I groaned when I read this from my former employer, PwC
“Our analysis suggests that up to 30% of UK jobs could potentially be at high risk of automation by the early 2030s, lower than the US (38%) or Germany (35%), but higher than Japan (21%).”
They join the many such reports from academics, analysts and authors that filled an entire chapter titled “Sum of all Fears” in my new book, Silicon Collar. I concluded
“The reality, however, is that when Oxford, MIT, McKinsey, and Gartner talk, the person on the street and even business executives typically just read the headlines, and when all of these big brands agree on something, it solidifies readers’ overall impression—in this case, pessimism.”
The PwC analysis suffers from the same “lab rat” mindset of so many such studies. In an annex, PwC explains their approach
In the present study, we first recreated the dataset from Arntz, Gregory and Zieharn (AGZ, 2016). This comprised US data from the Programme for the International Assessment of Adult Competencies (PIAAC) database, merged with automatability data from FO. However, these sources use different occupation classifications: the 702 O*NET occupations from FO were classified using the Standard Occupational Classification (SOC) 2010 codes, whilst the PIAAC database contained occupations classified using the first 2-digits from International Standard Classification of Occupations (ISCO-08) codes.
By invoking work by “AGZ” and “FO” the study may gain credibility in the academic world. Business executives would be much more impressed if they had looked beyond supply side growth in capabilities of AI, robotics and other automation technologies. They should have gone to the demand side and validated the results with actual practitioners who would have told them about maturity and economics of automation and why they plan to continue to hire human workers for a long, long time. They could have surveyed the technology market to see who, if anybody, is developing “frankensoft” machines which have cognitive skills, limb dexterity, visual and speech capabilities and countless other skills humans bring to most jobs. They could have factored societal “circuit breakers” to automation I cataloged in this article, Slow-Motion Automation.
It’s a shame given that as a firm, PwC has access to executives across countries and industries. As a result, you get some head scratching conclusions from the PwC analysis:
- Japan is significantly less susceptible to automation than other developed countries? It is the largest robotics supplier to the world, has vending machines for literally everything, and is leading the world with robotic hotels and restaurants. It also has the most striking labor shortages with an aging population and very low immigration and therefore high incentives for automation.
- Teachers are low-risk for automation? The study explains it away as “Although the considerable growth of e-learning shows that there is scope for automation in education, this may widen access to courses rather than replacing human teachers altogether.” What e-learning promises is more just in time training. That threatens traditional school and university models. It also allows for foundational courses to be delivered as a “shared service” – virtually across schools and universities.
- The more literate a worker the lower the risk of automation? Tell that to IBM Watson, which has shown ability to consume vast amounts of medical journal information that an average doctor or surgeon cannot ever expected to keep up with. IBM is starting to call that “augmented intelligence” to supplement the capabilities of oncologists and other very “literate” occupations.
To PwC’s credit they hedge their pessimism a bit
“However, in practice, not all of these jobs may actually be automated for a variety of economic, legal and regulatory reasons.”
“Furthermore new automation technologies in areas like AI and robotics will both create some totally new jobs in the digital technology area and, through productivity gains, generate additional wealth and spending that will support additional jobs of existing kinds, primarily in services sectors that are less easy to automate.”
Unfortunately, what will stick from the report is the sensational conclusion that over 30% of work populations in the UK, US and Germany could be lost to automation in just over a decade.
The sad part is few of these studies, after scaring the world, come back and acknowledge they were wrong
Gartner, another former employer of mine, has projected: “By 2018, digital business will require 50% fewer business process workers.”and by that time “more than three million workers globally will be supervised by “robobosses””
We are a year away and the predictions are nowhere near likely, but what is the probability Gartner will come back and publicly say mea culpa?
Cross-Posted at Medium