OECD: Job Loss from Automation May Be Less Severe than Feared, but Still Painful

OECD: Job Loss from Automation May Be Less Severe than Feared, but Still Painful

A recent report from the Organization for Economic Cooperation and Development finds that the number of jobs at risk of displacement due to automation in the coming years is probably smaller than previous forecasts have estimated. Nonetheless, the tens of millions of workers in developed countries are still at risk of having their jobs replaced or radically altered by AI and robotics. The Verge’s James Vincent summarizes the report’s findings:

The researchers found that only 14 percent of jobs in OECD countries … are “highly automatable,” meaning their probability of automation is 70 percent or higher. This forecast … is still significant, equating to around 66 million job losses.

In America alone, for example, the report suggests that 13 million jobs will be destroyed because of automation. “As job losses are unlikely to be distributed equally across the country, this would amount to several times the disruption in local economies caused by the 1950s decline of the car industry in Detroit where changes in technology and increased automation, among other factors, caused massive job losses,” the researchers write.

The analysis from the OECD, an inter-governmental organization representing the world’s 35 richest countries, is considerably less disconcerting than previous studies that have calculated the risk of automation at anywhere from 30 percent to fully half of all the work currently being performed globally. One difference between this study and previous ones, Vincent explains, is that it pays greater attention to details like whether a job can be fully or only partly automated and the variations among jobs that may have the same title but whose work differs substantially:

Read more

Walmart Workers Actually Like Their New Robot Coworkers

Walmart Workers Actually Like Their New Robot Coworkers

Last October, Walmart announced that it was rolling out shelf-scanning robots at 50 stores throughout the US after piloting them at a smaller number of locations in Arkansas, Pennsylvania, and California. The robots are taking over some of the menial busywork that used to occupy employees on the store floor: checking shelves for out-of-stock items, incorrect prices, and wrong or missing labels.

At the MIT Technology Review, Erin Winick recently talked to Martin Hitch, chief business officer at Bossa Nova, the San Francisco-based robotics firm that created the machines, about how employees and customers were reacting to them. While you might expect employees to resent having their work automated or fear that the robots would put them out of a job, Hitch said employees “instantly become the advocates for the robot”:

One way they do that is by giving it a name—the robots all have Walmart name badges on. The employees have competitions to see what the right name is for each robot. They also advocate for the robot to the general public. It’s the store staff saying, “It’s helping me.” We see them now defending the robot.

Read more

Walmart Pilots Robot Floor Cleaners, Raising Automation Jitters

Walmart Pilots Robot Floor Cleaners, Raising Automation Jitters

Walmart, the largest private employer in the US, is testing a self-driving mechanical floor scrubber in five of its stores near its headquarters in Bentonville, Arkansas, LinkedIn managing editor Chip Cutter noted in a recent blog post—and it’s not the only automated technology the big box giant is looking into:

The machine resembles a traditional scrubber but comes equipped with similar technology used in self-driving cars: extensive cameras, sensors, algorithms and Lidar for navigational mapping. Think of it as a Roomba crossed with a Tesla. A human must first drive the device to train it on a path; it can then operate largely independently, including when a store is open to customers. If a person or object gets in its way, it momentarily pauses and adjusts course. …

Walmart has said it wants to automate tasks that are “repeatable, predictable and manual,” giving its people more time to focus on higher-value work like customer service and selling.

Many of the menial tasks involved in retail work are ripe for automation, and Walmart is by no means the only major retailer experimenting with new technologies. Lowe’s rolled out an autonomous retail service robot called “LoweBot” in the San Francisco Bay Area last year, while Amazon continues to invest heavily in its robotic workforce. Being such a massive employer, however, Walmart can affect the entire US economy with its labor market decisions, so any changes it makes are bound to attract attention. In this case, Walmart’s move highlights concerns about the impact of automation on the workforce: What happens when the country’s largest employer no longer needs so many employees?

Yet Walmart is sensitive to this concern, Adam Pasick and Karen Hao write at Quartz, and has stressed that it is not using machines to replace employees:

Read more

McKinsey: Up to 30 Percent of World’s Work Could Be Automated by 2030

McKinsey: Up to 30 Percent of World’s Work Could Be Automated by 2030

When it comes to the threat or promise of automation, experts are divided as to whether AI and robotics will eliminate jobs en masse or merely automate rote tasks and free up more of workers’ time for innovation and creativity. McKinsey has put out some interesting research throughout the year in which they attempt to forecast the impact of these new technologies on the workforce. In January, they released the attention-grabbing headline finding that half of the work currently performed by humans could be automated with already-existing technology. Though fewer than 5 percent of jobs can be automated entirely, their research found, most jobs could have at least one third of their component tasks automated today.

In an update to that work published this week, McKinsey takes a closer look at the various factors that will drive automation in the coming decades—such as technical feasibility, cost of deployment, and labor market considerations—and concludes that “between almost zero and 30 percent of the hours worked globally could be automated by 2030, depending on the speed of adoption.” The effects will not, however, be evenly distributed among occupations:

Activities most susceptible to automation include physical ones in predictable environments, such as operating machinery and preparing fast food. Collecting and processing data are two other categories of activities that increasingly can be done better and faster with machines. This could displace large amounts of labor—for instance, in mortgage origination, paralegal work, accounting, and back-office transaction processing. … Automation will have a lesser effect on jobs that involve managing people, applying expertise, and social interactions, where machines are unable to match human performance for now.

