Robotics: can your work be automated?

A recent trend in Shared Services and Outsourcing is Robotics Process Automation (RPA). In the early days labour arbitrage and industrialized delivery models were used to optimize processes and bring down cost. Twenty years later we see that additional value needs to come from new initiatives and RPA is a good candidate to deliver onthis requirement. From the perspective of individual businesses this may be a promising trend, but what about the impact for employees? Or to make it personal: will you still have a job when robotics really take off?

A popular video on this topic is labelled ‘Humans Need Not Apply‘. It sketches the impact of automation, gives some great examples and compares the trend to the early days of industrial revolution. Self driving cars are close although we will have some technical challenges (snow, fog) and probably some liability and insurance issues to take care of. But then we can move quickly to an economy where a driver’s license is not sufficient to get a job. But apart from transport and logistics the trend also touches on medicine, industry and many other sectors. If artificial intelligence can win a game of GO from the world champion, then deep learning will quickly overtake the biological capabilities in analysis and diagnosis in other fields as well.

Different views on the same topic

The BBC published some research on which jobs can be automated in the next twenty years. According to this research the chance that the job of a Management Consultant will be automated is listed as 7%. Even so, with 45% of the total number of jobs under threat, this is a disruptive trend that will have a major impact for all our clients and the economy as a whole. How should we prepare for something like that?

Others are more optimistic about the future of human workers in an age of automation. In his TED talk David Autor suggests that our current worries are hardly different from the ones of our ancestors. The fact that we can’t predict the solution yet does not mean there isn’t one: