1

Crafting New Classroom Experiences Fast—The AI Way

by Kamyar Goudarzi, Assistant Professor of Management

I knew I wanted my last 408 Business Policy session to be about strategy failures—how they unfold, and more importantly, how to learn from them and particularly, how to avoid them. The interest in talking about failures had sparked in a conversation when Dave Wyman and I were interviewing one of the candidates for a faculty position in the management area. Interestingly, and only a short time before that conversation I had listened to a 4-part series on failures on the Freakonomics Radio podcast. One of the things that really stuck with me from the podcast was the idea of a “pre-mortem” analysis (you can listen to that episode here “How to Succeed at Failing, Part 4: Extreme Resiliency” – https://freakonomics.com/podcast/how-to-succeed-at-failing-part-4-extreme-resiliency/). At the time I thought, that’s it— I need to do something with these ideas on failures since there’s too much focus on explaining successes but not failures.  

 So, while I had put “strategy failures” as a topic in my course calendar at the start of the semester, I hadn’t really given much thought to what exactly I would do for that session. The day before that class, which I admit is pretty late to be planning a class, and wondering how I could turn these ideas into a hands-on experience that applies to a strategy class I turned to ChatGPT and their newly released o3 model: 

 Me: “I want the final session of a one hour capstone strategy class to focus on failures—give me engaging activities with solid learning outcomes such as doing a premortem analysis.” 

 ChatGPT sketched a “Reverse Shark Tank” format: teams draw a famous firm from a hat, plot out a scenario for a plausible five-year collapse, then pitch that downfall in 90 seconds. Peers fire back questions, then vote on whose failure story feels most credible. I loved it! but managing slips of paper and finding a magician’s hat to make it cool seemed cumbersome. So, I decided to make something that can help with the logistics, and I could reuse in the future and share with other colleagues. 

 Me: “Let’s build a single page website that has a button like a hat and when you press it …” 

 In seconds (19 to be exact) I had the first prototype (screenshot below). Thanks to ChatGPT’s live code preview you could see what the app looks like. I then refined the UI: “make the 🎩 bigger,” “create a password field to allow resets only by an instructor with password”, “show real-time vote percentages”—all in the chat window. In a couple of hours, I had an app packaged which I then deployed on Surge (again with the help of ChatGPT) and you can see the result here: https://premortemapp.surge.sh; ChatGPT also gave me a QR code to put on my slides!  

 Well, it’s not perfect yet, but it was ready in just a couple of hours—a task that a few years ago would have taken days which meant I wouldn’t have even attempted it or more likely been impossible given my very amateur programming skills. Those two hours of conversing with ChatGPT and some more tweaking (e.g., making sure each student can only vote once, further UI edits, allowing instructor to change the number of teams, etc.) all came down to this slide. Even the infographic is a ChatGPT creation by the way! 

 What I took away: 

1. AI as an educational partner: ChatGPT isn’t just a coding helper; it’s a co-designer for pedagogy. With a well-crafted prompt, it can brainstorm meaningful activities, prototype interactive tools, and finetune UIs on the fly—making it effortless to experiment with new formats and deliver new classroom experiences.
2.  Speed of prototyping can transform teaching: A fuzzy podcast insight and a brief conversation a few months ago became a live, deployable app and classroom exercise in an afternoon. That agility means last minute inspiration can easily be transformed to real learning. I have started thinking about creating new and more complex exercises.
3. Digital interactivity drives engagement: Moving the random draw and voting to students’ phones and a single interface can shift attention from logistics to idea generation. The thrill of tapping 🎩 and watching results live keeps everyone invested and saves time.

 So, whether you’re designing group projects, simulations, debates, labs, or any hands-on exploration in any field, blending active learning with AI powered prototyping can transform engagement and deepen understanding across disciplines. AI might just become your most adaptable and efficient classroom assistant.