least of which is Scott Alexander’s objection at the beginning of this article to what he calls “sneaky welfare”. The other objection is that this won’t work- and we’ve proven it doesn’t work- it’s impossible to stop new technologies from changing the workplace no matter how hard you try.

The Luddites tried to stop the Industrial Revolution, and they are remembered as an insult for their troubles. Once a new, productive technology exists, it will always gain ground and traction because of the potential it has to out-compete anyone who doesn’t use it. At best, you can delay its adoption- at worst, you will be mercilessly steamrolled as everyone else adopts the technology and your company withers and dies out.

At best, you can delay its adoption- at worst, you will be mercilessly steamrolled as everyone else adopts the technology and your company withers and dies out.

So what to do?

A bold, controversial answer seems to be taking hold among most economists- even ones who don’t like it admit it might be the only feasible way out. And the answer is this:

We will need to start paying people not to work.

That is a very hard sentence to accept. For the entirety of human existence, work has been coupled to income. You work, you get paid- you live. It’s as simple as that. There feels (at least to me) something very wrong about paying someone to sit and do nothing all day, watching TV and eating Cheetos.

And yet it might be the only way to not end up in a brutal, poverty-stricken dystopia. Even the authors above, Brynjolfsson and McAfee, concede that people will become mostly idle- even though they don’t seem to like the idea.

We will need to start paying people not to work.

This solution is called a basic income. On its surface, it’s a very simple idea: everyone gets a fixed amount of money every year with no testing or work requirements necessary. It’s an idea that is gaining traction among worldwide economists for combating the problem of workplace automation, and there are even some Nobel laureate economists who advocate it as a solution.

(Fun fact: In Gene Roddenberry’s Star Trek, this is how the futuristic United Federation of Planets economy operates. It was one of the first mainstream depictions of a post-scarcity economy, and it seems as though we might get there centuries before Roddenberry thought we would).

It’s a very early prototype idea at the moment, and there are differences in opinion as to how it would be implemented, but essentially the idea is to provide a small basic income to people that would allow them to live. If they wanted to simply sit and do nothing all day, they could- but if they wanted to challenge themselves and become artists, volunteers, or writers, they could do that as well.

It’s an interesting way to view the rise of the machines- not as job eaters, but as freedom granters. Instead of a bleak world of chronic poverty and unemployment, it offers us a picture of a utopia where humans are free of the burden to work for a living, able to pursue their innermost desires safe in the knowledge that their kids won’t be out on the street if they fail to make an income.

It’s an interesting way to view the rise of the machines- not as job eaters, but as freedom granters.

Of course, it’s not a certainty that we need something like this. Some economists are even doubtful that the “Machine Age” will displace workers at all. Some economists believe that jobs will simply migrate to new sectors- much like in the Industrial Revolution, jobs didn’t disappear- they simply moved to different sectors as we switched from an agrarian / artisan economy to an industrial one, and as we switched from an industrial one to a service one with the information revolution.

Their position is that something similar will happen during the machine revolution, and that no adjustment will be necessary- but that position is looking increasingly shaky as the evidence piles up that automation will be far more effective at eliminating jobs than any innovation that came before it.

So in the end, I suppose, it’s up to us. Will we allow the machines to usher in an era of brutal unemployment, or will we work together to bring about an age of utopian post-scarcity where humans look to the stars, unburdened by the stress of having to work?

I think I know which one I’m rooting for. How about you?