The Human Side of Machine Learning

As enterprises bring machine learning into their organizations, many pundits predict that it will lead to massive layoffs.

Yet in a recent study we developed with the Economist Intelligence Unit, “Making the Most of Machine Learning: 5 Lessons from Fast Learners,” we found evidence that highly skilled human employees will be vital during the transformation and beyond. The “people” part of the business isn’t going anywhere.

However, the integration of machine learning will demand completely new ways to define roles and responsibilities, new skills to either build or co-exist with advanced algorithmic labor and perhaps most importantly, a culture built to continuously evolve and learn along with its artificial intelligence (AI) capabilities.

Indeed, 75 percent of organizations surveyed that are already seeing benefits from machine learning, which we call the “fast learners,” say that they expect to retrain employees as they increase their use of intelligent automation.

The emerging machine learning enterprise will be staffed by people dedicated to, as Erik Brynjolfsson and Andrew McAfee of the MIT Center for Digital Business have said, “racing with machines as opposed to racing against them” from the very top of the organization and throughout it.

Keeping up with the machines is part of a larger challenge: digitally transforming the organization to create fallow ground for new technologies such as machine learning, the Internet of Things (IoT), and Big Data, and analytics to survive and thrive. Avoiding the kind of organizational resistance that has plagued previous major technology shifts, such as the enterprise software wave of the early 2000s, will be critical.

Heading off resistance is the responsibility of company leadership. In that regard, fast learners have a head start. C-level executives at the fast learners are engaged with machine-learning strategy to a higher degree than in other organizations. Fewer fast learners suffer from a lack of strategic clarity about machine learning – as opposed to previous enterprise software efforts, when top executives often checked out.

Having a clear strategy for digital transformation more broadly and machine learning specifically will also be important if organizations are going to attract a new breed of intuitive and inquisitive technologists – who possess not simply programming skills, but deep understanding both of data science and the business – to build their machine learning capabilities. The skills may be hard to find – a lack of available external machine learning expertise was named by fast learners as a top challenge.

What’s more, executives must enlist their skilled non-IT professionals to work together with technologists to develop the organization’s machine learning capabilities over time. Accounting firm and early machine learning adopter EisnerAmper, for example, has been hiring data scientists, business analysts, and application developers into its enterprise technology group over the last eighteen months. The newcomers partner with the firm’s accounting, auditing, and tax leaders to develop new capabilities.

Over time, the machine learning enterprise will begin to function differently, adjusting business processes, staffing models, and learning and development programs to adapt to the speed and scale at which machines can learn, says Stanton Jones, director and principal research analyst with ISG. Organizations are moving from a focus “in which people are driving a process that is supported by technology,” he says, “to one in which technology is driving a process supported by people.”

Machine learning enables companies to exponentially increase the scale of their capabilities without increasing staffing – a.k.a. “non-linear growth.” People will still be involved, but at a higher level, managing, analyzing, or acting upon the machine learning output.

The exponential value starts to accrue when machines augment and complement human skills. “That’s more of a partnership with machine intelligence,” says Cliff Justice, principal in KPMG’s Innovation and Enterprise Solutions team. “You’re going after new ground. You’re innovating faster.”

At EisnerAmper, machine learning is the engine driving the company’s transformation for the digital era, enabling it to move beyond basic auditing and accounting to becoming a strategic business adviser to its clients. The firm has developed smart auditing tools, where the software actually learns how to learn in order to make the process more effective and more efficient, and it plans to launch a full machine-learning driven audit practice within three years.

That will free up EisnerAmper’s practitioners to spend more time providing clients with high-level advisory services and strategic consulting while offering traditional auditing services at a lower price than competitors.

Fast learners in our study are embracing the organizational and cultural shifts required to succeed with machine learning. Indeed, among the study respondents who have just begun dabbling in machine learning but have not yet seen benefits, only 50 percent say they are planning to retrain employees for the machine learning era – 25 percent less than the fast learners.

The fast learners have recognized that the value of machine learning comes with the right combination of human and digital labor. That may explain why fast learners say that organizational resistance is less of a challenge than in other organizations.

As a result, they have a head start not just in developing machine learning capabilities, but in adapting their enterprises for a near future in which the integration of man and machine learning will be a competitive necessity.

Learn more by downloading and reading the full study.

Dan Wellers is the Digital Futures global lead and senior analyst at SAP Insights. Dr. Dirk Jendroska is head of Strategy and Operations Machine Learning at SAP.

This story originally appeared on the Digitalist.