Are we really headed towards a future without jobs? This is THE question of our time, everyone is paying attention to the remarkable explosion of progress in technology. It’s a question that’s been asked so many times in the past, perhaps this time we should be asking “is it different this time?” The fear that automation might displace workers and give rise to mass unemployment goes back at least two hundred years to the Luddite revolts. It’s an evergreen concern.
Is it different this time?
Most people haven’t heard of the Triple Revolution Report. A prominent report put together by an expert group including Noble Laureates and presented to the President of the United States. The conclusion of the report was that the US was on the brink of a huge social and economic upheaval because industrial automation will put millions of people out of work. The report was delivered to President Lyndon Johnson in March 1964. Over 50 years ago and while broadly speaking a degree of automation has occurred, it hasn’t really had the effect predicted. It’s fair to say that the alarm has been raised repeatedly, but thus far it’s been a false alarm. Indeed, what has happened so far is that while recent technological advancements have indeed decimated whole industries and killed off occupations, at the same time progress has lead to entirely new industries and occupations.
One branch of thinking is that as the current technology, automation and AI, advances new industries will rise, and new occupations will replace the old ones. New kinds of work will emerge, ones that today we can’t even imagine. This has been the story so far and it turns out that the new jobs that have been created have been better than the old ones. They’ve been safer, more engaging and more comfortable. And of course, they have paid more.
However, for a particular class of worker, this story has been quite different. For these workers, the technology has completely decimated their work, and it hasn’t created any new opportunities at all. These workers are horses. 100 years ago the US horse population was c.22 million, by the mid-60s it was 3 million, with the majority of the 3 million not being ‘workers’. This gives rise to a question, is it possible that a significant number of human workers will be made redundant in the same way that horses were?
Clearly, horses are very different from humans but if we look back a hundred or so years and look at the work they were doing, we can see that this work is now replaced by technology. We still talk about engines in terms of horsepower, that’s the power that an engine can generate. Horses replaced human energy and strength and then technology, in the form of engines, replaced the energy and strength of horses. The problem for horses in this instance was that they couldn’t adapt to changing circumstances, there’s nothing else that they could usefully do beyond adding value to human recreation.
As humans, this is precisely the issue that we face today. The machines that will replace humans in the near future are not just replacing our energy and strength, they are replacing our thinking, learning and our ability to adapt. They’ll do all of this better than we can. Machines are now encroaching on the very human qualities and strengths that have protected us thus far. These are the things that have made us relevant to the current economic model.
3 key reasons why it is different this time
There are 3 key reasons why the impact of technology will be different this time.
Firstly the kind of progress we are talking about is truly exponential. Moore’s Law, specific to microprocessors is one indicator but progress is much more broadly based than that. We’ve seen around 30 doubling in computational power since the late 1950’s when the first integrated circuits were fabricated. We are now at a point where we are going to see an extraordinary amount of absolute progress. In the coming decades we’re going to see things that we’re really not prepared for. Things that will astonish us.
The second key reason is that the technology, in a limited sense, can ‘think’. They can make decisions, learn and solve problems. The most important force behind this ‘thinking’ is a technology called machine learning.
Machine Learning is at the root of advancement
Machine Learning is the most highly disruptive, scalable and powerful technology in human history. One excellent example of what is achievable is what Googles Deep Mind division was able to achieve with its Alpha Go system. The ancient game of Go has an infinite number of gameplay configurations. There are actually more game play possibilities than there are atoms in the universe. What that means it that it would be impossible to build a computer to win at Go in the same way that it could win at Chess. The limited number of moves and configuration in chess means that winning it, for a computer, means brute force, i.e. utilising the power of the computer to run through every single scenario every time it takes a turn. A top level Go player finds it impossible to articulate what it is they are thinking about when they play. They play by intuition and their learned sense of the game.
You’d have thought that playing Go, because of the very human traits it requires to play it well, would be safe from automation. We are currently tending to draw a very distinct line between the jobs we imagine that AI might do in the future and those that are safe. The routine, mundane and essentially predictable jobs are ripe for automation. Current estimates are that these jobs are c.50% of the economy. On the other side of the line we imagine are unique to humans. The fact that Google solved the playing and winning the game of Go blurs the line in terms of jobs security.
It’s not just human strength and energy at risk this time
These emerging technologies are rapidly climbing the skills ladder. They’re not confined to purely threatening blue collar, low skilled jobs that require little in terms of education. Accountants, Lawyers, Analysts, Journalist, etc are all threatened and are likely to be challenged in the future. As a minimum, we are going to face a future of high unemployment. The UK unemployment rate in June 2017 is 4.3% and it’s hard to see if it will go any lower. That isn’t the lowest it’s ever been but we might actually be seeing the last of these high rates of employment.
High rates of unemployment, should this occur, with put a huge amount of stress on society. For example, we’ll see the opportunities for self-actualisation diminish right in front of us. This may well lead to an epidemic of depression. There will also be a huge economic issue. Gainful employment distributes income and therefore purchasing power. It creates both supply and demand. A vibrant market economy relies on lots of consumers who are capable of purchasing the products and services that are being produced. Without this, we face a declining economic spiral.
The question then becomes, what exactly could we do about this?
On the one hand the future looks utopian, it’s one where we work less, have more leisure time and spend more time with our loved ones. We can do things that we find genuinely rewarding, which initially sounds very appealing. But is it realistic to think that this comes without problems? There may well be a significant income distribution problem. A lot of people are going to be left behind. We are going to have to find a way of replacing the income achieved by traditional employment,
Could a future Universal Basic Income be the answer?
Universal Basic Income (UBI) is an idea that is generating a huge amount of interest right now. Essentially UBI would see everyone (that’s the ‘universal’ bit) being given a basic level income for the duration of their lives. There are various models but in the main it is not means tested, which means you’d get it even if you were working. This of course means that those actually in work would need to contribute more. The idea being that if you don’t need it you’d save it for the inevitable bad times and/or retirement. One of the main criticisms of UBI is that it removes the incentive for individuals to strive. This striving is one of the key reasons humankind has progressed in recent millennia. For example, imagine you were a student, struggling to get to grips with your chosen subjects and thinking about dropping out. Currently the thing that might keep you going is the vision of what your life would be like with no employment based on a poor education with no qualifications.
What would be our incentive in the future?
Now imagine a future where that vision is removed and no matter what you’re going to get the same income as those around you. The creates a perverse incentive for you to give up and drop out. We could tweak the UBI however to incentivise individuals to pay people that graduate more than those that drop out. Essentially we could incentivise those that align with societies vision of what is valuable. Some might see this as a scary thought. The current market-driven economy and society allow individuals, in the main, to align themselves with their own vision. They can see for themselves the dropping out would cause them pain, not because the State says so, but because the labour market says so. This incentive scheme could be extended to charity and community work. It could even be extended to environmental work. We could introduce a points-based system that recognises and rewards positive inputs aligned with whatever it is that society values. The issue here is exactly who is it that decides what is to be valued. If we can’t agree on the cause and effect of global warming for example, how are going to agree on what actions would impact change?
Technology will pull the cart in the future
We have a basic human need to find meaning and fulfilment and we need to occupy our time productively. How are we going to find that in a workless world? One of the main objections to UBI is that we will end up with too many people riding in the economic cart and not enough people pulling the cart. Clearly, in the future machines, AI & and Automation, are going to be capable of pulling the cart for us.