Sound Off: Artificial Intelligence

Should we consider AI and machine learning as technologies full of promise or peril?

Brian Uzzi, Richard L. Thomas Professor of Leadership and Organizational Change

As AI gets better at human decision-making, it could potentially take jobs away from human beings. It’s not just the worry about losing your job, it’s the worry about losing your status and self-esteem. AI has the potential to make human beings feel like the machine is actually better than we are as an entity, because it does the things that make us human better than we do.

It may be that our response is to begin rejecting technology and embracing religion or to put value on the human spirit. It could also lead to tremendous stratification, where you have a caste system of elite human beings who run the machines, then you have the machines, then you have all the other human beings. Or maybe we redefine what it means to be human. We have a new outlook with the premise that all human beings are created equal. It’s a big planet. Maybe all three happen simultaneously and separately.

Mark Knickrehm ’84, group chief executive for Accenture Strategy

AI will have a more significant impact on how work is done rather than on jobs, with a positive influence on labor productivity that will continue to expand over time. And while some jobs will indeed go away, vast amounts of existing jobs will be augmented by AI, allowing people to do more, faster. But looking further out, many completely new jobs will also be created in the process. The fascinating thing is that AI is happening in a very condensed timeframe, bringing attention to its risk alongside the opportunity.

Accenture Strategy has started to work with Northwestern to enable our people with the right skills needed to manage the things that AI can do for clients. Northwestern, through McCormick School of Engineering dean Julio Ottino and Sally Blount, former dean of the Kellogg School of Management, helped us craft a specialized program for our strategists and data scientists, taught by gifted professors who are bringing great substance and teaching approaches. In the last two years, we’ve trained many of our people on how best to up-skill their data-driven capabilities.

Adam Waytz, associate professor of management and organizations

Humans and machines can partner effectively in several ways. One way is to divide up moral decision-making. Machines are good at being objective decision-makers. It’s optimal for us to rely on machines to make pure utility-maximizing decisions and then for humans to make corrections if they feel any moral rules were violated.

A second framework involves letting machines do the computational work of sifting through gobs and gobs of data, where humans can provide a check on that work and feed their input back to the machine, augmenting the heavy-duty, robotic work with the human expertise.

The third framework is letting robots handle some emotional labor. For example, customer service involves managing others’ emotions. We typically dislike talking to a robot when we need to reach customer service, but some robotic systems can authenticate you based on your voice. If you have the robot manage that experience, then the customer service rep can handle more specific inquiries and not deal with a frustrated consumer.

Tell us what you think of Northwestern Magazine in a short online survey by Aug. 15, and you’ll be entered to win one of eight $25 Amazon gift cards.