Artificial intelligence in and beyond the classroom
Asper professor incorporates AI in the classroom, studies implications in the hiring process
Asper assistant professor Wenxi Pu explores artificial intelligence (AI) in his teaching and research, curious about the implications of this evolving, sometimes disruptive, technology.
“AI is embedded in so much of our daily life now. It is not just a tool; it is actively shaping and mediating social processes, shaping how we view the world and how we view ourselves,” he says.
In discussing the impact of AI on business, Pu points to a shift from automation to data-driven machine learning. AI was initially introduced in the workplace to perform simple, repetitive and rule-based work, but today AI can complete more complex tasks, recognizing patterns in data and even generating content.
Pu contends that like business, education will continue to be transformed by AI, prompting educators to evolve as well, developing new teaching and testing techniques to keep up. Rather than prohibit students from using AI resources like the now infamous ChatGPT, Pu brings the technology into the classroom, teaching students how to use it instead of instructing them to avoid it.
“We should actively train students to learn how to use AI effectively and responsibly, which is why I incorporate ChatGPT in my BComm and MBA courses,” says Pu. “At the same time, we need to focus on enriching students’ understanding of course materials, on developing their critical thinking skills in class and on cultivating their communication and interpersonal skills.”
In addition to teaching students how ChatGPT works and how to work with it by creating more effective prompts, Pu teaches students how to imagine themselves in relation to AI and to consider its downsides.
“A generative AI like ChatGPT can create new content, but it doesn’t have any sense of what the content means. AI also has the potential to shape our values and create echo chambers of perspectives or polarization,” he says.
For Pu, it is important that students and users of AI remain “in the driver’s seat,” and think of themselves as responsible for the work that the technology generates.
“I also teach them to give the machine time to ‘think’ if it doesn’t produce the right material on the first try. They can ask the machine ‘are you sure about that?’ to generate alternative responses” he says.
His teaching practice encourages students to collaborate with the technology and think critically about the content it produces and, in turn, reminds them to ask the same questions of themselves.
Pu’s investment in teaching students to work with AI demonstrates his acknowledgement that AI technology isn’t going anywhere anytime soon. His research is similarly invested in understanding how organizations use AI and the possible implications of this use.
In a working paper, Pu and colleagues investigate the use of AI and machine learning algorithms in hiring practices, specifically the use of automated video interviewing (AVI) tools. In AVI, jobseekers record video responses to text-based interview questions. Machine learning algorithms then analyze these responses, assessing their answers as well as their tone of voice, verbal styles, facial expressions and more.
The algorithm evaluates qualities that the hiring organization can specify, including the future job performance, personality and organizational fit of interviewees.
Pu and co-authors ask whether individuals who experience mental health challenges face barriers when using this software, and their preliminary findings suggest that individuals with depression, anxiety and ADHD scored lower in the interview rankings.
While the researchers are still preparing their findings for review and publication, their work suggests a need to revisit the invisible stigma of mental health challenges and to further investigate the use of AI in hiring processes.
AI, besides transforming the hiring processes, is becoming part of business more broadly.
Says Pu, “While many companies are still in the adoption stage of AI, advancements in the technology make it easier for even more companies to use and implement it. Effective integration of AI requires companies to rewire their organizational structures, to unify data organization and to create more collaboration between units.”
Pu’s work as an educator and researcher demonstrates that the changing relationship between individuals and AI technology is worthy of careful attention.
This is why he commits to teaching future business leaders more than just how to use AI. He also teaches students how to imagine themselves in relation to AI, how to see its positive and negative potential and how to recognize it as something that can reflect—and shape—how we see the world and ourselves.