Workers are losing ground against machines

Big Data and cloud computing empower smart machines to do human work, take human jobs

2/4/2013
BY PAUL WISEMAN AND BERNARD CONDON
ASSOCIATED PRESS
PG&E employee Art Liscano, 66, reads a meter at a house in Clovis, Calif. The number of meter readers in the U.S. fell from 48,000 in 2000 to 36,000 in 2010. Every day, PG&E  replaces 1,200 old-fashioned meters with digital versions that can collect information without human help, generate more accurate power bills, and even send an alert if the power goes out.
PG&E employee Art Liscano, 66, reads a meter at a house in Clovis, Calif. The number of meter readers in the U.S. fell from 48,000 in 2000 to 36,000 in 2010. Every day, PG&E replaces 1,200 old-fashioned meters with digital versions that can collect information without human help, generate more accurate power bills, and even send an alert if the power goes out.

Second in a series

WASHINGTON — Art Liscano knows he’s an endangered species in the job market: He’s a meter reader in Fresno, Calif. For 26 years, he’s driven from house to house, checking how much electricity Pacific Gas & Electric customers have used.

But PG&E doesn’t need many people like Mr. Liscano making rounds anymore. Every day, the utility replaces 1,200 old-fashioned meters with digital versions that can collect information without human help, generate more accurate power bills, even send an alert if the power goes out.

“I can see why technology is taking over,” said Mr. Liscano, 66, who earns $67,000 a year. “We can see the writing on the wall.” His department employed 50 full-time meter readers just six years ago. Now it has six.

From giant corporations to university libraries to start-up businesses, employers are using rapidly improving technology to do tasks that humans used to do. That means millions of workers are caught in a competition they can’t win against machines that keep getting more powerful, cheaper, and easier to use.

To better understand the impact of technology on jobs, the Associated Press analyzed employment data from 20 countries and interviewed economists, technology experts, robot manufacturers, software developers, chief executive officers, and workers who are competing with smarter machines.

The AP found that almost all the jobs disappearing are the midskill, midpay jobs — jobs with salaries ranging from $38,000 to $68,000 — that form the backbone of the middle class.

In the United States, half of the 7.5 million jobs lost during the Great Recession paid middle-class wages.

Those jobs are being replaced in many cases by machines and software that can do the same work better and cheaper.

“Everything that humans can do a machine can do,” said Moshe Vardi, a computer scientist at Rice University in Houston. “Things are happening that look like science fiction.”

In the United States, more than 1.1 million secretaries vanished from the job market between 2000 and 2010, their job security shattered by software that lets bosses field calls themselves and arrange their own meetings and trips.

Over the same period, the number of phone operators plunged by 64 percent, word processors and typists by 63 percent, travel agents by 46 percent, and bookkeepers by 26 percent, according to Labor Department statistics.

Does technology create jobs? Of course. But at nowhere near the rate that it’s killing them off — at least for the foreseeable future.

Here’s a look at three technological factors reshaping the economies and job markets in developed countries:

Big data

At the heart of the biggest technological changes today is what computer scientists call “Big Data.” Computers thrive on information, and they’re feasting on an unprecedented amount of it — from the Internet, from Twitter messages and other social media sources, from the barcodes and sensors being slapped on everything from boxes of Huggies diapers to stamping machines in car plants.

According to a Harvard Business Review article by Andrew McAfee and Erik Brynjolfsson of the Massachusetts Institute of Technology, more information crosses the Internet every second than the entire Internet stored 20 years ago. Every hour, they note, Wal-Mart Stores Inc. collects 50 million filing cabinets’ worth of information from its dealings with customers.

No human could make sense of so much data. But computers can. They can sift through mountains of information and deliver valuable insights to decision-makers in businesses and government agencies.

“What’s different to me is the raw amount of data out there because of the Web, because of these devices, because we’re attaching sensors to things,” said Mr. McAfee, principal research scientist at MIT’s Center for Digital Business and the co-author of Race Against the Machine.

Software entrepreneur Martin Ford said, “Organizations are collecting huge amounts of data about their internal operations and about what their employees are doing.” The computers can use that information to “figure out how to do a great many jobs” that humans do now.

