Why a Computer may be your Next Manager

As we continue to instrument every facet of modern work, it’s beginning to enable data-driven management. The flip side of this data-driven augmentation, however, is that it may make the act of management the prime target for automation. Could it be that it will be the managers that go first?

Everyone takes turns guessing how long it will be until their job is automated out of existence. After every guess, everyone drinks.

There is a steady drumbeat of news and analysis that predict a certain demise of much of modern work. These predictions have left many rank-and-file corporate workers trying to sort out what’s really happening and what to do next.

But while they sit at the bar drinking and commiserating, their managers are sitting somewhere in the corner feeling slightly sorry for their employee’s predicament and then silently toasting their good fortunes to have worked their way up the corporate ladder to safety.

The safety of corporate managers, however, may not be quite so assured.

In “Are you ready for the data driven management revolution,” I discussed the fact that as we continue to instrument every facet of modern work, it is beginning to enable what I called data-driven management.

The flip side of this data-driven augmentation, however, is that it may make the act of management the prime target for automation.

Could it be possible that it will be the managers that go first?

The chatbot manager

At least one company is already going down this road.

Zerocracy has introduced a chatbot it designed to manage programmers. And it believes that it may represent the future of management.

The company’s CEO, Yegor Bugayenko, explained that many of the problems that he and his co-founders saw in their prior work with development teams had nothing to do with bad code and everything to do with bad management.

“Only about seven to ten percent of failures are caused by technical problems, by technical incompetence,” Bugayenko shared. “The rest of them are caused by management mistakes. People just forget stuff. They miss deadlines. They don’t remember something, they don’t synchronize information.”

Moreover, as they reflected on past successes and failures, they saw that there were two parts to being a good manager: leadership and project management. They further realized that most great leaders struggled with project management and that even the best project managers often let emotion and politics interfere with getting the best work out of people.

They figured they could automate the management part of the equation.

“I’m not talking about things like meetings, conversations, and motivating people — leadership — all that stuff stays,” said Bugayenko. “But the routine parts like planning, risk analysis, risk quantification and qualification, time management, scope control — we can teach a computer to do that.”

And so they did, starting with themselves. The company uses its management chatbot, what it calls Zerocrat, for its internal development.

The transition wasn’t always easy, but to Bugayenko’s thinking, it also made clear to him which workers would thrive in an automated future and which might not.

“When new programmers join us, they expect to be in the office and to be managed by people. They are used to being micromanaged,” Bugayenko explained. “We say, look, you have no managers anymore. You’re completely free. Just talk to the computer, and the computer tells you what to do and evaluates your results. The computer knows whether you’re good or bad, you don’t need to please your manager anymore. Just focus on your results and satisfy the requirements.”

It’s unquestionably a different way of working. The question is, will it work?

A holocratic future?

This vision of a chatbot manager brings to mind another idea that has been bouncing around for some time: the flat, manager-less organization.

Perhaps beginning with Gary Hamel’s provocative 2011 Harvard Business Review article, “First, let’s fire all the managers,” there has been a continuous fascination with the idea of the so-called holocratic enterprise — a self-managing organization.

Immortalized by companies such as Zappos, Gore-Tex, and Morning Star, leadership experts and pundits have celebrated this movement toward a self-managing future. Still, the widespread adoption of holocratic principles has failed to take root in any serious way.

Part of the reason for this limited adoption is that while self-organization sounds good on paper, it is complicated to accomplish and sustain in practice.

In a recent Aeon article, “No boss? No thanks,” professors Nicolai Foss and Peter Klein make a case for why this heralded manager-less approach doesn’t work.

They explain:

“Unfortunately, the bossless-company narrative is dead wrong…Despite big changes in technology and demographics, and increasing globalization, the basic idea of a firm, the nature of ownership and responsibility, and how people coordinate tasks are the same as always…Decisions have to be made about what to produce and how to produce it. Workers need information, tools and equipment, and motivation. And some individuals or groups need to bear the final responsibility and be held accountable for the firm’s actions – the buck has to stop somewhere. All of this is as true today, in our knowledge-based, networked, empowered, startup economy, as it was during the heyday of the large industrial corporation of the 20th century.”

Bugayenko, of course, agrees with the parts about the need for motivation and accountability. I’d guess that he also recognizes that workers need someone to decide what to produce and when, to coordinate tasks, to deliver information, and to allocate resources.

The issue, it would seem, is if it’s better for a human or a computer to serve that role.

Perhaps what has been missing from the conversations around the self-organizing enterprise is the chatbot?

A ready workforce

The general perception (and one that I shamelessly pandered to in my opening) is that people are already afraid of computers taking their jobs and will, therefore, be even more reticent to take orders from a machine.

I admit that when I first heard Bugayenko’s description of the chatbot-managed workplace, I shuddered a bit. Won’t workers recoil at the idea of a machine reducing them to no more than a means to get a job done?

Maybe not.

Oracle and research firm, Future Workplace, conducted a study in the middle of last year in which a stunning 93% of people said that they would trust orders from a robot. The study makes it clear that despite the fear-mongering, most people are embracing a more measured attitude to the encroachment of robots, artificial intelligence (AI), and automation in the workplace.

“As this study shows, people are not afraid of AI taking their jobs and instead want to be able to quickly and easily take advantage of the latest innovations,” said Emily He, SVP, Human Capital Management Cloud Business Group, Oracle. “To help employees embrace AI, organizations should partner with their HR leaders to address the skill gap and focus their IT strategy on embedding simple and powerful AI innovations into existing business processes.”

Bugayenko believes that the reactions from his employees — both good and bad — validate this point. Not everyone likes it, he made clear — particularly those employees who prefer being micromanaged. But he also found that his best employees love it.

“Our programmers really enjoy [being managed by a bot], because everything is predictable,” he explained. “If you fail, you know exactly what’s going to happen…and it’s not because your project manager is in a bad mood. All the bot cares about is the metrics. So you can effectively complete the jobs it gives to you, and be the best programmer you can be.”

Turning work upside down

Having a chatbot as a manager will, unquestionably, introduce a host of new workplace paradigms. Organizations that adopt this approach will replace often-political corporate cultures with algorithms, strict rules, and the elimination of emotion.

In many cases, this by-the-numbers, automated type of management model will make it easy to compare performance and result in a gamified culture that will have its own motivational effect. Whether that will be good or bad remains to be seen.

This approach will not replicate the way a human manager might manage a team — but perhaps that’s a good thing.

A study by MIT’s Computer Science & Artificial Intelligence Lab (CSAIL) found that not only did workers prefer a robot to a human in directing work, but that it resulted in better team performance.

The research team created an experiment in which two humans and one robot had to collaborate to complete a series of tasks. The team with a robot in charge and the humans doing the work was not only more effective, but also one in which “the workers were more likely to say that the robots ‘better understood them’ and ‘improved the efficiency of the team.’”

There are two big messages when it comes to the automation of management.

First, no one is immune from the transformation of work — even, and maybe especially, managers. The only way to ensure continued relevance is to perpetually adapt and evolve your skills. Moreover, you should focus your energy on developing and refining those skills which are the hardest to automate.

Second, as we continue down this road, the rules of work will change forever. Whether we eventually see full-fledged holocratic enterprises develop or if it is whole new automated management models that evolve, it is becoming increasingly unlikely that the traditional methods of organization and management will survive unchanged.

As an enterprise leader, the question for you will be if you have the courage and fortitude to reimagine everything — even if that means automating yourself out of a job.