ROBOT BOOM
A PANDEMIC- FUELLED SURGE IN AUTOMATION COULD HASTEN JOB LOSSES AMONG HUMANS,
About half of all the activities people are paid to do in the world’s workforce could ... be automated by adapting currently demonstrated technologies. That amounts to almost US $15 trillion in wages. — Mckinsey consulting
Increasingly computers can ‘see’ too, often better than we can.
Have you begun your new year reappraisal of life? Are you pining for a fresh start? A new career? Or perhaps, after the miserable novelty of 2020, a return to the certainties of old is most appealing?
If it’s the latter, bad luck. In the workplace, from office to farmyard, big changes are coming.
Even before COVID, we heard plenty about the threat of automation. Artificial intelligence and robotics would combine to make many humans redundant, the doomsters predicted. Since the outbreak began, research suggests, it has only “accelerated the pace of technological change.”
Views diverge on how dramatic the impact will be. Pre- pandemic, the OECD suggested that, in the rich world, nine per cent of jobs were at high risk of automation. The consultancy PWC, meanwhile, estimated that just under a third of jobs in the U.K. were at high risk. U. K.- based consultants McKinsey went further: “About half of all the activities people are paid to do in the world’s workforce could potentially be automated by adapting currently demonstrated technologies. That amounts to almost US$15 trillion in wages.”
Even so, it suggested such dramatic automation was likely to see the majority of jobs change. A mere five per cent would completely disappear, in a way that has happened to many supermarket cashiers. Between 2011 and 2017, a quarter of checkout staff lost their jobs, replaced by machines that were at first annoying but soon became second nature.
The pandemic has accelerated and exacerbated the technological cleavages already ripping apart the world of work, creating bigger winners and losers.
Pr e-C OVID, the lowest-paying jobs were most at risk of automation; during it, they have become the most at risk of leave or layoffs. Research by the economist Carl Frey shows that high earners were five times more likely to be able to continue working — it is hard to be a waiter by Zoom. Those who have wanted to keep working in person faced greater risks. In the U.K., the Office for National Statistics has released data showing higher COVID death rates for store clerks, checkout cashiers, security guards and taxi drivers, among others.
Those risks might have been worth taking if they were short- term routes to long- term job security. But a wealth of data point to the opposite.
Humans will not return after the pandemic to jobs that have been automated during it. Hotel check- in clerks, toll collectors on roads — these occupations are not suffering a blip, they are slipping into the history books. For them there will be no V-shaped recovery.
In this way, the COVID recession is not very different to the recessions that have preceded it.
“Since the Eighties, almost all employment losses in routine occupations, which are relatively easier to be automated, occurred during recessions,” says a recent analysis of pandemic effects on employment — titled Forced Automation by COVID-19? — by two U. S economists. These jobs were then “substituted by technology during the recoveries, leading to ‘ jobless recoveries.’”
Any recovery may be welcome now, but a recovery that leaves many lower earners (who have braved greater risks of dying from COVID) without a job, will be a bitter one.
There are some upsides. In the U.S., the pandemic has led to an astonishing burst of entrepreneurialism, with new companies founded at a record rate. That is encouraging because new companies grow fast, and need to hire people. By contrast, Britain’s economy is stuffed with barely productive zombie companies. If COVID’S destruction turned out to be creative — replacing stalled, automatable businesses of old with roaring new ones — we might be able to grow our way of the current economic malaise. There is a wrinkle in this happy scenario, however.
The Resolution Foundation argues that the idea of huge numbers of people acquiring new skills to change professions is mostly fantasy. If we lose our jobs, we don’t upskill, we tend to move down the ladder, to ever more precarious, lower-paid jobs. When Mckinsey looked at people who had lost tech-threatened jobs — such as waiters, PAS, salespeople and accountants — it found most ended up in other occupations that were also at risk.
Such trends are the result of rapid technological advances. In their recent State of AI report, Nathan Benaich and Ian Hogarth note AI’S rapid progress in a host of fields, above all Natural Language Processing (NLP), which allows machines the appearance of “understanding” words, as well as translating or summarizing them.
That makes for very effective virtual assistants, hence the decline of people in similar roles. But even technologists are not immune: NLP allows machines to turn instructions into code. Will we need coders in future?
Increasingly computers can “see” too, often better than we can. And machine learning is tapping into that hitherto uniquely human ability to spot new relationships between cause and effect, enabling better decision making. Computers are becoming rapidly more competitive in tasks once deemed possible only for humans.
How competitive? In a recent fighter- jet dogfight organized by the U. S. Defense Advanced Research Projects Agency ( DARPA), AI beat a true- life Top Gun 5- 0. “The standard things we do as fighter pilots aren’t working,” the pilot reported.
Such exercises show what AI software is capable of when linked to cutting- edge hardware.
Little wonder industrial robots are racing through industry, deploying vision, mobility, gripping and sensing that rival our own, and power, endurance and virus resistance that far exceed us. International Federation of Robotics statistics show that the number of industrial robots worldwide has almost doubled, to 2.7 million, in the last five years. China alone installed 140,500 last year. Britain, fewer than 2,000. Such machines, allied with 3D printers, could accelerate the current trend of reshoring manufacturing and supply chains to developed economies like Britain. But their very purpose means they will create few highvalue jobs rather than many, comparatively expensive, flesh- and- blood roles on assembly lines.
So, which jobs are at risk in Britain today? The reality is grim. Across Europe, some 59 million jobs — a quarter of the total — are estimated to be at risk, says Mckinsey. Almost half were already at risk from automation before the pandemic arrived. Now they are doubly vulnerable. In the last decade telecoms, financial services and education thrived while manufacturing and agriculture declined. Clerical and office support roles, and work on production lines account for 55 million of those 59 million disappearing jobs, and are at greatest risk.
It would be easy to write this off as a transition from manual labour to a “knowledge economy” in which qualifications and technical skills are everything. But that is not true. While such skills will be important, some of the biggest growth areas will demand emotional and social skills that the robots, for now at least, cannot master. Teaching, social services and health care, and the like. The latter alone could add almost five million jobs this decade.
One reason is our aging society. More people will require looking after. And that will have another effect: fewer workers — an estimated 13.5 million fewer in Europe by 2030.
So remarkably, even if COVID leads to a wave of layoffs, some employers may still find it hard to find candidates to fill jobs.