When I was in elementary school in the late 80s, I vividly remember reading a text about the soon-to-be-redundant role of the coffee delivery person (a common institution in the Mediterranean and the Middle East) due to the rise of automated coffee vending machines (yes, vending machines were NOT science fiction in the 80s). The text made the emotional argument for the loss of income of an individual whose skills' future utility in the economy seemed dim. It also made the argument that machines don't offer the interpersonal exchange of a business transaction, an element vital to human wellbeing. I wonder if that served back then as a hint to me that they were, in fact, two separate products.
Fast forward thirty years and not only are automated coffee machines almost completely extinct, but we have more coffee shops, coffee delivery services, and coffee professionals than ever before. What happened? Why was the prediction of automation-induced mass unemployment wrong?
The trouble begins with our tendency to view the world ceteris paribus. A job becoming redundant because of automation does not mean that those redundant people will never be able to offer value to the economy. In the past, the vast majority of people were engaged in plowing fields. Now only a tiny percentage of people are involved in agriculture, yet we consume more quality food than ever and we have people working in jobs that our ancestors could not imagine possible. Obviously, we are missing something when blaming automation for unemployment, because if that were the case, with the rapid growth of technology, almost no one would be able employable today since few people would be able to offer value in today's technology-driven economy.
And this is the point. A job is not something the economy gives you. A job is simply the relative perceived value you can provide to the economy under given circumstances. It's what you can give to other humans that they value and are willing to reciprocate.
It is true that technology and automation change the parameters and criteria by which we value labor in the world economy. However, is that necessarily a bad thing?
Let's go over the argument that automation makes people redundant. Let us take the example of the accountant. In the near future, it is projected that automation will render most accountants obsolete. Yet, can you have accounting without accountability? Won't you still require someone's signature and liability (skin in the game) to ensure that numbers are as authentic as possible? Won't you still need supervisors to supervise, monitor and maintain automated accounting systems so that they don't run wildly out of control? Won't you need new kinds of job descriptions to describe the new positions created for those who support such automated systems?
Yet it's true, automation does cost some people their jobs in the short run, as it makes then irrelevant. Does that mean that they will never find work again? Far from it! They will be able to enter a more vibrant job market because the increased profitability from automation allows employers to spend more elsewhere, or to increase their capacity, or invest in research for innovative products that require new types of job descriptions. This means that the economic cycle, which definitely does not shrink due to automation, becomes enriched with more activity, more volume and more velocity, either in the form of more goods and services being demanded or in the form of new goods and services demanded and supplied due to automation.
Let us now take the example of music. When the gramophone was invented, musicians felt that their trade would become obsolete in an era when the service of music could only be provided by live performance, something only very rich people could afford at the time. Plus, music was never enjoyed at the consumer's convenience, even for the richest; it's not like they had an iPod and headphones in their pockets to listen to whatever they wanted whenever they felt like it. The gramophone question was this: Why would anyone pay a musician when music could be recorded and reproduced endlessly by a machine? So they logically (at first) felt that the demand for musicians would decrease drastically. Yet today, not only do we have available to us almost infinite amounts of music to choose from, live or otherwise, but the average musician gets paid more than ever and even the poorest among us can afford to buy and enjoy music, not to mention that any teenager with minimum music education can easily become a bedroom music producer and contribute with unique creativity. We have more musicians, more music professionals, better wages, and lower prices in music than ever before, all because of technological automation.
So why didn't automation destroy music? Easy. Because technology enabled the industry to grow. A gramophone made an alternative to live music performance affordable. This was translated into a drastic increase in demand. Demand for live music may have been reduced at the beginning, but more musicians had an option to compose, record and sell records. The increased competition further lowered prices, which in turn stimulated demand even more, perhaps disproportionally more than the lowered prices, meaning that increased music sales outweighed price reduction losses. Fast forward to the radio, cassette player and Napster, and today anyone with a $20 smartphone can enjoy more music than any aristocrat 100 years ago. And musicians on average tend to be much better off now than then.
Not only do we now have countless examples of automation actually bringing more to the collective economy while improving our lives (proof by correlation), but we also have the rational reasoning behind it (proof by causation). Automation has no point if the increased profit it yields cannot be spent. If employers can sack all of their employees for sci-fi androids like Data from Star Trek, what will they be producing and for whom, if most of the population is unemployed? It makes no sense for anyone to apply automation if that is what it meant. Increased profitability from automation WILL be spent, and it will be spent on people who earn. Money is like water. It always finds a way to fill the gaps, one way or the other, sooner or later. And even if we do reach a technological level where every single human task can be performed by machines, then we will either all live like gods in a money-less paradise with free abundant stuff where no one needs to compete for limited resources with anyone, or we will exclusively produce other commodities, such as philosophy, debate, comedy, art, ideas, and perhaps, hugs
And to answer the question "How can redundant specialists adapt to new technologies and market dynamics?", I offer a different approach: Perhaps it is the employers who will need to adapt first. They will have to. Otherwise, who will buy their products in a shrinking economy resulting from mass unemployment? Increased economic activity due to automation creates spikes in demand, in some fields more than others, and that increase in those fields will require more labor, and perhaps new types of labor too, labor that automation hasn't caught up with, or may never will. And it's the employers in those fields that will have to compete with each other to acquire top talent. In their efforts to compete, they will have to adapt their requirements. In the IT sector, for example, an extremely competitive field for employers, a university degree is not always required if a professional can demonstrate top talent. If the IT labor market was not competitive for employers, and labor demand was much lower than labor supply, employers would be requiring more and more qualifications to screen the multitudes of oversupplied talented professionals at their disposal. This example serves to illustrate how market dynamics can be a win-win if healthy progress is allowed to run its course.
