Thursday, 19 June 2014

Robots and the future of work

NESTA's book "Our work here is done: visions of a robot economy" was launched yesterday. I had the honour to contribute a chapter to the book, and also to speak at the launch yesterday.

The launch, hosted by Stian Westlake, featured presentations by Ryan Avent, Izabella Kaminska, Ellie Truitt, Nick Hawes and the awesome Carlota Perez, and demonstrations of real live robots. I think the robots rather stole the show - perhaps that is a sign of things to come!

The text of my speech at the launch is below.

Robots and the future of work

Last September, a research paper by Frey and Osborne scared the world. It concluded that as many as half of all American jobs could be automated. Not content with the loss of jobs from offshoring, the capitalist system now threatened further destruction of the American way of life.

Not surprisingly, the “robots will eat your job” movement went into overdrive. The Atlantic listed some of the jobs most at risk:
  • telemarketers
  • sewers ( though I thought all those jobs had already gone to Bangladesh)
  • insurance underwriters
  • mathematical technicians (watch out, quants!)
  • tax preparers and accounts clerks
  • librarians
But they also listed some of the jobs least at risk. Social workers. Therapists. Managers. Fire fighters. And, for some reason, dentists.

What is interesting is the nature of these jobs. The jobs most at risk from automation are those that don't involve the creative, the unpredictable and above all, humans. The jobs least at risk, by and large, are those that do. Creative jobs also feature in the “least at risk” category.

But of course both lists are jobs that exist now. Yet the jobs of the future may be very different. It's very hard to predict what jobs that exist now will still exist in 5 years time, and even harder to foresee what jobs there will be in 50 years time that don't exist now. When I was a small child, no-one could have foreseen that 50 years time I would be writing articles using a computer and distributing them on something called the Internet.

And yet – some things never change. Although the medium I use to write did not exist 50 years ago, writing did. So did reading. So, thankfully, did imagination. A robot could attend a conference and report on the proceedings: that type of journalism is one of the jobs most “at risk”. But creative, thoughtful writing remains the province of humans, and gazing into my crystal ball, I predict that will remain the case. The core skills of reading and writing – perhaps redefined as “comprehending and communicating” - together with the disciplines of thinking and imagining, will remain at the heart of what humans do in the new world of work.

And so will caring. Indeed it is probably fair to say that “caring” jobs will feature largely in the workplaces of the future. But they won't be caring jobs as we know them now, largely concerned with looking after people's physical needs. Those aspects of care can, and in my view should, be automated. The care that humans uniquely give to each other is not fundamentally physical, though there might be a physical component to it. As any hairdresser will tell you, the most important part of her work is the coffee & chat with her clients. That's where a human adds value. A robot will do a perfectly good haircut, but it won't be an experience. The time will come when we will pay more to be served by a human, because we value the emotional interaction – even though the physical aspect could be better done by a robot. Indeed, perhaps the hair salons of the future will be staffed by humans who are experts in understanding their clients' needs and directing robots in how to meet them – an interesting combination of the creative, the social and the technical.

Indeed that fusion of creative, social and technical is likely to define the work of the future. And the young people of today already understand this. They may not turn in good results in traditional school subjects, but they are experts in the worlds of social media, crowdsourcing and imaginative use of technology. The skills that will be important in the future are learned through play, not formal education.

The days when “work” meant spending long hours pushing paper around or endlessly repeating the same part of a production cycle are over. Those jobs are going – not because we no longer need information to be manipulated, not because we no longer need to “make stuff”, but because robots are much more suited to these kinds of routine tasks than humans are. And to me, this is not a disaster. This is a wonderful opportunity. We can eliminate drudge work and free humans to spend their time interacting with each other, caring for each other and creating clever and beautiful things to brighten up everyone's lives.

Of course, this requires a shift of attitude. For much of history we have valued producing more than caring: caring jobs are poorly paid relative to producing jobs, or even not paid at all. But in the future, robots will be able to produce all we need far more efficiently and cheaply than human workers. The jobs in which humans will be needed, and therefore valued, will be the ones where robots can't cope. And there is one thing above all that robots cannot do – and that is understanding and responding to human needs and wants. That will always remain the primary responsibility of humans. Communicating, empathising and innovating will be the highly-valued skills of the future.

