Supercharge Your Workflow:
How I'm Using AI in My Instructional Design Courses.
How I'm Using AI in My Instructional Design Courses.
In the world of Instructional Design (ID), we're constantly searching for ways to work smarter and create more engaging learning experiences. I've found that integrating Artificial Intelligence into my workflow has been transformative, and the key is to stop thinking of AI as just a tool and start seeing it as a collaborator.
Different AI platforms have unique strengths, so knowing which one to use for a specific task is crucial for efficiency. This month, I want to highlight two practical ways I’ve successfully woven AI into my design process, improving both my productivity and the end-user experience.
1. Building Intelligent, Interactive Forms
Instead of using static, one-size-fits-all questionnaires, I used AI to create a dynamic form for a course I'm developing. This form adapts to the user's input, creating a personalized and more effective learning interaction.
Here was my process:
Train the Model: First, I "trained" the AI by providing it with the core content of my course—learning objectives, key concepts, and FAQs. This essentially turned the AI into a subject matter expert on my specific topic.
Create Smart Qualifiers: I designed the form to ask branching questions. Based on a learner's response, the AI presents a relevant follow-up question, guiding them down a personalized path instead of a rigid, linear one.
Integrate Learner Support: For tricky questions, I built in an AI-powered hint system. Users can request a "clue" or a "hint," and the AI provides tailored guidance based on the course material it was trained on.
Automate Follow-up: Finally, I connected the form's results to an email system. Upon completion, a summary of the user's responses and performance is automatically sent to them and/or their instructor for review.
2. Streamlining the Training Needs Analysis (TNA)
After the success with interactive course forms, I realized this approach was perfect for another critical ID task: the Training Needs Analysis (TNA). A TNA is all about asking the right questions to uncover performance gaps, and AI is the perfect partner for this kind of exploration.
By applying the same principles, I created a TNA survey that digs deeper than a standard questionnaire. For instance, if an employee indicates a weakness in "communication," the AI can immediately ask targeted follow-up questions like:
"Are you referring to written communication (emails, reports) or verbal communication (presentations, meetings)?"
"Can you provide a recent example where you felt your communication could have been stronger?"
This allows me to gather rich, detailed data right from the start, making the entire analysis process faster and far more accurate. It transforms the TNA from a simple survey into an insightful diagnostic conversation.
A Look Under the Hood: Scaffolding My AI Prompts
To get the sophisticated, interactive form I wanted, I couldn't just throw a single, complex request at the AI. I quickly learned that the prompts themselves needed to be incremental. I had to scaffold my instructions, much like we do for our learners, building from a simple foundation to a more complex final product.
Here is the four-step process I used:
Step 1: The Definition Prompt
This first prompt is simple and high-level. It establishes the core purpose of the form without getting bogged down in details. My goal was to give the AI the general topic and intent.
"Please create a form that asks questions that help to choose a topic to create a compelling personal and social digital story that matters to you. It can be based on class discussions, personal stories, or creative imagination."
Step 2: The Qualifier Prompt
Once the AI understood the basic goal, I followed up with a more detailed "qualifier" prompt. This is where I injected the pedagogical framework of my course, giving the AI specific guidelines and the scope it needed to work within. I instructed it to build the form's logic around a specific reflective process for the user.
First, the form would prompt the user:
"Think about your own experiences, interests, or values. Write a short paragraph about something you care about."
Then, it would use that input to guide them through deeper reflection:
"Why does this matter to me?"
"What message or feeling could I share with others through this story?"
"How could my experience connect with an audience the way the digital stories we studied did?"
The final instruction was for the AI to process the user's answers to these questions. Based on that input, it provides several tailored story ideas along with a sample outline for each. The result is a perfect balance: the ideas are broad enough for the learner to make them their own, but the structure is guided enough to dissipate the procrastination that comes from a lack of clarity.
Step 3: Refining the Layout and Design
The initial results from the AI were functionally correct but didn't match the visual aesthetic of my course. This is a common step where the designer's eye is crucial. I identified the specific hex code for the main accent color in my style guide and decided to simplify the overall look.
My next prompt was direct and specific, leaving no room for ambiguity:
"Now let's focus on design and layout. Make it minimalistic, with a white background and accents using color #6293C0. No emojis."
This simple command instantly aligned the form's appearance with the rest of my course materials, proving that for design elements, precise and clear instructions are key.
Step 4: Adding Advanced Functionality (Email Integration)
A form is great, but I needed a way to capture the results. My goal was twofold:
Give the user a chance to receive a copy of their results via email (which they could optionally input).
Get a copy of every submission sent directly to my project email (edthestorylab@gmail.com) for review.
This step was perhaps the most difficult, especially with my limited coding experience. After asking around and consulting with various AI chats, the recommended solution was to use a third-party service to handle the email function from the web page. That's when I was introduced to EmailJS.
The initial configuration was simple, but then I needed to integrate it into the code. This is where I shifted my workflow. I had been using ChatGPT and Claude.AI, but I decided to complete this task using Gemini AI. The primary reason was its feature of being able to see the generated code and the rendered output side-by-side, which is incredibly helpful for troubleshooting visual and functional elements simultaneously.
I needed to locate the part of the code that handled email commands and provide my specific EmailJS account information. It looked like this:
JavaScript
const EMAILJS_CONFIG = {
serviceId: 'your_service_id',
userTemplateId: 'your_user_template_id',
publicKey: 'your_public_key'
};
I'll be honest—this wasn't a one-click fix. I had to troubleshoot a lot to get the variables right and ensure the function was called correctly. But after a few hours of focused work, I had a fully functional system. Now, the creative ideas generated in the form don't just disappear; they land directly in the user's and my inboxes, ready for the next stage of the storytelling process.
Final Thoughts: AI as a Creative Partner
Witnessing the first result of this process was genuinely surprising. More than just executing a command, the AI helped organize my core concepts into a practical and usable structure. It acted as a creative partner, activating different perspectives on the problem that I might not have reached on my own.
It wasn't perfect on the first try, naturally. But after just a few more iterative passes—refining the prompts and tweaking the output—the tool was good to go. This journey reinforces the idea of AI as a powerful collaborator in our workflow. With clear, scaffolded instructions and a willingness to refine, we can leverage these tools to build smarter, more personalized learning experiences faster than ever before.