The ‘cloud’

In the old days — say, five years ago — businesses that had to track lots of information needed to install servers in their offices and hire technical staff to run them. “Cloud computing” has changed everything.

Now, companies can store information on the Internet — perhaps through Amazon Web Services or Google App Engine — and grab it when they need it. And they don’t need to hire experts to do it.

Cloud computing “is a catch-all term for the ability to rent as much computer power as you need without having to buy it, without having to know a lot about it,” Mr. McAfee said. “It really has opened up very high-powered computing to the masses.”

Hilliard's Beer in Seattle, founded in October, 2011, bought software from the German firm SAP that allows it to use cloud computing to track sales and inventory and to produce the reports that federal regulators require.

“It automates a lot of the stuff that we do,” owner Ryan Hilliard saids. “I know what it takes to run a server. I didn’t want to hire an IT guy.”

Smarter machines

Though many are still working out the kinks, software is making machines and devices smarter every year. They can learn your habits, recognize your voice, do the things that travel agents, secretaries, and interpreters have done.

Microsoft has unveiled a system that can translate what you say into Mandarin and play it back, in your voice. The Google Now personal assistant can tell you if there’s a traffic jam on your regular route home and suggest an alternative. Talk to Apple’s Siri and she can reset an appointment.

Besides becoming more powerful and creative, machines and their software are becoming easier to use. That has made consumers increasingly comfortable relying on them to transact business. Which has eliminated jobs of bank tellers, ticket agents, and checkout cashiers.

People who used to say, “Let me talk to a person. I don’t want to deal with this machine,” are now using check-in kiosks at airports and self-checkout lanes at supermarkets and drugstores, said Jeff Connally, CEO of CMIT Solutions, a technology consultancy.

So machines are getting smarter, and people are more comfortable using them. Those factors, combined with the financial pressures of the Great Recession, have led firms and government agencies to cut jobs the past five years, yet continue to operate just as well.

How is that happening?

● When the Great Recession struck, the Seattle Police Department didn’t have money to replace retiring officers. So it turned to technology — a software system that lets police officers file crime-scene reports from laptops in their patrol cars.

The software was nothing fancy, just a collection of forms and pull-down menus, but the impact was huge. The shift from paper eliminated the need for two dozen transcribers and filing staff at police headquarters and freed desk-bound officers to return to the streets.

● In South Korea, Standard Chartered Bank is expanding “smart-banking” branches that employ a staff of three, compared with an average of about eight in traditional branches. The bank has closed a dozen full-service branches, replacing them with the smart branches, and expects to have 30 more by the end of the year.

Customers do most of their banking on computer screens and can connect with Standard Chartered specialists by video-conference if they need help.

● The British-Australian mining giant Rio Tinto Plc announced plans last year to invest $518 million in the world’s first long-haul, heavy-duty driverless train system at its Pilbara iron ore mines in Western Australia. The trains are part of what Rio Tinto calls its “Mine of the Future” program, which includes 150 driverless trucks and automated drills.

● A driverless car introduced by Google in 2010 and the Pentagon’s drone aircraft are raising the specter of highways and skies filled with cars and planes that can get around by themselves.

“A pilotless airliner is going to come; it's just a question of when,” James Albaugh, chief executive officer of Boeing Commercial Airplanes Group, said in 2011 to IEEE Spectrum magazine.

Unmanned trains have arrived. The United Arab Emirates introduced the world’s longest automated rail system — 32 miles — in Dubai in 2009.

Driverless cars will have a revolutionary impact on traffic one day — and the job market. In the United States alone, 3.1 million people drive trucks for a living, 573,000 drive buses, 342,000 drive taxis, or limousines. All those jobs will be threatened by automated vehicles.

● Phone companies and gas and electric utilities are using technology to reduce their payrolls. Since 2007, for instance, telecommunications giant Verizon Communications Inc. has increased its annual revenue 19 percent — while employing 17 percent fewer workers.

Verizon also makes it easier for customers to deal with problems themselves without calling a repairman. From their homes, consumers can open Verizon’s In-home Agent software on their computers. The system can determine why a cable TV box isn’t working or why the Internet connection broke — and fix the problem in minutes.

The program has been downloaded more than 2 million times, Verizon said.

Next week: Will smart machines create a world without work?