How technology will destroy low-wage and middle class jobs the world over
It’s likely that machines will be smarter than us before the end of the century – not just at chess or trivia questions but at just about everything, from mathematics and engineering to science and medicine.” – Gary Marcus, professor at New York University
There is an ongoing crisis of jobs in India and the world that does not provoke as much discussion as it should. The Labour Bureau revealed this year that India added only 135,000 jobs in 2015, compared to 419,000 in 2013 and 900,000 in 2011. That is vastly inadequate considering that 12 million people reportedly enter India’s workforce each year. The news is grim elsewhere too. The jobs situation in the US propelled Donald Trump’s presidential campaign notwithstanding the distaste for his rhetoric and personality. Youth unemployment in Europe is rife. South Korea’s shipbuilding industry eliminated 20,000 jobs this year. Australia has one low-skilled job advertised for every six people looking for work.
The situation is usually attributed to the 2008 financial crisis and the slowdown of China’s demand for raw materials and manufactured goods. But there is also a recognition that advancing technology is displacing jobs, exacerbating the effects of globalisation which has created an unequal world of winners and losers. The World Bank president Jim Yong Kim has pointed out that automation technology threatens 69% of jobs in India and 77% in China.
Keeping in step with changes in technology and their effects on society is not easy given the pace of change. A fascinating, indispensable book to understand where technology is headed, how it will affect the world and what policy choices governments will confront in the future is Martin Ford’s Rise of the Robots: Technology and the Threat of a Jobless Future. Ford’s book, which won the FT and McKinsey business book of the year award in 2015, largely deals with the US but it is relevant for middle classes worldwide which tend to see their own future as interwoven with that of the West and the developed world.
These are some of Ford’s key arguments:
The great displacement
Advances in automation technology will threaten a lot of blue-collar jobs. The “visual perception, spatial computation and dexterity” of robots, for instance, is improving rapidly and their use is in industries increasing. Tesla, the electric car company, for example, uses 160 robots to assemble 400 cars per week. Foxconn, a manufacturer of Apple devices, plans to introduce a million robotics in its factories. Robotics are a focal point of China’s “Made in China 2025” plan; it “has set national goals of producing 100,000 industrial robots a year and having 150 robots in operation for every 10,000 employees by 2020.” There are now machines to make sushi, pick almonds and strawberries in farms and arrange merchandise in warehouses. Momentum Machines has a device that makes burgers from start to finish. The company wants to target restaurants, convenience stores and food trucks and its co-founder has candidly conveyed that his device “isn’t meant to make employees more efficient.” “It’s meant to completely obviate them.”
The advent of artificial intelligence (AI) means that computers are now capable of doing (certain) white collar jobs as well. For instance, software can perform statistical analysis of baseball games, pick out notable events and generate sports stories for readers that can pass off as one written by regular journalists. A powerful AI engine called “Quill” is now used by media outlets, including Forbes, to produce automated articles in fields including, sports, business and politics. One entrepreneur predicts that in 15 years 90% of news articles will be written algorithmically. Quill is designed to be a “general purpose analytical and narrative-writing engine”; it is able to generate business reports drawing on a variety of sources, including databases, “financial and sales reporting systems, websites and even social media.”
The Quill engine is just one of the many new software applications capable of doing the knowledge work that has been the preserve of humans. We are in the age of machine learning, defined as a form of artificial intelligence which gives the computers the ability to learn without being explicitly programmed or, as Ford puts it, “a technique in which a computer churns through data and in effect, writes its own program based on statistical relationships it discovers.” IBM’s Watson technology thus was able to absorb around 200 million pages of information to beat champions of the unpredictable quiz show Jeopardy! in 2011. Watson’s technology is now used in medicine; it processes information from medical textbooks, journals and clinical studies to act as a diagnostic tool and improve treatment plans at major medical centres in the US. IBM has announced that it working on making online shopping sites replicate the personalised service in a retail store. Ford says once these are implemented in the sphere of customer service, “huge numbers of offshore call centre jobs are poised to be vaporised.”
Such technologies are set to be applied in other fields. The IBM announced in 2013 that it was moving the Watson artificial intelligence system to the cloud so that developers can link to the system directly and incorporate the cognitive computing technology into their software applications and apps. Google also offers developers a cloud-based machine learning application and a “large-scale compute engine that lets developers solve huge, computationally intensive problems by running programs on massive supercomputer-like networks of servers.”
