Dr. Keiwan Wind is an Assistant Professor at McMaster University’s DeGroote School of Business. A software engineer by training, he holds an MBA, a master’s in health informatics, and a PhD in Data Science in Public Health, where he investigated the mechanisms shaping perceived health through social network analysis. His interdisciplinary background informs his teaching of Digital Transformation: Thriving in a Changing Landscape (MBA K610).
Teaching digital transformation presents an interesting paradox: the world changes faster than we can document it, let alone teach it. AI, at the heart of this transformation, cannot be ignored, excluded, or resisted. Instead, Dr. Wind sees the classroom as a place to question its nature, capabilities, limitations, and implications. His guiding question is simple: If AI can do what we ask students to do, what exactly are we preparing them for?
I sometimes describe my approach to teaching AI as a Defense Against the Dark Arts class. Not because AI is inherently dark, but because students need to learn how to use increasingly powerful technologies without surrendering their judgment, agency, or, eventually, their livelihoods to them.
My concern is not simply that students might use AI to complete their assignments. It is that if an assignment can be entirely outsourced to AI, we may be teaching a skill that employers can also outsource. Why should students invest their time, effort, and tuition in developing capabilities that are already becoming replaceable?
This question has fundamentally changed how I design my courses.
I approach education as a journey from compound ignorance (not knowing that we do not know) toward simple ignorance (becoming aware of what we do not know). AI makes this journey particularly important. It can generate convincing answers in seconds, sometimes creating an illusion of understanding where little understanding exists. My goal, therefore, is to help students become better at questioning answers rather than merely answering questions.
This is also where Hilary Lawson’s (2005) concept of closure becomes relevant. We do not encounter reality in its entirety. We make sense of it through limited closures, shaped by our experiences, knowledge, assumptions, and positionality. Recognizing these limitations is not a weakness; it is an invitation to learn from one another. I try to bring these different closures into conversation, hoping that, like pieces of a puzzle, they help us develop a richer, though never complete, understanding of reality.
Each in-person class begins with a Tertulia, a roughly 30-minute open conversation exploring the philosophical roots of the technologies we study. We discuss blockchain through anarchism and decentralization, cybersecurity through the paradoxes of privacy and security, and AI through the hard problem of consciousness.
These conversations are deliberately not about finding the correct answer. They are about recognizing assumptions, entertaining competing perspectives, and asking questions we might otherwise overlook.
I see myself as a facilitator rather than an instructor in these conversations. My role is not to tell students what to think, but to help them recognize how their backgrounds and experiences shape their understanding of the world. I encourage them to share their closures without worrying whether their perspectives are sufficiently sophisticated or worthy of discussion. Sometimes the most interesting insights emerge from perspectives that initially seem unrelated. My task is to connect these pieces, challenge their apparent contradictions, and explore what we might understand together that none of us could see alone.
I want students to develop human prompting: the ability to ask meaningful questions of themselves, others, and the world, before worrying about how to prompt a machine.
Another important change is introducing systems thinking and a critical realist perspective into the course. We live in a world of increasingly complex, interconnected problems. Yet we often approach them by isolating variables, identifying correlations, and proposing solutions without sufficiently understanding why the problems exist or persist.
Students work in teams of 4-6 members on a real-world wicked problem, progressing through three phases.
This progression is intentional. AI can help students search for information, organize evidence, summarize interviews, or explore possible relationships. But the responsibility for interpreting evidence, questioning assumptions, explaining causal mechanisms, and defending recommendations remains with students. The objective is not to make AI use impossible, but to make outsourcing intellectual work insufficient.
I encourage students to use AI for repetitive tasks and as a learning companion. It can help them understand unfamiliar concepts, explore alternative explanations, and critique preliminary ideas. However, students must remain responsible for the questions they pursue, the judgments they make, and the conclusions they defend. The wicked problems are selected collectively and their findings and conclusions are evaluated in-person via oral presentations.
After all, the competitive advantage of being human should remain with the human.
The most encouraging feedback has come from students’ engagement with Tertulia conversations and wicked-problem workshops. They particularly appreciate the opportunity to discuss questions that rarely have straightforward answers and to connect technological developments to real organizational and social challenges.
One student once told me, almost apologetically, that after three weeks of class, they were leaving with more questions than answers. My response was simple: “That is the point!”
For me, this captures the most meaningful impact of the redesign. Students are encouraged to move beyond identifying problems and prescribing familiar solutions toward recognizing complexity, questioning their assumptions, and understanding the mechanisms that produce observable outcomes.
These capabilities are not entirely AI-proof, nor should we pretend they are. But they are harder to outsource because they involve contextual judgment, engagement with people, intellectual responsibility, and the ability to decide what matters.
I do not yet have systematic evidence demonstrating better long-term learning outcomes. What I observe is greater willingness to question seemingly obvious explanations and engage with the uncertainty inherent in complex problems.
My next step is to move from systems thinking toward system dynamics, helping students build simulation models and digital twins of complex problems. They can experiment with interventions, explore unintended consequences, and understand how reasonable decisions sometimes produce undesirable outcomes.
AI will be part of this process, not its replacement.
One lesson has become increasingly clear: We in academia cannot afford to be merely reactive to technological change and let corporations redesign our programs to supply workforce for their version of future. Universities are uniquely positioned to imagine, question, and shape alternative futures.
Ultimately, the future of education does not lie in competing with AI at producing answers, but in cultivating the curiosity, judgment, imagination, and intellectual humility to recognize which questions are worth asking.
And perhaps that is where education should have been all along.
—
Wind, K. (2026), From Answering Questions to Questioning Answers, Teaching in the Age of AI: Examples, From Answering Questions to Questioning Answers, Licensed under Creative Commons BY-NC-SA 4.0.
Lawson, H. (2005). Closure. Routledge. https://doi.org/10.4324/9780203995792
Teaching in the Age of AI