Blogs & Comment

Why robots aren’t ready to manage human workforces (yet)

Robots make excellent task-masters and time-keepers—but that’s not the same as the real job of managing people

Softbank’s “Pepper” Robot

Pepper, a customer service robot, assists customers at Japanese telecom company Softbank’s retail locations. (Tomohiro Ohsumi/Bloomberg/Getty)

Robots are having a moment. The move of driverless cars and drones from sci-fi to reality has unleashed a torrent of bestsellers, blog posts and TED Talks, each pontificating on what it all means. We’re now flooded with prophecies of how machines will take over every facet of human life. Film and television are teeming with critical darlings—Ex Machina, Person of Interest, Humans, even Avengers: Age of Ultron—that ponder the power of artificial intelligence in daily life. BB-8, the spherical (and, it must be said, fictional) hero of Star Wars: The Force Awakens, is currently more famous than most Hollywood starlets.

Given our current fascination with all things algorithmic, it should come as no surprise that robots are apparently about to take over corporate life too. In fact, it’s already happening: Companies such as Hitachi and Amazon have started managing workers’ day-to-day tasks using AI. According to technology research firm Gartner, there will be some three million workers worldwide reporting directly to robots by 2018.

The appeal is obvious—robot bosses can automatically and with a high degree of accuracy accomplish exponentially more than those of us constrained by fallible old meat brains. They can tirelessly monitor every facet of a worker’s productivity and process reams of data to arrive at statistically optimal decisions. They don’t demand six-figure salaries, and they don’t ever take vacation. Moreover, underlings don’t appear to fear them. Take a much-trumpeted 2014 study out of MIT, in which research subjects underwent work trials with both human and robots in charge and found they preferred being managed by the machines.

If you look at the data then, robot bosses add up. This may explain why, in an uncharacteristically breathless piece in The Economist’s “World in 2016” supplement, writer Lucy Kellaway predicted that managerial authority will be replaced by artificial intelligence in the very near future. Her rationale for this prediction: “The manager’s three main functions—checking up on people, chivvying them and judging how they are doing—will either be deemed no longer necessary or will be given to machines to do instead.”

She’s right, in that robots can do all those things. But she’s wrong in another, somewhat more crucial way: Those aren’t the things managers actually do. At least, not the good ones. What we need bosses for in 2016 is not timekeeping or bureaucratic busywork—garden-variety iPhone apps can take care of that. We need bosses who equip businesses with promising talent; who excel at creativity and lateral thinking; and who have the emotional intelligence to herd the complicated urges and behaviours of their employees—most of whom, all hype aside, remain stubbornly human—toward building something great.

Scientists are working to develop AI that recognizes emotions and follows an appropriate decision tree in response, but they have yet to come up with a machine with the nuanced skill to handle Emily’s frustration that she’s twice been passed over for promotion or to channel Amit’s enthusiasm for a promising side project. Empathy and ambition are still human qualities.

If that sounds too soft, consider this: There’s little proof that people are any more productive or hard-working when they’re under the watch of a machine. In fact, there’s evidence to suggest the opposite is true. Not long ago, researchers at the University of Manitoba’s Human-Computer Interaction Lab conducted a study in which participants were asked to do 80 minutes of painfully dull drudge work under the supervision of a human, or that of a computer named Nao. Nao was able to keep 46% of subjects on task for the full duration—not bad, until you consider that the human was able to motivate 86% of subjects to work until the end.

For all the talk of flat hierarchies, the truth is that innovative modern workplaces are more than stimulus-response task factories. They’re run by people who yearn for creative challenges and fulfilling opportunities, and who need the guidance of good human managers—not hard drives—to get them there.