Curious, decision-making machines
These trends are pushing computer technology towards awe-inspiring capacities. Machines are now “starting to demonstrate curiosity and creativity.” The Creative Machines Lab at Cornell University set up a double pendulum and developed a software that would independently plot its chaotic movements and come up with fundamental natural laws, which it did – including Newton’s Second Law. The software, called Eureqa, had complete control of the experiment and it was taking decisions on how to release the pendulum in order to discern mathematical possibilities. Eureqa is an example of genetic programming that allows computer algorithms to design themselves; it was made available on the internet in 2009 for others to use and is now producing results in a range of fields, “including a simplified equation describing the biochemistry of bacteria that scientists are still struggling to understand.” Genetic algorithms “have produced designs that are competitive with the work of human engineers and scientists in a variety of fields, including electric circuit design, mechanical systems, optics, software repair and civil engineering. In most of these cases, the algorithms have replicated existing designs, but there are at least two instances where genetic programs have created new, patentable inventions (emphasis added)”. Manuela Veloso of Carnegie Mellon University says that science is now working on making AI systems explain themselves – answering questions such as “why are you saying that?” or why are you recommending this?” – to make it easier for humans and AI machines to coexist.
In the light of these Ford writes that the frontier for IT’s takeover of high-skilled jobs is advancing. This is most evident in the sphere of financial trading where at least half or perhaps as much as 70% of stock market transactions on Wall Street are now done by robotic algorithms through machines communicating over high-speed fibre optic networks. Wall Street banks announced, as a result, massive layoffs even while the stock market was rising in 2012 and 2013. The number of financial workers on Wall St dropped from 150,000 to “barely more than 100,000” between the turn of the century and 2013.
In a fascinating snippet, Ford points to a 2013 paper in the scientific journal Nature about a group of physicists who studied global financial markets and identified “an emerging ecology of competitive machines featuring ‘crowds’ of predatory algorithms,” and suggested “that robotic trading had progressed beyond the control – and even comprehension – of the humans who designed the system.” Assessing the disruptive effect of technologies, researchers at the University of Oxford’s Martin School conducted a study in 2013 of over 700 US job types and concluded that nearly 50% of jobs will ultimately be susceptible to full machine automation. Ford writes that computers don’t “need to replicate the entire spectrum of your intellectual capability in order to displace you from your jobs”, they only need “to do the specific things you are paid to do.”
Technology alone not to blame
So, notwithstanding the promises of politicians the threat of a jobless future and the consequent social crises is very real. One thing needs to be clarified though. Technology is not the only reason for the loss of jobs now; technical advances are happening within an economic context that has been stacked against employees / workers for decades. From the 1950s to the early 70s, wages of US workers rose in tandem with productivity owing to mechanisation. The link between productivity and wages snapped subsequently; in 2013 the average worker earned about 13% less than in 1973 (after adjusting for inflation), even as worker productivity grew by 107 percent and the cost of “big-ticket items” like housing, education, and health care soared. Wages have stagnated or decreased over time, the share of labour in national income has declined as businesses keep more of the profit to themselves – and they either let go of labour or have offshored jobs in order to cut costs.
Advances in automation (and the 2008 recession) exacerbate these trends and we now have a situation where middle class jobs are under threat in a range of fields including customer service, retail, catering, administration and even driving (owing to the advent of self-driving cars and trucks). There is increasing incidence of job market polarisation in the developed world where jobs previously done by the middle class are disappearing and the labour market is divided between low-wage service jobs and highly paid skilled professional jobs. Economists have observed that middle class jobs tend to permanently disappear in recessions and the jobs that are likely to be created during recoveries are in low-wage sectors. The assumption that education and reskilling on the job will lead to new opportunities does not pan out in reality. Many graduates work in low-wage temporary jobs with weak benefits.
The one serious impact of all this is that a majority of population across the world does not have enough money to spend on goods and services. Workers are consumers too, as Ford points out, and their lack of discretionary income reduces the revenue of companies that create products, thereby perpetuating the recession. This then creates a systemic crisis for capitalism. The situation does not seem very dire when one sees buoyant quarterly returns of companies but the fact is that a large chunk of consumer spending is driven by the rich, not by the rest of the population. In 2012, the top five percent of US households were responsible for 38 percent of spending, up from 27 percent in 1992. The bottom 80 percent contributed about 39 percent, down from 47 percent in 1992.
This is not a tenable scenario, with millions of workers either unemployed, underpaid or about to lose jobs. Rise of the Robots underlines that we have to rethink the assumption that education and reskilling on the job will lead to better prospects. Governments do need to grapple with the reality of jobless futures. The rise of Trump and other nationalist demagogues elsewhere, the turn against globalisation and cosmopolitan values are all symptoms of the crisis in employment worldwide. There are policy options that economists recommend but establishments have not been courageous enough to contemplate them or open them up for public discussion.
The views expressed are personal. Twitter: @SushilAaron
Part II of this series examines measures that political leaders can explore in the wake of technological change.