Custom GPT for Learners

We have been looking at how

learning designers can use generative AI tools as a partner for their

learning design activity. So far, we have looked at examples where we created

different requests and prompts for tools like ChatGPT to help us with

different learning design tasks. But we can also look at generative AI tools

to help us track, engage all the

different elements of a learning experience

that we are designing. Our design process

will result in a large number of items

to keep track of. Learning objectives,

personas, activities, quizzes, assignments,

media elements, and so on. After a while, as the learning

experience gets bigger, it will become more

challenging to see both the big picture and the small details

about the experience. We can turn to

generative AI tools to help with all of this too. Aside from querying the normal generative

AI tools like ChatGPT, we can also create custom GPT

tools to support learning designers with tracking and assessing the different

elements of their project. Let's look at an

example of how we can create such a custom GPT, and some ways we might use it. For this example, let's

look at how we can create a custom GPT for a

class that I have taught with Dr. Rebecca Cantana

here at the university. The course is XR in education, a course about how we can

use technologies like virtual and augmented

reality to support learning. While we've taught the

course in-person at the university, in this example, we are creating a fully online

version of this course, and the custom GPT we are creating will help us with

that learning design task. Let's look at how we can create a custom GPT for

this course that we can then use to answer different design questions we might have about the course. This GPT is intended solely for the learning designers of

the course and not students. We can think of this custom GPT as

another design partner. We can go into ChatGPT to

create a new custom GPT, and then we can start

configuring it. First, we can give it

a name that reminds us this custom GPT is a design partner for our

XR and education class, so we'll call the GPT we are creating XR in Education

Design Partner. Next, we describe the GPT by saying that it serves

as a repository of completed design artifacts to assist in the development of

additional course materials. We can then provide instructions for how we would like the GPT to operate by saying that this

GPT provides up-to-date information about the

current design status of various artifacts

and course elements. Design team members are

encouraged to upload finalized artifacts to the GPT to ensure that it

remains up-to-date. Design team members could

query this GPT to develop new materials and ensure that the content is coherent,

consistent, and aligned. We can then set up some

conversation starters. You can think of

these as shortcuts to tasks you might frequently

want the GPT to do for you, and so these links will be displayed on the front

of the GPT page. Here, we create the following

conversation starters. One for course elements

that need to be developed, another for course elements that are aligned

with objectives, the third one called course elements that are

aligned with our personas, and finally, coherence and inconsistency in

the course design. As we create these

conversation starters, you'll see them displayed as buttons on the front of the GPT. This way, we can quickly

have the GPT work on these tasks when we press

the corresponding buttons. Finally, we need to

give the custom GPT some information or knowledge about our XR in

education course. This information is found in different course files

we have for the course, so we can upload documents

like the course syllabus, a sample reading, a

quiz, some personas, some initial scripts

that we created about learning and affordances, etc. We can do this initial upload of information about the class, but as we move forward with

the design of the course, we will be adding new

information that we create, such as new quizzes and scripts or perhaps new personas

and learning objectives. Once we have all of this

information uploaded, we can create the custom GPT for our course and share the GPT with other members

of our design team. Remember that our

GPT currently only has some initial information

about our course, but that's enough to

let us see how we might use our custom GPT with

some small examples. Let's use some of our

conversation starters. First, to ask our GPT about what course elements are aligned with our

learning objectives. Here, the GPT returns with

quite a bit of information. It does start out by giving

an assessment of how it sees the course aligning elements of the course with the

learning objectives. It notes that the

course structure incorporates

theoretical concepts, practical applications, and immersive technologies to enhance learner engagement

and understanding. It then goes on to list some of the main course elements and how those elements align with

learning objectives. Now, as always, you,

as the designer, should review this feedback for accuracy and to see that

it all makes sense. But as you can see, as you start adding more course elements

to the GPT over time, you can continue to use the

GPT to provide you with useful summaries

and overviews to help you gauge different

aspects of the course. Let's look at another example. We can start again at the front page of the

GPT with a new chat. This time, we are

asking the GPT to tell us what course elements

still need to be developed, given what we've noted

in the syllabus. Here, the GPT gives us a summary of what it thinks

we still need to do. For example, it tells

us that we still need to develop certain

course modules, and it describes elements

of those modules. It reminds us that, especially for a course on

extended reality, we should develop some more interactive elements

and simulations. It describes some quizzes and materials that

we should create, and it reminds us that

we need to think about some additional

accessibility and inclusivity features

for the course. There's more that we could do, but this should give you an idea about how learning designers can create custom GPTs to support the learning

design process. Creating a GPT is still

a somewhat new activity, so you may need to really

experiment and use different types of resources

to create a useful GPT. But hopefully, you can

see how developing a custom GPT can really

serve as a design partner in different ways to help

you keep track of all the elements that

go into creating an expansive

learning experience.