The Schumpeter column in The Economist, a newspaper, yesterday reported that
One of the most widely quoted statistics of recent years….. that 47% of American jobs are at high risk of automation by the mid-2030s
taken from a research paper published in 2013 by two Oxford academics, Carl Benedikt Frey and Michael Osborne has been widely misunderstood and/or misquoted .
Schumpeter reports that,
after crunching the numbers, the model concluded that occupations accounting for 47% of current American jobs (including those in office administration, sales and various service industries) fell into the “high risk” category. But, the paper goes on, this simply means that, compared with other professions, they are the most vulnerable to automation. “We make no attempt to estimate how many jobs will actually be automated,” the authors write.
I came across this piece of research in 2013 in the pages of – wait for it – The Economist. What the Economist reported in 2013 was that Osborne and Frey had shown that 47% of all categories of jobs would disappear over a couple of decades. This clearly allows for the possibility that the actual number of jobs that would cease to exist might be much more or much less than 47% of the total number of jobs currently available.
In the 1960s my maths teacher, Mr Garland, tutored me to successfully fail maths “O” Level three times. Once a week he would test us on 20 basic maths questions then give us the answers and allow us to mark our own tests. On one occasion when my answers did not match his I raised my hand and said
“Sir…I’ve got…..” but he interrupted me saying
“I don’t care what you’ve got Tibbetts it’s wrong”
Little wonder you might say that I gave up trying and settled for failure.
I am not sure that Osborne and Frey know what they are talking about but I am sure that lots of experts commenting on their findings don’t know what they are talking about. And I am sure that the results of machines taking over our work will be a mixed blessing improving many lives and degrading or destroying many others and I am absolutely certain that it doesn’t matter what answer the Economist has got, it’s wrong.