In The Second Machine Age, the authors refer to some research by Daron Acemoglu and David Autor. This 2010 research examined US and EU labour data and use of the concepts of routine/non-routine and abstract/manual work within it.
Acemoglu and Autor use the definition of “routine” for a job task that is “sufficiently well understood that the task can be fully specified as a series of instructions to be executed by a machine”. They explain that “(a)bstract tasks are activities that require problem-solving, intuition, persuasion, and creativity” and that “(n)on-routine manual tasks are activities that require situational adaptability, visual and language recognition, and in-person interactions. Driving a truck through city traffic, preparing a meal, installing a carpet, or mowing a lawn are all activities that are intensive in non-routine manual tasks.”
The findings of Acemoglu and Autor´s research into wage and job polarisation found that both types, manual and cognitive, of routine work are not faring so well, whereas non-routine in both are doing okay. Brynjolffson and McAfee refer to the classifications Messrs A² use in their research as a “two-by-two matrix” but then they failed to give me the picture in their book.
As I think about the jobs of the future and ready-reckoners for how to determine these, I am looking for ways to simplify this in my own mind. The 2×2 B&M suggested (but never drew) is a simple way to assess whether you need to be re-skilling because your role is in the danger zone of things likely to be comparatively easy to be done by a machine instead of you (or your successor). Use the more granular clarifications made by Acemoglu and Autor, and you can plot the dot your job makes. I have drawn the missing picture for my own interest. Next iteration will be a pirate´s map version. Kidding.
