In 2011, a book by two MIT professors, Race Against The Machine, ignited a widespread debate about jobs of the future. In short, the study concluded that the sluggish employment growth in the last 15 years was due to increasing automaton in the workplace and that a large percentage of jobs were susceptible to automation. Others studies like the 2016 study by Choi, Manyika and Merimadi found that many categories of jobs had a high probability of being automated in the future. Recently, in 2017, McKinsey Institute came out with their own estimates suggesting the loss of 800 million jobs worldwide due to automation.
Quiet naturally, this has elicited a lot of concern. With the world population set to peak at 9 billion within this century, it is fairly reasonable to ponder the future of jobs in the face of increasing automation. These studies, and more like them, have alarmed naysayers who envision an apocalyptic scenario of social anarchy, brought on by an army of jobless people around the globe. But as we shall see in the following lines, making predictions about the future is a tricky business.
The unease about technological advances and job displacement is not new. Way back in 1589, Queen Elizabeth refused to grant a patent for a new knitting machine since she feared large-scale unemployment of manual knitters. In the early phases of the industrial revolution, a group of manual workers (later dubbed the ‘Luddites’), resorted to smashing machines since they saw it as a threat to their jobs. Karl Marx viewed machines as a capitalist invention for taking care of labour-related problems. In 1927, the then US Secretary of Labour James Davis opined that automation would bring large-scale unemployment. President Kennedy proposed a law in 1963 (later signed) to deal with job losses due to automation. Wasily Leontief, the 1973 Noble prize winner in economics, wrote an article in 1982 expressing fear of high unemployment due to technological advances.
Yet none of these fears materialized and technology has continued its relentless march unhindered without large-scale unemployment. To understand why, we’ll have to understand the nature of automation’s application to our work and economy.
A robot is not what we normally envision, a piece of metal walking around. In fact, it’s just an algorithm, programmed to do repetitive stuff. Washing machines, for example, are programmed to carry out certain functions repetitively. So are computers. These are robots, reflection of Artificial Intelligence (AI) powered by algorithms, and they have been around for some time now. Did all these robots take away our jobs? Let’s look at the evidence.
When Automated Teller Machines (ATMs) came along in 1970s, many foretold the end of bank branches and tellers. In fact, the number of branches increased. In 2010, the number of tellers was higher than it was in 1980, which squarely went against the predictions of doomsayers. A similar example exists in the form of car manufacturing. The assembly and manufacturing lines are now dominated by robotics, yet the persons required for producing a car is higher than ever. In the 1980s, as spreadsheets like excel and other computer software applications were applied in the workplace, there were predictions that accountants will see their demise. But numbers indicate that the growth in accountancy and data-related jobs has far outpaced the demise of traditional bookkeeping.
So what’s going on? Why have machines not displaced human labour and killed off all the jobs? To unravel this puzzle, one needs to understand about how technology and humans interact. Specifically, it is critical to distinguish between technology’s role as a complement or a substitute to human labour. Humans represent physical labour, while technology represents intellectual, intangible capital that is applied to physical labour. When it acts as a substitute, it displaces physical labour. But when it acts as a complement, it creates more opportunities and newer avenues of work and production for physical labour. This has been the historical case and explains why advances in technology and robotics haven’t killed off jobs. Technology has historically acted as a complement rather than a substitute. While a certain profession may see demise, others normally spring up to more than compensate for those losses. A cursory look around us brings forth many examples of such developments. Cars were a rarity in Pakistan at one time, but now roads are choked with them. Did it kill all the jobs associated with horses, carts and buggies? It may have, but it also spawned a whole new industry in terms of car related services. And let’s not forget other offshoots like hoteling industry. As cars became widespread, people increasingly traversed all over Pakistan, giving birth to a new profession in the form of hoteling services, where millions now find their sustenance.
Can we, then, opine that today’s predictions of widespread job losses will also turn out to be wrong, as they have been in the past? It is a complex debate that does not have a straightforward answer. For a long term perspective on this issue, interested readers should read Joel Mokyr’s History of Technological Anxiety and Future of Economic Growth: Is This Time Different? What we can take away from the above discussion are two things: technological advances and automation have not caused levels of unemployment that had been predicted by many. And two, past trends cannot be taken as an assurance that the future would follow the same path. In other words, there is no guarantee that technological advancements of today will not create large scale unemployment in the future. In fact, some respected economists are now openly expressing fears that the rate of technology-induced job destruction is very likely to pick up.
Larry Summers, a towering figure in economics, is an ardent proponent of this theory. Similarly, Nicholas Eberstadt’s book Men Without Work foresees tough times ahead. The main argument here is that future technological developments will not act as a complement (as it previously has), but as a substitute. Moreover, they point to an important aspect of today’s technology: it is based on the type of artificial intelligence that evolves. In other words, it is now about self-correcting machines that think like humans. Only recently, a robot-solved the Rubik’s cube puzzle in record time, something that was thought unimaginable just a decade ago.
