The Challenge of Automation

Obama gave a farewell address this past week and he addressed a variety of topics, including trade and the problems of jobs going overseas. The interesting thing he said, however, was missed by most commentators. He argued that…

“…the next wave of economic dislocation won’t come from overseas. It will come from the relentless pace of automation that makes many good middle-class jobs obsolete.”

This reflects my own views. Studies show that even when jobs have been repatriated, as some have, to put manufacturing closer to end-consumers for example, there are far fewer jobs now than there were a few years ago when the jobs went overseas.

Automation continues at a vigorous pace, and, driven by intelligent or cognitive software, it will continue even more rapidly in the years ahead. A recent report from Japan, for example, reported that 34 employees at Fukoku Mutual Life Insurance company were laid off after an IBM Watson AI application made them redundant. These employees were responsible for calculating payouts for employees with health claims. The application was not only capable of “reading” the documentation submitted, but could review medical histories, analyze photos and evaluate physician notes to determine the nature and severity of the problems being treated.   Moreover, the new AI application did the analysis and prepared the documentation in a much shorter time, increasing the overall productivity of the company’s response by 30%. The application cost about 200 million yen and, even with maintenance factored in, it will save the company about 140 million yen a year. Thus the application will generate a ROI in considerably less than two years. A leading Japanese research organization suggested, in a report, that nearly half of all jobs in Japan would be automated in this fashion by 2035.

If this is true in Japan, it is equally true in the US. Compared to this kind of automation, letting a few more factory jobs slip overseas, or keeping them in the US is really of little significance.  Moreover, knowing how important productivity and profitability are to most organizations, this type of automation will undoubtedly to be applied as quickily as possible.

It’s not our place here to consider the social implications and the possible social solutions, which lie in the realm of politics and not technology. Obama suggested some possibilities, and I suspect that he wasn’t thinking “big enough.” I suspect the problem will require serious taxation and a guaranteed income for all citizens. But, as I say that isn’t our problem. Our problem, as process practitioners and organizational change facilitators, is to begin thinking of where cognitive technologies can be used to facilitate automation.

Managerial and clerical work that currently relies on analysis and “low level decision making” seems to be the most likely target of opportunity at the moment. We need to start developing the analytic and process design skills to define how such work can be automated by more efficient processes supported by cognitive applications.



  1. Martin Hrabal says:

    I agree with the inevitable trend of automation. But I have different view on social implacations which can be called “machines take work from people”. In history we were witnesses of extinction of jobs and whole industries (candle makers when electricity occured, coachmen when automobiles appeared etc.). Nevertheless, these people always found employment in new industries.
    From economics we know that after cost reduction in a process (e.g. automation) profits will increase. But these profits are (mostly) spent or invested and thereby give rise to new entreprenerial opportunities and new jobs (where unemployed can transfer).
    So no “serious taxation and a guaranteed income”, but free market which would support innovation and movement of capital and people.

  2. Martin, I hope you are right. I know that you have been right in the past. When the Industrial Revolution began, some 70% of English workers were farm laborers. Some 50 years later most were working in factories, and automation made it possible for the remaining farm workers to produce all the food needed. Today, in the US, only about 2% of the population lives on farms and they produce way more than everyone else can eat. So the shift has certainly happened in the past.

    At the same time, its clear that as we become more productive, we free up capital that encourages the founding and growth of new businesses. That process continues today.

    The question is whether we can still continue all this is the future. There is some data suggesting that we are reaching some limits. I certainly hope increased productivity will increase jobs. In any case we need increased productivity to raise the standard of living for everyone, and we will undoubtedly find some way to accomidate the changes. As I suggested, that really isn’t the focus of this blog. As process people, we need to focus on how best to achieve the increased productivity.

Speak Your Mind