Read more

Who Will Remember How to Do the Robot’s Job?

Who Will Remember How to Do the Robot’s Job?

In the coming decades, automation is expected to displace vast numbers of employees performing routine tasks in a variety of industries, including Asian factory workers and low-wage jobs in the US manufacturing, logistics, and service sectors. One study posits that half of the world’s work can already be automated using current technologies, though in many cases it is not yet economically efficient to do so. The jobs most ripe for automation are those involving routine, repetitive, manual tasks that robots can perform more rapidly and precisely than humans without experiencing boredom, injury, or fatigue.

Nonetheless, many of the people who have held these jobs up until now possess knowledge that may still be valuable to their employers even after they are replaced by machines. In that light, Quartz’s Sarah Kessler considers what automation will mean for training and knowledge retention as fewer and fewer employees remain who remember how to do the robots’ jobs:

As GE has automated its factories, its executives say they still value experience like [assembler Bill] Knight’s, if only because it helps them catch machine errors and, in some cases, better understand how to program machines. “If there’s something wrong in the calibration of this equipment, I need to know that,” says Denice Biocca, GE’s head of HR for supply chain and services. “I need to have some sort of understanding of what [the process] should be in order to know, ‘the machine says it’s tight, but I know it’s not.”

Read more

Is the Post-Work Future a Myth? Not for Everyone

Is the Post-Work Future a Myth? Not for Everyone

Central to the debate over automation and the future of work is the fear that robots and artificial intelligence will take over so much of the work people currently do that there will be no jobs left for human beings. Some believe we may be looking at a not-too-distant future of mass structural unemployment due to automation, while the more optimistic crowd contends that the current wave of automation will ultimately create more jobs than it destroys, just as the steam engine, the automobile, and the computer did in past periods of technological change. Political scientist Ruy Teixeira is in the second camp, arguing at Vox that the fears of a post-human workforce don’t square with the economic data:

If [futurist Martin] Ford’s “rise of the robots” were taking place, we would be seeing very rapid productivity increases today (fewer workers, larger output). We’re not. Instead, productivity increases have been abominably slow in recent years — a mere 1.3 percent per year, just over a third of the rate at the end of the last century.

Another indicator that the robots are gaining on us would be an exceptionally high rate of “occupational churn,” the rate at which the job structure is changing as some occupations decline and others grow. In a study of Census data going back to 1850, economists Robert Atkinson and John Wu found instead that the rate of churn in recent decades has been exceptionally slow — slower, in fact, than at any other period in their study.

Greg Ip made a similar point in the Wall Street Journal last month, also pointing to the slowdown in productivity and arguing that, if anything, robots aren’t taking over enough of our jobs. “Instead of worrying about robots destroying jobs,” Ip wrote, “business leaders need to figure out how to use them more, especially in low-productivity sectors.”

In these pieces, Teixeira and Ip write to dispel the “jobless future” hypothesis and do so quite effectively. Yes, this same debate occurred in the 1950s and yet we’re still here. But what makes this time different—and indeed it is different—is that technological change tends to proceed at an exponential rate, and we are reaching a point in which new technologies are rapidly making entire industries redundant.

This quality of rapid change cannot be said about people, and this includes our skills, our ability to adapt, and our willingness to change social and political norms. The difference in the speed of change is what makes this threat of a jobless future so real, even if productivity growth remains incremental.

Read more

Microsoft’s Build Conference Hints at the Workplace of the Future

Microsoft’s Build Conference Hints at the Workplace of the Future

Build, Microsoft’s annual developer conference taking place in Seattle this week, is focusing heavily this year on AI and machine learning, and how the company plans to embed these technologies in the workplace of the near future. Wednesday’s keynote demonstrations showcased what Mark Sullivan at Fast Company calls “a vision of the workplace of the future where workers are surrounded by all manner of cameras, sensors, and other recording devices connected to internet-based AI services”:

Microsoft showed demos and videos of the “intelligent edge” in a variety of forms, in a variety of use cases, and in a variety of industries:

  • A heart patient was walking around wearing a sensor. He began to get tired, so the sensor sent that data up to the cloud for processing, and a nurse was notified to bring him a wheelchair.
  • A camera detected an employee accidentally tipping over a barrel containing a dangerous chemical, information it sent up to image-recognition software in the cloud. Some other database likely helped determine that the liquid in the barrel was hazardous. Presumably an alarm was sent to a cleanup team.
  • An employee in a shop was spotted taking a selfie while brandishing a jackhammer. The brain in the cloud recognized the employee, the activity, and the setting and concluded he was behaving recklessly, then contacted a supervisor.
  • Someone else in a shop was seen not wearing safety goggles. Alarm. Supervisor notified.

All this involves some sophisticated, on-the-fly AI. In the words of the presenter demoing the intelligent edge developer tools at Build: “The solution is running more than 27 million recognitions per second across people, objects, and activities.” But the use cases Microsoft showed onstage sound equal parts helpful and intrusive. Sure, getting a heart patient back to bed or detecting a dangerous chemical spill are health-promoting. But the notifications to the supervisor suggest a completely different, and possibly unintended, consequence of the technology.

“There is benevolent surveillance and then there is just surveillance,” Sullivan worries, “and the Microsoft technology could work in both scenarios.” But CEO Satya Nadella spent part of his time on the Build stage Wednesday dispelling the notion that Microsoft is out to turn the workplace into a surveillance state:

Read more