Hub Designers win Best-In-Track at OLC for AI-Powered Instructional Design

A professional conference setting featuring McMichael and Huffman presenting at the OLC event. The image is divided into two sections. On the left, attendees sit around a table covered with a blue tablecloth. McMichael, wearing a black vest, gestures while speaking. Other participants listen, with notebooks and water bottles in front of them. On the right, Huffman stands at the front, holding a microphone and gesturing as she presents. Attendees in the foreground take notes and engage in discussion.

Jonathan McMichael and Sue Huffman, learning experience designers at the Learning and Teaching Hub, were awarded the Best-in-Track honor by the Online Learning Consortium (OLC) – a globally recognized leader in advancing quality digital education. Their session, “AI Design Challenge: Make Your Own Instructional Design Tool!”, was selected as a standout in the Innovative Learning Environments and Technologies track at the 2024 OLC Accelerate Conference in November.

Huffman emphasized that this recognition reinforces her “commitment to openly sharing” processes, successes, and challenges to support others. McMichael echoed the importance of disseminating lessons learned, sharing that “all the tinkering we’ve done with AI” has not only benefited ASU, but has also “resonated with the larger educator community.” 

This session served as a milestone to share key insights from their work as part of ASU’s AI Innovation Challenge.

AI Innovation Challenge

Graphic image of the AI Innovation Challenge website on a laptop. The website reads, AI Innovation Challenge. Arizona State University, in collaboration with Open AI, proudly presents the AI Innovation Challenge.

As part of ASU’s partnership with OpenAI, the university launched the AI Innovation Challenge in 2024, inviting faculty and staff to propose ways to maximize ChatGPT Edu accounts to enhance their work. Initially, the Hub submitted two project proposals, which were later streamlined into a single initiative. The Hub’s development team focused on developing custom GPTs—specialized versions of ChatGPT tailored for specific tasks–primarily focused on instructional design and supporting faculty in their teaching.

Through the AI Innovation Challenge, McMichael, Huffman, and their colleagues at the Hub developed more than 13 custom GPTs, ranging from AI agents that help refine learning objectives and assignment descriptions, to AI thought partners that assist in designing authentic assessments.

This challenge provided a foundation for deeper exploration into AI’s role in instructional design.

AI as a thought partner, not just a tool

A key principle emerged from their exploration as part of the AI Innovation Challenge: AI should be a thought partner, rather than just a tool. Instead of automating tasks, this approach ensures that AI enhances strategic decision-making, course design, and student engagement. 

The Hub’s GPTs do not replace the expertise of faculty or designers, nor are they merely content generators. Instead, they are designed to enhance collaboration between subject matter experts and designers, with the overarching goal of fostering student success. Developed with the Fulton Schools of Engineering instructional design context, educational theoretical frameworks, and intentional workflows in mind, these tools promote collaboration, dialogue, and iterative refinement—ensuring that human expertise remains central to decision-making.

Four individuals are working at a computer. A table sign that mentions AI is in the foreground.

Their tools reflect a core insight from research on AI-powered tools education [1]:

Thoughtfully designed AI makes human expertise matter more. When educators build their own AI tools, they not only leverage technology but also refine their expertise, explore new possibilities, make informed decisions, and engage in deeper professional dialogue about teaching and learning.

AI is most effective when used collaboratively. Educators see greater benefits when AI supports backward and intentional planning—starting with learning goals and designing tasks accordingly. The difference here is seeking input from AI tools, rather than seeking an output. [1]

“I knew we were onto something when I saw how using some of our AI prototypes transformed our conversations with faculty and fellow instructional designers,” McMichael shared. “Initially, there were concerns that AI might disengage users, but we found the opposite; thoughtful engagement actually increased as we integrated these tools into our workflow.”

The impact of their work

At OLC, McMichael and Huffman’s presentation highlighted the power of AI instructional design tools to enhance course design, student engagement, and faculty support. Their session provided a hands-on look at AI’s role in enhancing human creativity and decision-making.

