Summary: When used with purpose, custom GPTs don't just help teach—they simulate real-world professional challenges, allowing students to experience consequences, test reasoning, and refine thinking—all in a safe environment. Professor Michael Harden’s business students aren't just learning theories anymore; they’re interacting with AI-powered decision makers that challenge them, question them, and sometimes flat-out reject their proposals. And that’s exactly what makes the learning stick.
Why Students Zone Out—and What Actually Pulls Them Back In
Lectures are necessary. Readings have their place. But neither replicate the messiness of real decisions, where time is short, stakes are real, and human behavior is unpredictable. This is why engagement in business classrooms is tough to manufacture with static materials. Students often know there's a right answer somewhere in the back of the textbook, so they're not truly risking anything when they pick a side in a case discussion or answer a quiz.
What Professor Harden saw—and what most educators already feel—is that students want relevance. They don’t want to just “learn” marketing theory. They want to pitch. They want to try something bold—and then adjust when it doesn’t land. That’s hard to design into weekly curriculum. But it’s exactly what custom GPTs enable, if you use them wisely.
Custom GPTs Give Theory a Voice
The value of a custom GPT lies in its structure. It doesn’t just blurt out search results. It operates within a persona—an angry client, a skeptical stakeholder, a panicked co-worker. And because it can be trained to respond consistently within that context, it creates the illusion of a real-world counterpart. This is where students perk up. Why?
Because GPTs push back.
When a student pitches a half-baked ad campaign, the AI playing the role of a cautious executive challenges the ROI. When a student tries to resolve a personnel issue by "laying down the law," the AI mimicking a defensive employee doesn’t just accept that—it escalates. These interactions force students to think, adapt, and iterate in real time.
These aren’t throwaway simulations. Students remember them. Some even get rattled. And that’s the whole point—because when it matters, they’re engaged. And when there’s friction, learning happens.
Building Engagement with Structure, Not Surprise
Let’s be honest: AI in education has a bit of a Wild West problem. Tools are used recklessly or replace actual teaching, which defeats the purpose. Professor Harden avoided that trap by anchoring the role of the AI firmly within the learning objective.
Each custom GPT was tailored for a job. One might act as a hiring manager during a mock interview. Another plays a passive client who won't respond unless provoked with the right questions. It’s not just about having the AI there—it’s about scripting its boundaries, expectations, and reactions intelligently enough that it behaves like a real person in a recognizable context.
This structure makes space for both action and reflection. It's not just, “Did you get the right answer?” It's, “How did you approach the conflict? What assumptions tripped you up? What would you do differently next time?”
Students Feel the Stakes—Without Real-World Risk
One of the hidden benefits of AI-driven role-play is it creates psychological safety. Ironically, many students who are silent during class discussion are the most vocal with the GPT. Why? There’s no fear of being judged by peers. Yet the role-play still feels consequential.
Students get to test drive professional behavior in a safe but high-fidelity way. That gives hesitant students courage to try new approaches. It levels the playing field between confident talkers and quiet thinkers.
And here’s an insight a lot of professors miss: riskless participation is better than no participation. The traditional classroom can reward quick talkers, not deeper thinkers. GPTs rebalance that.
The Instructor’s Role Changes—But Doesn’t Shrink
Professor Harden didn’t use AI to teach for him. He used it to collect material for deeper teaching. Instead of asking students to summarize a reading, students are asked to reflect on what the GPT challenged them on. Why did their approach backfire? What part of the framework or model did they forget to apply?
This reframes the professor as an observer and guide. They don’t sit back—they watch how students interact with AI, then debrief with depth. They can spot flawed reasoning more easily and help students unlearn lazy thinking habits. GPT becomes a lens, not a substitute.
Seen this way, the voice of the professor doesn’t get quieter. It shifts frequency—away from dictating and toward discussing. That’s where real education happens.
Practical Ideas for Implementation
For professors thinking about integrating GPTs into their classrooms, start simple. Don’t overdesign. Don’t pretend the GPT is perfect.
- Design 1-2 interaction points across a semester that simulate key professional roles your students will encounter.
- Brief your students clearly. Tell them what the GPT will be doing and what kind of feedback they're expected to capture during the exchange.
- Follow up the simulation with written or voice-recorded reflections where students break down the problem they faced and how they handled it.
- Use those reflections to launch class discussion—not starting from theory but from real attempts and stumbles.
This doesn’t need to replace what’s already working. It's not a revolution—it’s an amplifier. It brings alive the business challenges that students too often only see in hypotheticals.
The Bottom Line
AI doesn’t fix disengaged students. But it creates new surfaces to grip onto, new corners of curiosity, and new chances to fail (and recover) in a way that makes learning personal.
What professor wouldn't want that?
Done right, custom GPTs don't just make classes more high-tech. They make them more human. They mirror what students will soon face in boardrooms, cubicles, client meetings, and crisis calls—only now, they get to practice with a safety net.
No more passive participation. No more fake decisions. Just real engagement—and real growth.
#ClassroomAI #ActiveLearning #AIinEducation #CustomGPT #BusinessEducation #StudentEngagement #HigherEdInnovation
Featured Image courtesy of Unsplash and Maccy (yioua5jPPjc)