This post is co-written with Julia Begbie, who develops cutting-edge online courses as a director of a design college in London.
Some people (including us) think that within a generation or two, many or most people will be unemployable because machines will perform every task that we can do for money better, faster and cheaper than we can.
Other people think that humans will always remain in paid employment because we will have skills to offer which machines never will. These people usually go on to argue that humans will need further education and training to remain in work – and lots of it: we will have to re-train many times over the course of a normal career as the machines keep taking over some of the tasks which comprise our jobs. “Turning truckers into coders” could be a slogan for these people, despite its apparent implausibility.
There are several problems with this policy prescription. First, we do not know what skills to train for. One school of thought says that we will work ever more closely with computers, and uses the metaphor of centaurs, the half-man, half-horse creatures from Greek mythology. This school argues that we should focus education and training on STEM subjects (scientific, technology, engineering and maths) and downgrade the resources allocated to the humanities and the social sciences. But a rival school of thought argues that the abilities which only humans can offer are our creativity and our empathy, and therefore the opposite approach should be adopted.
Secondly, the churn in the job market is accelerating, and within a few years, the education process will simply be too slow. It takes years to train a lawyer, or a coder, and if people are to stay ahead of the constantly-improving machines in the job market, we are likely to have to undergo numerous periods of re-training. How long will it be before each period of re-training takes longer than the career it equips us for? And is that sustainable?
Third, reforming education systems is notoriously difficult. Over the years, educational reform has been proposed as the solution to many social and economic problems, and it rarely gets very far. Education has evolved over the last 100 years, and teachers are more professional and better trained than they used to be. But as the pictures above illustrate, most classrooms around the world today look much the same as they did 100 years ago, with serried ranks of children listening to mini-lectures from teachers. The fundamental educational processes and norms developed to build up the labour force required by the industrial revolution have survived numerous attempts to reform them, partly because reforming a vast social enterprise which looks after our children is hard, and partly because the educational establishment, like any establishment, tends to resist change.
It therefore seems unlikely that educational reform will be much assistance in tackling the wave of technological unemployment which may be heading our way.
And oddly, this may not be a problem. If, as we believe, many or most people will be unemployable within a generation or so, the kind of education we will benefit from most is one which will equip us to benefit from a life of leisure: education that is vacational rather than vocational. This means a broad combination of sciences, humanities and social sciences, which will teach us both how the world works (the business of science), and also how we work as humans – from the inside (the business of novelists, artists and philosophers). This is pretty much what our current educational systems attempt to do, and although they come in for a lot of criticism (some of it justified), by and large they don’t do a bad job of it in most places in most countries.
Although educational systems probably won’t be reformed by government diktat in order to help us stay in jobs, they will be reformed in due course anyway, because new technologies and approaches are becoming available which will make it more personalised, more effective and more enjoyable. Some of this will be enabled by artificial intelligence.
New-ish techniques like flipped learning, distance learning, and competency-based learning have been around for years. They have demonstrated their effectiveness in trials, and they have been adopted by some of the more forward-thinking institutions, but they have been slow to replace the older approaches more generally. More recently, massive open online courses (MOOCs) were heralded as the death-knell for traditional tertiary education in 2013, but they have gone quiet, because the support technologies they required (such as automated marking) were not ready for prime-time.
MOOCs will return, and the revolution which they and other new approaches promised will happen. We will have AI education assistants which know exactly which lessons and skills we have mastered, and which ones we need to acquire next. These assistants will understand which approach to learning suits us best, which times of day we are most receptive, and which times we are best left to relax or rest. Education will be less regimented, more flexible, and much more closely tailored to our individual preferences and needs. Above all, it will be more fun.
The main contribution of education to technological unemployment will probably be to make it enjoyable rather than to prevent it.