IU Today is checking in from time to time with faculty and staff to learn how they have successfully used artificial intelligence tools in their work, research and teaching. We recently caught up with Amy Kinser, a professor in the Kelley School of Business at IU Bloomington.

Question: How have you used AI to accomplish tasks that aid your work/research, and what benefits have you seen?

Answer: My efforts are a part of the broader strategic initiatives at the Kelley School to create AI-resilient classroom experiences: learning designs that encourage students to use AI critically and responsibly rather than avoid it or rely on it blindly. My efforts have been guided and inspired by the Kelley AI Playbook.

Through the Kelley AI Playbook’s encouragement, I use AI daily in all dimensions of my work from teaching and course design to administrative processes. For example, administratively, I have used AI-driven workflows in Power Automate and N8N to extract international activity data from hundreds of faculty CVs for Association to Advance Collegiate Schools of Business reporting, saving many hours of manual review.

In one classroom example, I have been experimenting this semester with integrating IU-licensed NotebookLM directly into two very different courses:

• K201 (Foundations of Business Information Systems and Decision Making): A large, highly coordinated freshman course where I cannot alter the curriculum. Here, I use AI-generated after-class review videos branded as K201 Quick Hits, such as “Guardians of the Data.” These short videos summarize key concepts and are worth no points, yet roughly 70% of students watched them early in the semester (now about 30%), and many have commented that the accompanying Notebook helps them quiz themselves before exams.
• K360 (VBA and Application Integration): A senior-level course that evolves quickly with technology. Here, AI assists in producing before-class preparation videos branded as K360 Video Microcasts, such as “Loop There It Is: Mastering Repetition.” Each video is paired with a single PlayPosit question worth two points. Students in this course have shown stronger assignment performance and come to class better prepared compared to prior semesters.

In both cases, I use AI tools (primarily IU-licensed NotebookLM and IU-licensed ChatGPT Edu) to draft prompts and even generate branded/catchy video titles. AI has been tremendous for supporting the iterative design of instructional materials.

To better understand how students experienced AI integration in these courses, I added two custom questions to Indiana University’s Online Course Questionnaire evaluations. The first asked students to rate the statement, “The instructor helped students understand when AI use is or is not appropriate in the subject matter,” on a five-point scale. The second asked an open-ended question: “In what ways did the approach to AI in this course shape the quality of your learning experience?”

Student responses were strongly positive in both courses: 4.48 to 4.75 out of 5. Students frequently described AI as most helpful for explanation, debugging, and studying when paired with clear expectations about independent thinking. As one K201 student wrote, “When I didn’t have Amy to ask questions, NotebookLM was there to assist and really help a lot during studying,” reflecting how students experienced AI as a learning aid rather than a shortcut.

AI-resilient class time is the key to harnessing the potential for AI to enhance learning without replacing it. For example, in my senior-level K360: VBA and Application Integration course, I designed an exercise to help students experience how AI can assist in, but not replace, the human side of gathering and defining business requirements. While I have long taught what a business requirement is, students often found the concept abstract. This semester, AI made it possible to design a dynamic, interactive version of that lesson that finally clicked.

Students first received a short case scenario and worked in small groups to write three initial business requirements. We then reviewed those together using a Requirements Guide that emphasizes clarity, testability and stakeholder focus. Next, each group “interviewed” three simulated users: short narratives generated with AI assistance that contained realistic, messy curveballs and contradictions. These additional inputs required students to rethink and refine their requirements. Because AI allowed me to instantly create dozens of unique user perspectives, every team encountered different conflicts and trade-offs, closely mirroring what happens in real consulting or systems-analysis work.

After this, I introduced an AI Prompting Toolkit and asked students to use an AI tool to critique and improve their own requirements. They were then asked to reflect: What did the AI help with? What did it miss? Was it easier or harder to refine their requirements than they expected?

By the end, nearly all students agreed that writing business requirements is challenging, and that AI is a useful collaborator, not a substitute for analytical judgment. In fact, sometimes AI just sent them down tangential side-quests, distracting from their real goal.

Q: What have you learned about using AI that made it easier, more helpful or more targeted for your specific need?

A: What I’ve learned most about using AI effectively is that context and purpose drive quality far more than detail or word count. Early on, I focused on writing long, precise prompts. Now, I start by telling the AI why the task exists: who the learners are, what I need to achieve, and what constraints matter. When I supply that context, the AI’s output becomes not only more accurate but also more human.

I’ve also learned to treat AI as a co-creator that builds from my design, not as a blank-page writer. For example, when producing my course videos, I use ChatGPT to refine the instructional prompt, tone and flow until it captures exactly what I want students to experience. I then give that refined prompt to IU-licensed NotebookLM, which builds the final video.

I watch the video to ensure it was produced as I intended, and iterate again if needed. In this process, I’m still the author. But AI accelerates the production and frees me to focus on content and pedagogy rather than production mechanics.

Finally, I’ve realized that integrating AI into teaching isn’t just a technical exercise; it’s a design challenge. The goal is to shape experiences where students can use AI productively but must still think independently. Whether it’s refining requirements, prompting critically or evaluating AI’s mistakes, the real learning happens in the reflection that follows.

Q: What is one tip you’d share with a colleague about using AI tools?

A: Start small but start intentionally. Choose one specific task that drains your time or requires repetitive creative energy, such as drafting feedback, preparing summaries or structuring data, and explore how AI might assist with just that. Then evaluate whether the results are reliable, transparent and worth scaling.

One practical habit that has transformed my own results is asking AI to ask me questions before helping. After stating what I need, I invite the AI to clarify my request or flag missing context. This small adjustment gives the AI “permission” to identify gaps I didn’t notice, which leads to far more accurate, tailored and useful output.

Ultimately, the goal isn’t to get the AI to do the thinking for you; it’s to build a dialogue where each prompt sharpens both your own understanding and the AI’s response. And if you haven’t yet, I highly recommend taking IU’s GenAI 101 course for more practical prompting and usage tips. It is seriously good.

Note: IU offers a free AI course to staff, faculty and students to help them understand the tools available and how to use them.