by Edward Cone
News & Record
April 1, 2018
We live in an age of wonders, many of them painstakingly engineered to steal our jobs.
Machines have always replaced human labor — that’s why we build them, and they deliver on their purpose efficiently, remorselessly and, in the long run, to the greater economic good of society. But this is about to happen in some unprecedented ways, and it’s going to leave a mark.
An emerging generation of computers that can learn, see and talk on the phone promises — or threatens — to automate tasks long reserved for people. Analyzing legal documents, knitting apparel, driving a truck: The technologies broadly known as Artificial Intelligence and robotics can do it, or will in the not-too-distant future.
This labor-saving revolution has a lot of smart people worried. The most extreme predictions call for mass unemployment and the end of work as we know it, but that underestimates human capabilities and overestimates computers.
Yet even the rosy scenario is going to disrupt a lot of lives. Jobs are not fungible goods, and there is no guarantee that the people put out of work will be the same ones hired for new positions. We’re talking big numbers here: A recent paper by my colleague James Lambert and his team, “The A.I. Paradox: How Robots Will Make Work More Human,” projects that advanced automation will push 4.3 million U.S. workers out of their jobs by 2027, with a total of 6.5 million people — more than 4 percent of the current labor force — changing positions because of technology in that time.
Landing the new jobs is going to require learning fresh skills. The so-called STEM disciplines (science, technology, engineering and math) will be highly useful, but so will the creativity, empathy and teamwork that continue to set us apart from machines. People who master some or all of these areas will have a shot at lucrative careers. Those who don’t will be consigned to lower-wage service positions. There won’t be a lot left in the middle. (The broader impact of widening income disparities is a topic for another column.)
The path forward is clear: “The success of the U.S. labor market will be defined by how well it can equip the people displaced by technology with the skills that matter most,” Lambert writes. “The challenge for policymakers, business leaders and educators is to deliver a smooth transition to the future labor market.”
The bad news is that we are nowhere near ready to meet that challenge, as a nation or at the local level. Meaningful efforts are underway, but they lack the scale, coordination and urgency to get ahead of the problem. The United States is taking a patchwork approach to a systemic problem.
I asked Erik Brynjolfsson, a professor at the MIT Sloan School of Management and an expert on digital technology, about our preparedness for what’s about to happen. He responded in pungently non-technical terms. “We should be doing a better job of preparing ourselves. Instead, we’re impoverishing ourselves. It’s a shameful public-policy failure.”
From farm to factory to digital labor
Greensboro owes its existence as a modern city to a previous wave of workforce disruption. The small town saw explosive growth around the turn of the last century as people streamed off farms for jobs in mills, factories and offices that thrived on advancing technology. The same pattern played out across the state and the country, with agricultural employment plummeting from about 40 percent in 1900 to just 2 percent today.
More recently, a lot of those industrial-age jobs disappeared. Automation was just one reason for this decline, but many of the remaining or replacement industries are highly automated. That allows businesses to do more with fewer employees, which is good for profits and productivity statistics but less good if you’re looking for work in familiar places. Now, as machines move beyond repetition and brute force to subtler skills, the automation trend is kicking into overdrive. This time, office workers, professionals and service jobs are vulnerable, too.
We like to say that computers are getting smarter as they increasingly mimic human senses, learn from experience and guide machinery through the physical world. For the most part, though, applications of artificial intelligence remain highly specialized; the chatbot that replaces a call-center worker can’t do the job of a network security algorithm, and neither one can climb a cell tower like a repair drone. Machines with anything approaching general intelligence are still a long way off.
But today’s targeted applications are powerful, tireless and improving fast. Megabank J.P Morgan Chase & Co. is running software that can parse loan documents with great accuracy, saving hundreds of thousands of hours in human labor per year. Computers are very good at reading X-rays and increasingly adept at diagnosing illnesses. The hype around self-driving vehicles is Barnumesque, but if I were a young long-haul trucker, I would be exploring my career options.
This technology will have global impact. Garment manufacturing, for example, may return to the U.S. as computer-guided machines take on jobs long monopolized by nimble fingers. Yet the loss of jobs in Bangladesh won’t bring the same number of jobs back here, and employment in automated factories will require something new: the ability to work with intelligent machines. Those skills are in short supply.
The training deficit
Talk about skills training with Randy Parker, president of Guilford Technical Community College, and you’ll come away impressed. The needs of local industries like aviation are being targeted. A $30 million center for advanced manufacturing is about to open. Hot areas like cybersecurity are getting attention. But there is also a sense of running to stay in place.
“We’re not quite there yet,” says Parker, citing persistent skills gaps and the challenge of retraining older workers who didn’t grow up with digital technology.
The AI revolution? It’s just coming into view. GTCC is part of the consortium of local colleges, businesses, nonprofits and the public sector at the downtown Union Square campus, where the next phase of development is under discussion. There will be a focus on skills relevant to the new era, including design thinking, entrepreneurship, digital technology and digital labor, but the curriculum has not been finalized. “We don’t know how it’s going to grow or morph yet,” Parker says.
In Winston-Salem’s Innovation Quarter, Forsyth Technical Community College mixes a healthy dose of so-called soft skills into its training for people already in the workforce and for those seeking jobs. Still, says Jennifer Coulombe, dean of Business & Industry Services, “As a society, we’re not quite ready.” The needs go beyond what even the best community colleges can teach adults. “I want to train my own kids to be adaptable and flexible.”
Brynjolfsson pushes the same formula. “It’s not just training but core education for lifelong skills. The sweet spot is combining liberal arts with science and math.”
How’s that coming along? “Worse and worse,” he says. “We are not doing nearly enough at the national or state level. This has to involve every employer and nonprofit and workers themselves.”
I asked him about a recent report from Sweden, where workers get serious training and welcome emerging technology as a competitive necessity.
“That used to be America,” he said. “We had the best educational system, and we were confident we’d be successful. If you want to make America great again, you have to recapture that.”
Copyright © 2018 - The News & Record