Willamette University professor Tim Johnson encourages his undergraduate and MBA students to view the use of AI as a managerial task: to leverage the technology, one must design workflows, oversee resources, organize tools, and communicate those plans via prompts to the model. Using AI thus means managing AI. As the creator of one of Willamette’s first applied Artificial Intelligence courses, Johnson is helping future leaders learn how to manage the machine.
“When you're a manager, you're working with a team in order to execute some goal and achieve it,” Johnson says. “And that's basically analogous to what's going on with artificial intelligence — you're working with a series of models to perform some objective in the workplace you wouldn't just do by yourself.”
Johnson is no stranger to AI. The rise of OpenAI’s ChatGPT was most people’s first real introduction to artificial intelligence — but not for Johnson. Long before large language models made AI technologies mainstream, Johnson began paying attention to AI due to both his interest in the work of Herbert Simon, an early AI pioneer, and his time as predoctoral fellow at the Max Planck Institute’s Center for Adaptive Behavior and Cognition in Berlin, Germany. His interest in AI grew further through the years via his research using computer simulations and big data to model public policy outcomes and human cooperation.
Now he brings that deep experience and unrelenting curiosity to research about big questions like how AI models respond in scenarios that involve altruism and trust. These ideas also shape his classes on AI for undergraduate business and MBA students and exemplify the hands-on, comprehensive curriculum at Willamette’s Atkinson Graduate School of Management.
In Johnson’s class, students don’t just read about AI — they actually test the models on specific management tasks and evaluate how well they perform. Students gain experience testing AI models across all areas of business and organizations. One week, students test AI in business operations, examining how the tools could automate routine tasks like scheduling or process mapping. The next week, students see how AI fares in finance and accounting. Through such classroom activities, Johnson hopes that students not only gain the skill to use AI models, but also the ability to test the performance and assess the impact of those tools.
Understanding AI as skeptics, supporters, or citizens
While he doesn’t expect his students to be computer scientists or coding experts, Johnson believes that managers should understand technical aspects of AI models. Through such knowledge and rigorous testing, students can learn which workplace tasks are appropriate for AI and where AI will fail to meet the needs of an organization.
“My view is that without some degree of technical understanding, it's very difficult to identify use cases in which the tools are effective versus ineffective,” Johnson says. “So the process of exploration needs to be structured around testing.”
Many students come to the class curious about how they can use AI to become more effective professionals, but AI skeptics are also welcome. In fact, Johnson says that even the most reluctant adopters of AI technologies can benefit from understanding how the tools work.
“As with all new technologies, we should have healthy skepticism about AI and we should be judicious in our approach to it,” Johnson says. “By looking at the technology with a critical eye, folks will identify its constructive uses and poor uses more quickly. Accordingly, we need folks who are both critical of AI and willing to study it so that their skepticism can be part of a serious conversation about the technology.”
That is the bigger reason why Johnson is so committed to helping students think critically about the implications of AI: building a better AI future will require informed communities to participate in the decisions we make about this technology — before it’s too late.
“Pretty soon we're going to start grappling with the challenges of a world that changes at a rate that is very difficult for a person to keep up with,” Johnson says. “AI models can be used in the AI development process, which means that as these models get better, they can participate in the improvement of subsequent models. I think this feedback loop sets the stage for the biggest challenge on the horizon for humanity overall: how do we cope with a world in which technological change gets faster and faster?”
After taking Johnson’s course, students are empowered to take on those big questions — one AI prompt at a time.
