Hollywood might have you believe that the real threat of smart technology lies in the possibility of robots taking over the world.
Gene Roche, executive professor of higher education at William & Mary, is less sold on the idea of total robot world domination. He is, however, focused on perhaps a more startling and certainly more imminent reality: machines dominating the workforce.
“Many jobs that we thought only human beings can do are going to be done much more effectively by computers,” said Roche. “Take, for example, stock traders and bond traders — those are very well compensated people — but the evidence is growing that artificial intelligence can be much more effective in trading stocks and bonds than people can.”
As the newest appointed Center for Innovation in Learning Design faculty fellow, Roche is leading the charge at W&M to figure out how people will cope in a world of smarter machines and how the education system can prepare students for a drastically different job market. Roche’s interest in the topic stems as far back as the earliest days of the Internet, when he was working as director of career services at Hamilton College in New York.
“One challenge for the School of Education is to think about the possibilities,” said Roche. “What would society be like if 30–40 percent of the jobs that are currently held by college graduates could be done better and more cheaply by machines? What we know for sure is that it requires an education system that’s much different than we have now.”
Robots at work
The concept of robots taking over human jobs isn’t a new one, particularly in the manufacturing industry. Companies from automakers to Foxconn, the manufacturer of the iPhone, have replaced hundreds of thousands of manufacturing jobs with robots in the past decade. Other industries are under imminent threat, too. The rise of technologies such as self-driving cars could offer a wealth of benefits in terms of reducing the number of automobile fatalities but, Roche said, it will also deliver a major blow to the workforce.
“If we could eliminate 30,000 [automobile-related] deaths by having cars do most of the driving, how could we morally not make that decision?” said Roche. “But if we do, the economic ramifications are huge. You’ve got millions of people who make their livings driving trucks and taxis, repairing cars and writing liability insurance policies. All of that would change dramatically with large-scale adoption of autonomous cars.”
What’s startling for many is the breadth of capabilities in artificial intelligence: not only are machines able to learn, they’re also able to teach themselves (a concept called deep learning). Google recently built a program that taught itself to play a Chinese strategy game called Go and then beat the world champion — three times in a row.
Couple that with advances in robotics and the Internet of Things (or the capability of computers to collect information from physical, Internet-enabled objects like your phone, camera or thermostat) and the possibilities of smart machines are virtually endless. If robots can autonomously learn, strategize and be creative, even the most seemingly human jobs could be under threat, said Roche. Within the next century, machines could be selling cars, developing ad campaigns, performing surgery and penning bestselling novels.
“I worry that we’re going to lose jobs that traditionally have provided people with fulfillment and a sense of purpose,” said Roche. “What would our society look like if people didn’t have paid work to help provide meaning and focus to their lives?”
Preparing for the future
The forecast isn’t all grim, Roche points out, if society can learn to adapt. It’s no small feat, considering the fact that no one knows exactly what the world of smarter machines will look like in 100, 50 or even 10 years.
“Artificial intelligence will breed many new jobs, but at this point we can’t even conceive of what those are and what kind of competencies those are going to require,” said Roche.
Instead of focusing on specific skills, Roche is looking at the bigger picture — how can we design learning experiences that prepare future generations to work together and become problem solvers? This question is among those Roche will explore in his seminar course this fall on Educational Technology Planning.
“If we know the broad parameters of what deeper learning is and what it can potentially do,” said Roche, “then we can begin to ask ourselves what that means for each of us … Since we can’t predict the future, we need to make sure the education system is responsive enough and flexible enough to help students prepare for whatever changes they face.”