Given how artificial intelligence-led robots and machines are doing even non-repetitive jobs that require cognitive skills, most analysis on them paint a never-before kind of across-the-board job destruction with, unlike in the past, no new job-creating areas coming up—the 2016 Oxford Martin School-Citi report said automation in developing countries puts 85% the jobs at risk. New research by McKinsey Global Institute (MGI) paints a less grim picture—while around 5% of jobs that exist today are completely automated immediately, around 30% of the work done in around 60% of jobs can also be automated by 2055. As a result, while talking of very large productivity gain of 0.8-1.4% per annum, MGI point to the need for man and machine to work together to derive this benefit, and it is not just at low-skill jobs that the challenge will arise. A fourth of even a CEO’s job, MGI says, can be offloaded to an algorithm/AI entity. For the rest, it depends. By way of an example, a filing clerk and a strategist in a company both work with data; however, automation of data gathering and processing will summarily replace the filing clerk while it will complement the work of the strategist.
You May Also Like To Watch This:
In a country like India, MGI estimates 235 million jobs are at some kind of risk—in terms of wages, that’s a whopping $0.4 trillion that’s being talked about. How far robots will take over jobs at restaurants or at hospitals—the impact on manufacturing is expected to, in any case, be large—will depend on their costs. Given the speed at which costs of automation are falling, it is impossible to make a prediction, but clearly jobs that have low wages will get automated later. The policy lesson here is that governments who artificially hike wages through legislation will invite robotisation faster. The speed of automation is not in government hands either—if China lowers costs dramatically by automating its textile industry, India has no option but to follow suit, else Chinese imports will flood the markets. With the fate of so many going to be determined by their ability to work with robots and make the most of them, disparities in salaries simply have to increase dramatically—the productivity-driven gains MGI and others talk of will not accrue to the bottom of the pyramid. That means education policy has to be an integral part of jobs policy—people can no longer afford to be ‘trained’, they need to be ‘educated’ to not just cope with changing technology but to be able to deliver a higher order of skills since the routine/repetitive task will be done by machines. Few in government, sadly, seem aware of the ramifications, or the speed, of the changes in the offing.