“It became clear that participants [at OLC] came from a variety of contexts – university and department sizes, a variety of responsibilities, and diverse access to Large Language Models,” Huffman reflected, recognizing that the range of experience and knowledge among participants created rich discussions. “Randomly seated at round tables, I could hear groups connect over their differences and offer ideas and perspectives.”

Key takeaways from their presentation:

  • The AI-Tool Development Process. The pair shared insights from the Hub’s AI Innovation Challenge Project, where a structured development process guided testing, refinement, and feedback collection. GPTs were evaluated through real-world instructional challenges, with feedback gathered on usability, utility, and areas of improvement.
  • Building an AI Tool. Participants in the workshop completed a structured worksheet (Building an AI Tool Worksheet) to think deeply about their AI use case, output, interaction, and prompt.
    • Target Use Case. Participants grappled with reflective questions on a use case (e.g., Who is this for? How does this task fit into their bigger picture?) to put together a statement – “As a [Role], I want to [Task], so that [Goal].”
    • Target Output. Next, the worksheet had participants consider the output, and perhaps most importantly, what the Minimally Viable Product would be and the essential elements the output must contain.
    • Storyboard. To visualize the user experience, participants used storyboarding – through drawing, writing, or other creative methods – to map out key stages of an interaction and integrate back into their workflow.
    • Building a Prompt. The next step was prompt engineering—writing effective AI prompts that yield accurate, useful, and consistent results. Huffman and McMichael provided templates and examples to get participants started.

Explore their OLC slides, worksheet (including an example), and instructional design starter kit in this folder.

Lessons learned

McMichael and Huffman reflected that their biggest insights stemmed not just from building AI tools but from the collaborative, iterative process.

“The real breakthrough for me was when Sue developed the assignment description tool,” McMichael said on his biggest ah-ha moment. The Assignment Description Builder was an example they provided in their presentation at OLC. “Beyond its utility, developing it as a team showed me that we weren’t just building tools; we were building a framework for how to talk about and use AI meaningfully. That moment made me realize that if we could articulate this process, we could empower others to create their own tools, too.”

Huffman similarly reflected, “Interestingly, my ah-ha moment also came from building the assignment description tool! I spent so much time working through our storyboarding process – building a knowledge base and refining my ideal outcome – and less time engineering a prompt. I realized that building AI tools can expand my abilities and skills as an instructional designer.”

A professional conference setting featuring McMichael and Huffman presenting at the OLC event. The image is divided into two sections. On the left, attendees sit around a table covered with a blue tablecloth. McMichael, wearing a black vest, gestures while speaking. Other participants listen, with notebooks and water bottles in front of them. On the right, Huffman stands at the front, holding a microphone and gesturing as she presents. Attendees in the foreground take notes and engage in discussion.
On the left, McMichael is seated at a round table, wearing a vest and engaging in discussion. On the right, Huffman stands at the front, presenting to the audience at the OLC conference.

McMichael and Huffman’s worksheet and resources equips others with the tools to create, refine, and implement AI-driven solutions that enhance learning, decision-making, and instructional innovation. The presentation challenged attendees to experiment, iterate, and develop AI tools aligned with their expertise and values.

Their work reinforced a key message: the future of AI in education is not about what AI can do alone, but what we can do with it, together.

“Our goal wasn’t just to showcase what we built, but to set a precedent of openness, collaboration, and intentionality with how we approach AI,” concluded McMichael. “We hope this presentation helped attendees see that they, too, can take an active role in shaping AI’s impact in education.”

Interested in developing your own AI tools for teaching and learning? Join our GenAI Community of Practice and be part of the conversation!

References

[1] S. Keppler, W. Sinchaisri, and C. Snyder, “Backwards planning with generative AI: Case study evidence from US K12 teachers,” Aug. 13, 2024. [Online]. Available: http://dx.doi.org/10.2139/ssrn.4924786.