As with most iterations or applications of artificial intelligence, machine learning can either catapult the imagination to utopian climbs or conjure deeply dystopian anxieties. And while neither the stuff of Star Trek nor Terminator will come to pass anytime soon, machine learning will likely be to the post-industrial era what automation was once for the industrial, announcing both promise and perils across nearly every sector of business and society.
Are business schools— or more importantly, their graduates— prepared for the changes ahead?
Many schools are already integrating a deeper technical understanding of machine learning into their MBA curricula. They’re stacking on heavyweight courses in programming, data science, optimization, and neural networks, or cooking up cross-disciplinary programs such as MIT Sloan's Leaders in Global Operations, the Fuqua School of Business’ Masters of Quantitative Management, and Stanford's joint MBA and Computer Science MS.
In a world where computers perform more and more of the cognitive labors once reserved for humans, it makes sense that MBAs would tack closer to the role of engineers. Schools, in turn, openly encourage a pathway between the two.
Maura Herson, director of MIT Sloan's MBA program, happily notes that “MIT Sloan students are able to take classes shoulder to shoulder with students from across campus in the MIT Media Lab, or MIT’s Electrical Engineering and Computer Science program if they want to gain deep technical expertise."
According to John Blevins of UCLA's Anderson School of Management, this cross-pollination goes both directions, creating a richer culture for both.
“Anderson has always worked with our engineering school on campus and often have graduate engineers take courses in our MBA programs,” he says. “Many of our students are engineers themselves (around 20%, I believe) and some have masters and PhD’s in engineering before they come to our program.”
However, as some decision-making is increasingly delegated to algorithms and neural networks, we might ask whether machine learning threatens to upend the very nature of “management” itself. Could machine learning even lead to crisis among MBAs— or at the very least, a radical shift of self-image?
Educators greet such predictions with a healthy dose of skepticism. Thomas Roemer, director of the MIT Leaders for Global Operations acknowledges the advantage of a well-stocked war-chest of technical and analytical skills. Nevertheless, the importance of classic “soft skills” such as leadership and communication have not waned in the slightest.
“Data does not speak for itself, or rather on its own behalf,” he says. “Even the most thorough and conclusive data analysis needs to be communicated, trust in the messenger must be established, and desired change requires leadership. In other words, there is a difference between a front room analysts (analyst, manager and communicator) versus back room analyst (analyst only, even if potentially with more depth).”
Likewise, Jeremy Petranka, assistant dean of Fuqua's MQM, considers machine learning as portending less “transformation” than “acceleration.”
“For instance, customer segmentation has always been a core component to a successful marketing strategy, but with Artificial Intelligence and Machine Learning, and the amount of data that now exists, insights into segmentation that used to take months or years to develop can now occur in an afternoon.
“Through this mode, existing management analysis is faster and better, but not necessarily different. We are seeing business schools begin to incorporate more technical courses in their curriculum, which will help prepare graduates for this accelerated environment.”
So, while the MBA might borrow the tools of the engineer or data-scientist, they in no way assume the mantle or belong at the same desk. The best of the new MBA curricula are wise not to confuse them.
On a broader note, though, in all the rush of this “acceleration,” some business schools might not adequately sensitize their graduates to the ethical business implications of machine learning, despite the very real and enduring fears of the public. Automation led to many painful job losses in the manufacturing sector, and to this day, neither headlines nor heavily-impacted communities are likely to forget it.
When it comes to machine learning, the public foresees two possible outcomes: one utopian, the other dystopian.
In the utopia, machines have taken over all administrative drudgery, and the gains of innovation equitably benefit the whole of society; in the dystopia, machines have rendered even the most educated members of the knowledge sector superfluous and desperate for livelihoods.
What responsibility do business leaders and managers have for lessening the fears and ill-effects of machine learning, not just among the immediate workforce but for the greater society?
As UCLA Anderson's John Blevins puts it, “With data, comes knowledge. With knowledge, comes power. And with power, comes responsibility. We believe our students have a duty to lead their companies ethically as well as effectively.”
Thus, the more machine learning is integrated into their business, the more business leaders must master the so-very human skills of weighing greater efficiency against its social costs, or as Blevins mentions, honoring the “fiduciary responsibility to stakeholders while at the same time balancing the importance of good corporate citizenship.”
Fuqua's Jeremy Petranka cautions that “more than ever, managers will need to manage the fear that change induces while also pushing their companies forward. Business leaders who ignore the organizational and cultural effects that this kind of rapid change can cause will inevitably face a crisis.”
For today's MBA, the steepest challenges with machine learning will ultimately be the least mechanical.