A New Way of Theory-Building Using Machine Learning and Artificial Intelligence

Machine learning and artificial intelligence has taken over business practices like never before. From helping to make decisions to designing processes, machine learning and artificial intelligence has transformed the way businesses function. Yet, some academics argue that how academia can use machine learning and artificial intelligence to develop theories and provide to businesses has been largely ignored. Because of this, the gap between fast-paced industry and painstaking academia has only widened. In their research paper “Theories in Flux: Reimagining Theory Building in the Age of Machine Learning”, Professor Monica Tremblay and Professor Rajiv Kohli argue why it is important for this gap to be closed as well as how academia can use machine learning and artificial intelligence to close it.

In addressing the difference between industry and academia, Tremblay and Kohli note that the importance of replicability in theory building for academia does not necessarily apply to industry. When the two went to Silicon Valley to start their research, Tremblay learned that as opposed to academia, “industry does not think that replicability is important because all their decision making is highly contextual and constantly changing. The idea of replicable theories is like having a swiss army knife-- it kind of works to help, but not great. [Industry] would rather have a precision knife to go in and do exactly what they want it to do. Therefore, academics need to make theories that are more flexible and that can explain a highly contextual scenario.”

In turn, a great middle ground between providing theories in a quick manner while also ensuring that they are accurate, Kohli and Tremblay suggest, is the use of machine learning and artificial intelligence by academia. “In the past, we would have to do a survey of 200 people, analyze it, and come up with new knowledge. Now that we have 2 million records that can be analyzed very quickly [with machine learning], we can have more confidence in generating new knowledge without having to do the very rigorous analysis that we are used to doing” Kohli says.

As Kohli and Tremblay have experience in both industry and academia, they know the importance of bringing this way of thinking to the Mason School. “We want to be a contributor to adding new knowledge. To do this, we need to be relevant to industry and engage with it. If we are not relevant to them, businesses don’t see us as useful and we will be doing research in our own labs with no connection to the outside world” Kohli shares. According to Tremblay, by creating theories that are useful to industry, which machine learning and AI help ensure, “academia can uncover some really interesting business phenomenons that can contribute to practice”.

Professor Joseph Wilck, who specializes in operations and information systems management, also sees this type of research as beneficial to the Mason School. “I think what makes the business school a perfect place for Machine Learning and AI research is to ensure that it is applied across a variety of industries, and that students (graduates) are taking the latest knowledge into the workplace to enable further improvements in these different industries. Academia, and in particular business schools, are perfectly positioned and aligned to enable other industries since we are not directly tied to one over the other, and we teach knowledge and skills that are necessary for all industries.”

As machine learning and artificial intelligence become more embedded in industries, academic institutions can be a crucial part in utilizing these technologies to cater theories towards industry in a fast and accurate way. Professor Tremblay and Professor Kohli’s call to action can be the spark to closing the gap between the two stakeholders