If you’re interested in learning about how AI can support instructional design or the creation of course materials, you’ve come to the right webinar recap.
Are you curious about how generative AI and large language models (LLMs) can be used to design and create exceptional learning experiences? To give us some insight, we turned to Dr. Luke Hobson, Senior Instructional Designer and Program Manager at MIT xPRO, the Founder of Instructional Design Institute, podcaster and author.
Hobson joined forces with D2L to present a webinar called “Using AI to Design Better Online Learning.” Turns out, there’s a lot of interest in this topic. While we’d recommend watching the webinar on-demand at your leisure, the following piece is a brief recap of what was covered.
Your prompt is my command.
That’s an enticing promise, especially for educators who are notoriously stretched thin and underfunded. The allure of these new digital tools that could help educators create and design learning experiences is hard to resist. But could AI really be the answer?
According to Dr. Luke Hobson, it’s not perfect, but it can certainly help.
Hobson shares that he was interested to learn how LLMs and generative AI tools like ChatGPT, Google Bard and DALL-E, could be used to generate course materials and complete various time-consuming tasks including but not limited to the following:
He discovered that “from all of these different types of trials and experiments, there have been some things that certainly have worked quite well. While AI is not perfect and is never going to be perfect. In my opinion, [it’s a] kickstarter for ideas.” Like Robin to our Batman, it could be the trusty sidekick that educators can rely on when push comes to shove.
Walking a mile in someone else’s shoes is the essence of scenario-based learning.
These fabricated situations give learners a chance to apply theoretical knowledge to gain practical experience. As valuable as this kind of instructional approach may be, designing and creating them can be a challenge.
So, Hobson turned to ChatGPT-4. Based on his initial prompts, it was able to generate a scenario involving a fictional company. Once again, it wasn’t perfect right from the get-go, but he was able to refine and improve the output with additional prompts.
When satisfied, he turned to DALL-E within ChatGPT-4 to create a selection of logos based on his prompts, to make the fictional company even more believable.
As the saying goes, a picture is worth a thousand words. But finding the right image, navigating copyright, or attempting to design an image yourself is a heavy lift.
Hobson found remarkable success using DALL-E within ChatGPT-4. With his prompts, an image began to take shape.
Much like the Holodeck on the Starship Enterprise, simulations involve engaging in a hypothetical situation to understand what a similar experience could be like in the real world. For those unfamiliar with the Star Trek reference, think of a flight simulator and how young pilots can learn to fly while remaining safely on the ground. It’s highly effective and very costly tech.
“Wow, people love this if we do it right,” so why not try to make your own?
Hobson took to ChatGPT but ran into some unexpected resistance this time. AI pushed back, operating under the understanding that it was not capable of generating simulations. But, after some time, Hobson was able to convince it otherwise. Apart from this minor hiccup, Hobson uploaded several types of data to provide context and examples. “Now, the first type of prototype, it was not pretty,” but after a lot of “back and forth,” the simulation began to take shape. “Most importantly, folks, I was able to download it and to essentially make it my own.”
If generating simulations is a little ambitious, start with putting ChatGPT to use in role-playing script generation and in the creation of negotiation exercises.
Universal design for learning (UDL) is top of mind for many educators looking to create and design experiences that work for all learners.
Hobson asks, “What does this actually look like inside of the real world?” Taking and customizing the same content to fit various learners can be a monumental task.
Turning to Gemini, formerly Google Bard, Hobson uploaded an article he had written about spaced repetition. He then asked it to rework the content into two different versions: one that graduate-level students could understand and another for sixth graders.
There’s no end to how educators and course designers can lean into AI.
Hobson also suggests asking AI to access materials and assignments through a lens of diversity, equity, inclusion, justice, and belonging (DEIJB). It can offer guidance and recommend more inclusive language or wording.
While summarizing the Hobson webinar, we wondered what ChatGPT might have to add. So, we asked if an instructional designer were to turn to AI to help create educational materials and resources to reduce workload, what could they expect?
Source: ChatGPT 3.5 (screen capture, May 14, 2024)
Hobson began the webinar with a goal to inspire and provide actionable insights on how AI can support your instructional design aspirations, the generation of course materials and more. If there’s one main take away it would be, “I would highly encourage you to be able to experiment and see for yourself what exactly you can come up with.”
At the very least, remember “to keep having conversations, keep sharing, keep listening.” Because that’s how we teach and learn. Not convinced? Check out Hobson’s video below for even more pointers on how to use ChatGPT.
Grab a coffee, get comfortable and enjoy the complete webinar experience here. You’ll be glad you did.
Join us as we explore strategies and techniques to spice up your online courses and keep your learners learning. We’ll dive into Brightspace’s advanced analytics tools, customizable content and optimized course management.
Written by:
Educators and training pros get our insights, tips, and best practices delivered monthly