One of my favorite pro-technology arguments is the Instagram phenomenon. Instagram gave the opportunity to almost anyone to become a model or photographer, bypassing industry gatekeepers and getting a chance at global exposure without the Harvey Weinsteins of the world at the auditioning desk. This did not deprive established models of their work. Quite the contrary! Instagram models enjoy an alternative revenue stream through affiliations and advertising. Consumers get hooked to consuming more eye candy because it's so much more available and accessible now, thus making the pro model service more desired. And with an accelerated circulation of money due to new "jobs" like these being created, we all enjoy more buying power with our money due to that growth. This way, models are much more employable and much more niched than ever before. It's a win-win!
To sum up, automation does not bring unemployment long term. It may cause some jobs to be lost in the short term, but it is a necessary bump towards collective (or individual) progress and development. We should not fear automation, even though it can cause a frequent reshuffling of the employment market, causing uncertainty and fear. However, we should welcome it as a necessary hassle, like visiting the doctor.
The healthy market will not let society suffer due to automation. If automation caused massive unemployment, employers would be the first to suffer as their products and services would be demanded by fewer buyers (employed people with disposable income). And the market tends to balance itself because people are logical, industrious and awesome. We always find ways to create new products and services with each need of ours that is being satisfied by automation. we are even good at creating new needs for us! We've done that for thousands of years. And we've only just started! Don't believe me? Look at how the labor supply deficit in the IT sector suggests that the more technology improves the more people are needed.
Fearing automation is like fearing the dentist. It's perfectly natural to dread the sound of a dentist's drill. But we know that a little pain now will save us from a ton of pain later, not to mention it gives us that Colgate smile! (I am in no way affiliated with Colgate).
Article originally posted on Medium.
In my quest for philosophically induced breakdowns of cherished beliefs, motivated perhaps by the sheer challenge of confronting my own biases, I sought to identify cornerstones of ideological waves of thought, and then attempt to unleash the force of unsettling chaotic skepticism upon them. Because, beyond overdramatic flamboyant writing styles (such as my own), nothing does a greater disservice to a position than a bad argument. One such thought experiment was my skeptical analysis of the ‘infinite monkey’ theorem.
The ‘infinite monkey’ theorem asserts the hypothetical notion that a monkey typing randomly and unintelligibly for eternity could end up producing, to the exact detail, all the works of Shakespeare, given enough time (meaning, infinite time). This thought experiment is a basic argument for Darwinian evolution, as it supports the plausibility of randomly created complexity and order without purpose.
Even though the ‘infinite monkey’ theorem is widely accepted as logically sound, I will boldly (perhaps foolishly, who knows?) attempt to question its validity.
Let us re-examine the ‘infinite monkey’ theorem. It is the idea that infinite randomness will end up with infinite possibilities. I respectfully disagree, and here’s why: this theorem is based on the assumption that a monkey’s keystrokes are ‘sufficiently randomized’, and this rule is not random.
I don’t question Darwinian evolution here, but rather, I am simply skeptical (and cynical) of this theorem’s value as a valid argument for Darwinian evolution. The ‘infinite monkey’ theorem is a bad analogy because random monkeys do not type perfectly randomly. They type in the quite limiting manner which their clumsy thick fingers allow them. Therefore, given infinite time, we’d get infinite letter combinations of adjacent keys, and perhaps zero combinations of letters whose keys are of a certain distance apart on the keyboard (or old-fashioned typewriter — I miss those…). To perform otherwise would mean that we’re not dealing with a monkey but with a computer program that is ‘programmed’ to produce ‘sufficiently randomized’ letter combinations to ensure the creation of any and all sequences possible. However, this requires the setting of a rule, which in itself is not random. Infinite time does not guarantee infinite possible results. An easy way to see this is by picturing the infinite points on a ruler between 0 and 1. There are infinite possibilities of points on that line, but none of those infinite points will ever be point 1.01. Infinity, it seems, is finite.
To assume perfect randomization is to assume a purpose, which makes the ‘infinite monkey’ theorem self-defeating. Therefore, to end up with infinite random possibilities in infinite time, we require a rule, which in itself is not a random thing. This realization, like most of philosophy, serves to further confuse us in our search for the truth. But it makes the journey towards truth while sailing on the boat of philosophy much more interesting! :)
Article originally posted on Medium.