Since the future is uncertain and unpredictable, let’s keep our fingers crossed and hope that it never reaches a juncture where humans take recourse to smashing machines.
Quiet naturally, this has elicited a lot of concern. With the world population set to peak at 9 billion within this century, it is fairly reasonable to ponder the future of jobs in the face of increasing automation. These studies, and more like them, have alarmed naysayers who envision an apocalyptic scenario of social anarchy, brought on by an army of jobless people around the globe. But as we shall see in the following lines, making predictions about the future is a tricky business.
The unease about technological advances and job displacement is not new. Way back in 1589, Queen Elizabeth refused to grant a patent for a new knitting machine since she feared large-scale unemployment of manual knitters. In the early phases of the industrial revolution, a group of manual workers (later dubbed the ‘Luddites’), resorted to smashing machines since they saw it as a threat to their jobs. Karl Marx viewed machines as a capitalist invention for taking care of labour-related problems. In 1927, the then US Secretary of Labour James Davis opined that automation would bring large-scale unemployment. President Kennedy proposed a law in 1963 (later signed) to deal with job losses due to automation. Wasily Leontief, the 1973 Noble prize winner in economics, wrote an article in 1982 expressing fear of high unemployment due to technological advances.
Yet none of these fears materialized and technology has continued its relentless march unhindered without large-scale unemployment. To understand why, we’ll have to understand the nature of automation’s application to our work and economy.
A robot is not what we normally envision, a piece of metal walking around. In fact, it’s just an algorithm, programmed to do repetitive stuff. Washing machines, for example, are programmed to carry out certain functions repetitively. So are computers. These are robots, reflection of Artificial Intelligence (AI) powered by algorithms, and they have been around for some time now. Did all these robots take away our jobs? Let’s look at the evidence.
When Automated Teller Machines (ATMs) came along in 1970s, many foretold the end of bank branches and tellers. In fact, the number of branches increased. In 2010, the number of tellers was higher than it was in 1980, which squarely went against the predictions of doomsayers. A similar example exists in the form of car manufacturing. The assembly and manufacturing lines are now dominated by robotics, yet the persons required for producing a car is higher than ever. In the 1980s, as spreadsheets like excel and other computer software applications were applied in the workplace, there were predictions that accountants will see their demise. But numbers indicate that the growth in accountancy and data-related jobs has far outpaced the demise of traditional bookkeeping.
So what’s going on? Why have machines not displaced human labour and killed off all the jobs? To unravel this puzzle, one needs to understand about how technology and humans interact. Specifically, it is critical to distinguish between technology’s role as a complement or a substitute to human labour. Humans represent physical labour, while technology represents intellectual, intangible capital that is applied to physical labour. When it acts as a substitute, it displaces physical labour. But when it acts as a complement, it creates more opportunities and newer avenues of work and production for physical labour. This has been the historical case and explains why advances in technology and robotics haven’t killed off jobs. Technology has historically acted as a complement rather than a substitute. While a certain profession may see demise, others normally spring up to more than compensate for those losses. A cursory look around us brings forth many examples of such developments. Cars were a rarity in Pakistan at one time, but now roads are choked with them. Did it kill all the jobs associated with horses, carts and buggies? It may have, but it also spawned a whole new industry in terms of car related services. And let’s not forget other offshoots like hoteling industry. As cars became widespread, people increasingly traversed all over Pakistan, giving birth to a new profession in the form of hoteling services, where millions now find their sustenance.
Can we, then, opine that today’s predictions of widespread job losses will also turn out to be wrong, as they have been in the past? It is a complex debate that does not have a straightforward answer. For a long term perspective on this issue, interested readers should read Joel Mokyr’s History of Technological Anxiety and Future of Economic Growth: Is This Time Different? What we can take away from the above discussion are two things: technological advances and automation have not caused levels of unemployment that had been predicted by many. And two, past trends cannot be taken as an assurance that the future would follow the same path. In other words, there is no guarantee that technological advancements of today will not create large scale unemployment in the future. In fact, some respected economists are now openly expressing fears that the rate of technology-induced job destruction is very likely to pick up.
Larry Summers, a towering figure in economics, is an ardent proponent of this theory. Similarly, Nicholas Eberstadt’s book Men Without Work foresees tough times ahead. The main argument here is that future technological developments will not act as a complement (as it previously has), but as a substitute. Moreover, they point to an important aspect of today’s technology: it is based on the type of artificial intelligence that evolves. In other words, it is now about self-correcting machines that think like humans. Only recently, a robot-solved the Rubik’s cube puzzle in record time, something that was thought unimaginable just a decade ago.
Since the future is uncertain and unpredictable, let’s keep our fingers crossed and hope that it never reaches a juncture where humans take recourse to smashing machines.