Reflections on the future of work

The end of a year often encourages thinking about the future. My reflections here, however, are not purely objective as my daughter is expecting her first child and our first grandchild in Spring 2017.  Assuming that we are not wiped out by runaway climate change or nuclear war, how will the world of work look for that child when he or she becomes an adult?  

I believe it will be radically different and that we have to start planning for that future now. Sadly, I see no evidence whatsoever that today’s politicians, whatever their leanings, understand this and they remain stuck in a 20th century mindset. 

I came into computing in 1970, part way through the first ‘information revolution’. New technology always changes the world of work and the introduction of computers into industry displaced a large number of people from clerical jobs. However, 1970s computers needed a lot of support and new jobs were created. This, along with rapid economic growth meant that that we did not see a large rise in automation-created unemployment.

The second information revolution happened in the 1980s with the advent of microprocessors. Not only did this lead to personal computers, it also made manufacturing automation economically realistic for all sizes of business. This led to a loss of manufacturing jobs but PCs made the expansion of the service economy possible so many new jobs were created. However, those displaced from manufacturing generally had the wrong skills and background for these service jobs. A new workforce was created, which had a far higher proportion of women. In many traditional manufacturing communities, dependent on heavy industry, the women rather than the men became the main breadwinners. Some of these communities have never recovered from the disruptions that this caused.

The third revolution, which is still going on, was facilitated by the ubiquity of the Internet. The made globalised supply chains possible with the displacement of the majority of the remaining manufacturing jobs from the West to Asia. More skilled jobs were lost but the Internet allowed a further expansion of service industries. Again, we saw widespread changes in employment but not a net loss of jobs.

So, successive automation revolutions have created as many jobs as they have displaced. Many commentators therefore predict that the next emerging information revolution, facilitated by artificial intelligence will have comparable consequences. Yes, jobs will be lost but new jobs will be created partly in new areas and partly because of the economic growth stimulated by new technology.

Whilst I’m all for optimism, I think this view of technological change is unrealistic. The fourth information revolution will affect the service sector and the remaining manufacturing industry and vast numbers of people will lose their jobs. Initially, this will largely affect the least skilled jobs – Amazon shelf pickers, supermarket checkout operators, call centre workers and taxi drivers. But by the time my grandchild reaches adulthood, the effects will be felt in skilled jobs that require deep information processing – lawyers, doctors, HR managers, accountants and so on.

Anyone who denies these job losses will happen is simply kidding themselves. The key question is not ‘will we lose jobs’ but ‘can we create new jobs to replace these’. Of course, new jobs will be created – just as the internet led to jobs in search engine optimisation and web design, the AI revolution will create jobs like AI managers and regulators, robot teachers and so on. Hopefully, technology-faciltated economic growth will be used to increase the provision of socially valuable jobs in areas such as healthcare and education and will support the creative industries.

But there is really nothing on the horizon of the scale of the service industry. I find it very hard to believe that enough jobs will be created to replace those displaced. Remember that, while technology may not replace highly skilled jobs, it will create definitely create assistants for them. Take education for example. I think we’ll have the opportunity to improve education but this may not lead to more teaching jobs. Instead, automated systems will take over lots of tasks such as grading assignments leaving teachers to the arguably more important jobs of stimulating their students’ interest in a subject.

I don’t believe that my grandchild will inhabit a world of robot teachers and machines that care for the young and old. Whether automation will ultimately make these possible I don’t know but I believe that, as a social species, we will reject these technologies. But for sure, he or she will have to be highly educated and adaptable to find a niche in 2040s society. And, they will have to change their job regularly throughout their working life (as I have had to).

It is a truism of course that all predictions are difficult, especially those about the future. We can’t say for sure what the future will bring and what new technologies will emerge. But we know from our history of computing technology that the changes will be profound and they will take place over 10s not 100s of years. If our children and grandchildren are to have a future where they are happy, healthy and fulfilled, we have a responsibility to think about and promote the social changes that will be required to make this possible. 

2 thoughts on “Reflections on the future of work

  • Pingback:Rethinking education for an age of automation « Systems, software and technology

  • February 2, 2017 at 10:14 am

    An interesting post. I agree with this sentiment and I’m glad to see you offering an opinion that is not extreme (one way or the way).

    Two things that I’d like to add regarding employment though.

    First, many many decardes (since the start of the industrial revolution), economists have been predicting the end of the ‘working week’, where so much will be automated by machines and we will all only have to work 15 hours per week. Yet, we are in a world where (in the west at least) people are working longer hours than ever, many in careers that did not exist 50 years ago.

    Second, and related, as old jobs become displaced by automation and AI, we need not just consider the jobs that are created directly from this (that is, those jobs involved in the creation and maintenance of automation), but also indirectly from both the hard resources and labour that are freed from this. Automation almost always implies cheapest final products, which frees up purchasing power of consumers. If I paid $X dollars for this last year but now $X/2 this year, then I have a further $X/2 to spend on other things. I can spend that on new activities that would otherwise have not existed. This is why we see more jobs in areas now that machines cannot do; for example leisure activities. Entertainment was a thing only for the ruling class until quite recently, but now, most people can afford to watch live theatre. This is simply because > 200 years ago, we needed all hands on deck just to grow enough food and make things to survive. I suspect we will similarly see new types of careers and expansion of existing ‘leisure’ careers from the resources that are saved by automation.

    Like you said in your other post on this topic though, it will not be an easy transition for some, and we should not forget that! 🙂



Leave a Reply

Your email address will not be published. Required fields are marked *