The rise of the robot

aEarlier this month a robot killed a worker at car maker Volkswagen’s (VW) factory in Germany. Although the news caption about the killing fuelled speculation about a robot gone rogue, the explanation given by VW management was more prosaic in that it was an industrial accident as the worker had stepped inside the safety cage housing the robot to prevent worker injury. This news story did, however, bring into sharp relief the increasing importance of robots in the world economy and what the future portends for workers globally.
A concurrence of factors including relentless increases in computer processing capacity, declining data storage costs, lighter fabricating materials, software improvements, and augmentation in robotic sensory and machine vision systems has resulted in quantum enhancements of robot capabilities. A major problem area, which is the subject of debate among social scientists and technologists, is whether the increased reliance on robots – particularly in industry – complements workers and enhances labour productivity or whether increased robot versatility and capabilities will displace workers from many jobs and result in tsunami-like waves of large-scale involuntary unemployment.
The question in the title of the article written by two leading researchers in the field, Erik Brynjolfsson and Andrew McAfee both of the Massachusetts Institute of Technology (MIT), in the July-August 2015 issue of ‘Foreign Affairs’ is whether humans will go the way of horses? Their question references an essay written by the Nobel laureate economist Wassily Leontief who in 1983 cautioned: “The role of humans as the most important factor of production is bound to diminish in the same way that the role of horses in agricultural production was first diminished and then eliminated by the introduction of tractors.” Of course, horses can’t unionise and don’t have political rights so the analogy does not strictly hold but it does raise troubling

questions about the future direction of employment for unskilled workers and even skilled workers in professions as diverse as accounting, finance, law, journalism and medicine.
There are tasks that robots can do better than humans, given the current state of technology. These are jobs which can be categorised as routine such as crunching numbers, data mining of gigantic databases and assembling components in a factory. They are also needed for undertaking dirty and dangerous tasks such as clearing clogged sewerage lines, removing mines in former war zones or cleaning up of nuclear reactor waste after an accident where radiation levels are too high for humans (as in the Fukushima reactors in Japan). What they can’t do currently is what humans are good at: using judgement, thinking in abstract terms, making generalisations and dealing with uncertainty.
Researchers have indicated that a large number of jobs are on the cusp of being automated in the next five to ten years. One example is the truck or taxi driver. Many people have heard of Google and its driverless cars. These self-driving cars have logged over two million miles on the roads and have experienced only 11 minor accidents attributable mainly to human error. Major automobile companies in the US and Germany have also produced experimental versions of driverless cars. Thus the day when self-driving cars and trucks are a common sight on the roads is not far off.
There are many who disagree with the idea that robots will ‘eat’ jobs. The view that automation results in a reduction in the number of jobs in the economy has been derided by economists as the ‘lump of labour’ fallacy. The fallacy lies in assuming that there are only a fixed number of jobs in the economy. In the sceptics’ opinion human wants are insatiable and as goods prices fall over time with rising productivity demand for these goods rises and so does their production and, consequently, the number of people employed in producing goods and services.
Further, the human desire for variety in consumption means that rising real incomes will be spent on new goods and services that result from innovation and entrepreneurial ingenuity in an economy. That is why the employment-to-population ratio has grown over time for the richer countries for much of the last century. Not only have the number and variety of jobs increased, encouraging more women to enter the labour force, but so have productivity and compensation levels.
But as professors Brynjolfsson and McAfee point out in their books and in their interviews and essays, the relationship between employment, productivity and wages has been untethered since 2000. Whereas previously the trend lines for growth in productivity grew in tandem with the trend line for employment and median wages in the US, that has not been the case for the last 15 to 20 years. Labour productivity is still growing but median wages have actually fallen while the total employment number has stagnated in what they call a ‘Great Decoupling’. Some part of this phenomenon is undoubtedly due to globalisation and the off shoring of jobs to countries like China and India but the authors attribute the major portion of the blame to robotics and automation.
But doesn’t adoption of technology depend on its cost compared to other factors of production? It certainly does and here again the cost trend is tilting inexorably in favour of using robots instead of human labour. According to estimates prepared by Goldman Sachs, an investment bank, the payback period (the period of time it takes to recoup in cash the amount of investment in a project) for investment in industrial robots in China has dropped to 1.7 years in 2015 compared to 11.8 years in 2008 assuming a robot life cycle of 10 years. In 2016 the payback period is expected to reduce further to 1.3 years.
China is now the single largest market for industrial robots where numbers installed grew (albeit from a small base), according to the International Federation of Robotics, at an average of 36 percent annually in the period 2008-2013. As robot penetration into small- and medium-sized industry in China now takes off because of their cost relative to skilled labour cost the high robot adoption rate in China is likely to be maintained in the intermediate term.
The implications for wealth and income distribution are dire for almost all economies in the world – especially for those such as Pakistan’s where the majority of the population is illiterate and/or unskilled. Capital owners should win hands down at the expense of labour in the age of robotics which would be in line with economist Thomas Piketty’s thesis of a progressively widening gap between capital and labour. Labour productivity will no doubt increase considerably even as labour’s importance in production decreases over time but the benefits will accrue to those who own the machines and that class of highly skilled worker who can successfully adapt to the demands of the technologies deployed.
But if average household incomes shrink to intolerably low levels, who would purchase the goods produced in the factories and farms operated largely by robots? Would investment in human capital be cut back if unemployment becomes endemic in the world economy? How would saving and investment be affected if economic growth slows in the face of falling demand? These are some open-ended questions which policymakers, academics, and technologists have to keep in mind as we enter an age where as Professor Daniela Rus of MIT puts it: “It has become possible to imagine the leap from the personal computer to the personal robot.”
The writer is group director for business development at the Jang Group.
Email: iqbal.hussain@janggroup. com.pk