AI in TCM Education: A Journey into Building A Fully Asynchronous Course

Author

• Sung Woo (ORCID: 0009-0005-6410-9472)

(While this is a member-only story, I believe in making knowledge accessible. Here’s a free access link: Click)

The Question

How do you keep students engaged when there’s no classroom?

This is a question for every instructional designer working on Distance Education, for sure. I was not an exception. I tasked myself with developing our first fully asynchronous course in our Traditional Chinese Medicine curriculum. As someone who has experienced the magic of in-person TCM education, where students can feel the pulse, observe tongue diagnoses in real-time, and engage in spontaneous discussions, the challenge felt enormous.

But here’s the thing about challenges: they often lead to unexpected innovations.

The Birth of an Idea

Traditional TCM education thrives on presence — the immediate feedback loop between instructor and student, the ability to demonstrate techniques in real-time, the organic discussions that emerge from shared experiences. Moving this into an asynchronous format wasn’t just about digitizing content; it was about preserving the essence of TCM education while embracing new possibilities.

That’s when I started thinking about AI — not as a replacement for human instruction, but as a tool to enhance and streamline the course development process.

The Real Challenges We Faced

Before diving into solutions, let me share what kept me up at night:

How do you explain complex TCM concepts without the ability to read students’ facial expressions and adjust on the fly? How do you create meaningful discussions when students aren’t in the same room at the same time? And perhaps most importantly, how do you support and ensure that students feel supported and engaged in their learning journey?

These weren’t just technical challenges — they were pedagogical ones that went to the heart of effective TCM education.

Enter AI: A Surprising Ally

I’ve always been fascinated by AI’s potential in healthcare education, but using it to develop a TCM course? That was new territory. Here’s how we made it work:

Transforming Lectures into Learning Experiences

We started with the professor’s recorded PowerPoint lectures — solid content, but we needed more. I used GPT and Claude to help transform these recordings into comprehensive written materials. Not just transcripts, but structured content that could serve multiple purposes:

• Accessibility resources for students who need written materials
• Reference materials for developing deeper learning activities
• A foundation for future course iterations

The key was using AI not to create content, but to help organize and present existing expertise in more accessible ways.

Reimagining Discussion Forums

Let’s be honest — most online discussion forums feel like homework. “Post once, reply twice” becomes a mechanical exercise rather than genuine learning. I wanted something different.

Working with AI, we developed discussion prompts that connected abstract course concepts to tangible real-world examples. For instance, in our DNA and protein synthesis module, students compare biological processes to familiar systems — like comparing DNA transcription to following a recipe or building from blueprints. These analogies made complex concepts more approachable and discussions more engaging. In another discussion, students explore connections between Western medicine’s understanding of energy synthesis in the human body and TCM’s concepts of Qi and Meridian — bridging two different medical paradigms in a way that deepens understanding of both.

The Human Touch in a Digital World

Despite the advantages AI brought to course development, we never lost sight of the human element. Meeting the federal and the accreditation requirements for Regular and Substantive Interaction (RSI) wasn’t just about checking boxes — it was about ensuring real connection in a virtual space.

Our faculty still post weekly announcements, engage in discussions, and provide personalized feedback. AI helps streamline these processes (like generating discussion summaries), but the genuine instructor presence remains essential.

Building a Safety Net: The AI Course Assistant

As we developed the course, another challenge emerged: how do we support students who are new to fully asynchronous learning? Traditional courses offer immediate support through face-to-face interactions, but asynchronous learning requires a different approach.

This led to an experimental solution: a custom GPT-powered course assistant chatbot. The goal wasn’t to replace human support but to provide an additional layer of guidance. I designed the chatbot with a clear principle: “Be friendly but do not give away answers; encourage students to do their own work while providing helpful direction.”

The development process involved careful consideration of ethical boundaries. How do we make the chatbot helpful without enabling academic shortcuts? How do we maintain academic integrity while supporting student learning? These questions guided our iterative development and testing process.

We’re still refining this tool, but the early testing has revealed both its potential and limitations. For instance, we discovered that while the chatbot excels at explaining course logistics and clarifying concepts, it needs careful guardrails to ensure it doesn’t inadvertently provide assignment answers. This learning process has been invaluable in understanding how AI can support education while maintaining academic integrity.

The Road Ahead: Challenges and Possibilities

This journey has taught me that AI isn’t a magic solution — it’s a powerful tool that requires careful consideration. We still face important challenges:

The content AI generates needs thorough review. While it’s great at organizing and suggesting structures, ensuring accuracy and appropriateness requires human expertise.

We constantly think about how to design assignments that encourage genuine learning rather than AI-generated responses. It’s ironic — using AI to create assignments that can’t be easily completed by AI.

Here’s a practical example of how we’re addressing this challenge: We use AI tools like GPT and Claude to analyze lecture scripts and design discussion activities that require deep engagement with the material. These activities might ask students to discuss specific concepts or examples mentioned in the lectures, effectively “forcing” them to review the content thoroughly. You can’t meaningfully participate unless you’ve done the work — and that’s exactly the point.

But here’s the crucial part: Nothing goes directly to students. Every AI-generated activity gets vetted by course instructors and subject matter experts. Always. It’s about using AI to enhance our teaching capabilities, not replace our expertise.

These challenges have also opened up exciting possibilities. Could AI help us bridge the gap between Eastern and Western medicine more effectively? Could it assist in clinical documentation while preserving the nuanced art of TCM diagnosis? Could we develop more sophisticated ways to verify student understanding while maintaining the authentic spirit of TCM education? These are questions I’m eager to explore.

Looking Forward

This course development experience represents more than just our first asynchronous course — it’s a step toward reimagining how we teach TCM in the digital age. By combining traditional wisdom with modern technology, we’re not just adapting to change; we’re actively shaping the future of TCM education.

I’d love to hear your thoughts. Have you experimented with AI in your teaching or clinical practice? What challenges and opportunities do you see? Share your experiences in the comments below.

This is the first in a series of articles exploring the intersection of AI and Traditional Chinese Medicine. Follow along as we navigate this exciting frontier in healthcare education.