Back in the 1990s when we were in a recession, Jeremy Rifkin wrote a book called The End of Work, which argued that unemployment would remain high forever because of robots, information technology, yadda yadda, whatever. Cue the longest peacetime economic expansion of the century.
Now, we have a seemingly endless parade of books prattling on about how work is going to disappear: Brynjolfsson and McAfee’s The Second Machine Age, Martin Ford’s Rise of the Robots, Jerry Kaplan’s Humans Need not Apply, Susskind and Susskind’s The Future of the Professions: How Technology will Transform the Work of Human Experts (which deals specifically with how info tech and robotics will affect occupations such as law, medicine, architecture, etc.), and from the Davos Foundation, Klaus Schwab’s The Fourth Industrial Revolution. Some of these are insightful (such as the Susskinds’ effort, though their style leaves a bit to be desired); others are hysterical (Ford); while others are simply dreadful (Schwab: seriously, if this is what rich people find insightful we are all in deep trouble).
So how should we evaluate claims about the imminent implosion of the labour market? Well first, as Martin Wolf says in this quite sober little piece in Foreign Affairs, we shouldn’t buy into the hype that “everything is different this time”. Technology has been changing the shape of the labour market for centuries, sometimes quite rapidly. We will go on changing. The pace may accelerate a bit, but the idea that things are suddenly going to “go exponential” are simply wrong. Just because we can imagine technology creating loads of radical disruption doesn’t mean it’s going to happen. Remember the MOOC revolution, which was going to wipe out universities? Exactly.
But just because the wilder versions of these stories are wrong doesn’t mean important things aren’t happening. The key is to be able to lose the hype. And to my mind, the surest way to get past the hype is to clear your mind of the idea that advances in robotics or information technology “replace jobs”. This is simply wrong; what they replace are tasks.
We get a bit confused by this because we remember all the jobs that were lost to technology in manufacturing. But what we forget is that the century-old technology of the assembly line had long turned jobs into tasks, with each individual performing a single task, repetitively. So in manufacturing, replacing tasks looked like replacing jobs. But the same is not true of the service sector (which covers everything from shop assistants to lawyers). This work is not, for the most part, systematic and routinized, and so while IT can replace tasks, it cannot replace “jobs” per se. Jobs will change as certain tasks get automated, but they don’t necessarily get wiped out. Recall, for instance, the story I told about ATMs a few months ago: that although ATMs had become ubiquitous over the previous forty years, the number of bank tellers not only hadn’t decreased, but had actually increased slightly. It’s just that, mainly, they were now doing a different set of tasks.
Where I think there are some real reasons for concern is that a lot of the tasks that are being routinized are precisely the ones we used to give to new employees. Take law, for instance, where automation is really taking over document analysis – that is, precisely the stuff they used to get articling students to do. So now what do we do for an apprenticeship path?
Working conditions always change over time in every industry, of course, but it seems reasonable to argue that job change in white-collar industries – that is, the ones for which university education is effectively an entry-requirement – are going to change substantially over the next couple of decades. Again, it’s not job losses; rather, it is job change. And the question is: how are universities thinking through what this will mean for the way students are taught? Too often, the answer is some variation on “well, we’ll muddle through the way we always do”. Which is a pretty crap answer, if you ask me. A lot more thought needs to go into this. Tomorrow, I’ll talk about how to do that.