Our work here is not done. But it is set for fundamental change. In the post-technological world, the work of humans will be understanding each other, and in partnership with robots, caring for each other and for the wonderful world in which we are privileged to live.


  1. There is one problem with all this, which is that if humans retain the caring jobs, then a lot of caring jobs will be done by people who actually don't have all that much empathy or insight into other human beings. I am not so sure I want to be surrounded by "caring" people who are just over-blown social workers bent on "helping" me achieve whatever the then current PC lifestyle is.

    1. Then maybe developing social skills and empathy will be the key job of educators in the future.

  2. I'm deeply, deeply unconvinced by the "what jobs there will be in 50 years time that don't exist now" line of argument. Yes, I'm sure there'll be some. I don't believe for a second that there'll be anywhere near enough to go around. What's different this time is not just the ever-rising bar that humans need to clear to compete with machines,but the sheer rate of change - new jobs will likely be automated faster than people can retrain for them.

    There's a lot of merit to your shift toward caring jobs, but even that isn't a given. I think you vastly underestimate the extent to which humans can feel empathy with nonhumans. No, it's not "real", but that's not really the point. A cartoon deer isn't real, but millions of people sobbed themselves silly at screenings of 'Bambi'. And robotic carers may not be anywhere near as effective as good human carers, but they could easily be a lot better than bad human carers. No neglect, no loss of privacy or dignity, no abuse. Japan in particular, with its ageing population and antipathy to immigrant labour, is investing heavily in this area; I wouldn't bet against them.

    More generally, I think that the only hope for a non-dystopian future lies in some form of basic income. What scares me is that I can't see any route to get there that doesn't involve wholly implausible levels of selfless philanthropy on the part of our current economic masters.

    1. I think your notion of "caring" is rather narrower than mine. I did say in the post that caring for physical needs should be automated - indeed I have written about this before in relation to elderly care (see my post "In the countries of the old" on this site). But "caring" means much more than that. The "gentlewoman's companion" of old worked in a caring role, but she did not do physical work - that was the job of maidservants. Her job was to provide human companionship. And that, I submit, a robot cannot do. I suppose, in some kind of Asimov future, we could have robots with whom you can have a civilised conversation and a shared joke, so the "companion" of the far distant future might be a R. Daneel Olivaw. But that raises far bigger ethical questions, since for a robot to be a really effective human companion it would probably have to be sentient. We are still a very long way from that.

      I agree with you about basic income. Indeed the entire panel from the book launch agrees with you, and the American economist Noah Smith in the book suggested not only a basic income but a universal capital inheritance, too. I also agree that we are a long way from political acceptance of either of these - indeed in the introduction to the book Stian Westlake points that out. But that doesn't mean we should give up. Radical social and political change is needed.

  3. Unusually rosy-visioned for you, FC. I think the debate will turn to "What is PROPER work?" and "How should we share the wealth (and why should we, anyway)?" and "What are we going to do with all those useless people?" The current angle is to demonise and persecute the benefit class. Unless the political-business class is converted to your vision we face a period of ugly and cynical top-down-initiated divisiveness.

    1. This is, of course, a seven-minute speech, so it's hardly exhaustive. My chapter in the book discusses the nature of work in a robot economy at some length, and indeed I have done so previously on this site. I don't accept that people are useless, but I have also discussed in previous posts both on this site and at Pieria the question of how people can live and how wealth should be distributed in a robot economy - though I haven't finished with this subject by any means. But yes, this is a vision. If we don't articulate the vision, how can there be hope?

    2. > "we face a period of ugly and cynical top-down-initiated divisiveness"

      I think you're right, but it's a bit ironic in that the bottom will probably suffer at least as much as the top from a basic income, at least in the short term. BI is supposed to replace all existing benefits, but there's no way it could possibly cover the vast sums of housing benefit being paid out in places like London. If you thought the uproar over the bedroom tax was bad, the uproar over the effective mass eviction that would ensue will be much, much worse.