.webp)
In Module 2 we applied some of the theoretical foundations of Module 1 by exploring some practical case studies. Hopefully you’ve had a chance to rapidly iterate on an AI prototype based on a topic of your choosing.
In Module 3, we will explore strategies for implementing our prototype in a variety of learning environments and how to effectively iterate on feedback.
New technologies can create new possibilities for learning design. In this section, we will take a step back and try to look at the big picture as we consider how to take a more holistic approach to curriculum design.
Given the recent advancements in AI, there has been increased interest in its educational potential. This has spurred a dizzying array of publications on its educational use cases, many of which contain the fears and misconceptions that we have touched upon in Module 1.
In anticipation of these challenges, OpenAI, the makers of ChatGPT, have published their very own blog post on Teaching with AI (with some example use cases that you may find familiar!).
You should now be a pro at evaluating these prompts and what output a large language model might produce. You can even try refining your AI Template from the last module by copying and pasting some sections of these prompts to see what you get.
All these blog posts aside, throughout the remainder of this Module we will explore some deeper ways to think about teaching and learning with AI.
A good first step is to take a huge step back and determine the purpose for using AI in the first place.
There are many reasons why you and your students might want to use it. It's important to note here that in Module 2, we did not cover any use-cases related to writing essays or doing “traditional” learning activities but with AI merely tacked onto it. In other words, we mostly steered clear of doing the same old things but with newer technologies.
Instead, we could say that many of the examples we have encountered in Module 1 and 2 revolve around understanding the power of AI technology in driving motivation by providing experiences that are otherwise infeasible to accomplish at scale.
Another way to frame our purpose for using AI is by considering whether or not we want to use AI on students and/or use AI with students. We can define each simply below:
You can, of course, do both of these things and we strongly encourage you to do so.
One slight nuance we should address is that interacting with an AI agent might seem like an example of using AI on a student, but this is not necessarily the case for two reasons.
The first reason is that, as you should now know, these technologies are generative by virtue of their probabilistic nature. This means that, if designed appropriately, no two experiences will ever be exactly the same since these generative AI large language models provide infinite variety through clever and effective prompting. When a student engages in a conversation with a well designed AI agent, they are actively co-producing the conversation.
The second reason is that we recommend the AI agents you design be holistically integrated into a curriculum, as we will discuss below. When this is accomplished, what can seem like passive consumption turns into a stimulus that prepares students for future active learning opportunities.
Determining your purpose for using AI will depend on your learning domain and other factors. However, once you’ve spent some time thinking through this, you can now turn to identifying your learning audience and their learning needs.
A 2016 BBC piece entitled “What Do Prince Charles and Ozzy Osbourne Have in Common?” pointed out that they both share the following characteristics:
From the perspective of some analytic engines that decide which advertisements we see online, “the Prince of Wales and the Prince of Darkness are almost indistinguishable.”
However, while they share identical demographic information, you probably wouldn't want to send the same advertisement to both of them.
Personas shouldn't be about demographics. Personas should be about the problems and challenges people face.
Critics of this focus on demographic information alone when shaping personas suggest that when you define your persona in broad terms, what you target them with will be equally blunt.
We can extend this thinking to learning challenges or learning needs. When identifying what learning personas compose a target audience, we need to ensure we are also including their learning needs or learning challenges.
When we conduct such a Learning Needs Analysis for our target audience, we can better map out a cohesive learning journey and ensure we can solve their learning needs with surgical precision.
Identifying learning needs or learning challenges will vary significantly not only by learning discipline but also by factors like culture, demographics, and what you already know about your students.
Challenge 🦾: Revisit Module 2’s Case Studies where we outlined a Learning Problem and how Wonda was used to address a solution. Can you define what the learning needs were? What might have been missing or assumed in each case study?
Integrating new technologies into the curriculum is never easy, since they often require a cultural change – a pedagogical shift – not just a technological one.
Once you have defined your learning audience and their learning needs, you can start to think about how and where the AI Simulations can be situated within a learning journey. Viewing this map of your learners journey through a variety of different technologies is a great way to make visible this pedagogical shift.
Challenge 🦾: Revisit Module 2’s Case Studies where we outlined a Learning Problem and how Wonda was used to address a solution. Can you draw a mental map of a learning journey? When were Wonda’s AI features used? What came before and after it?
Implementing a technology driven learning experience will often involve layers of technologies. One important layer that may determine the success or failure of your learning experience is the learning management system (LMS).
In this part, we will discuss some of the things you might want to consider when implementing your Wonda AI experience through an LMS.
Every LMS has certain constraints and affordances which will prohibit or allow certain types of learning experiences to be possible.
An affordance, generally speaking, is a design concept that refers to the qualities or characteristics of an object or environment that suggests how it can be used or interacted with, based on its inherent properties.
In the context of a Learning Management System (LMS), affordances refer to the features and capabilities of the LMS that enable various types of learning interactions and experiences for both learners and instructors.
Most LMS’ have features that allow for some general types of learning interactions. We can categorize these 5 main types of content delivery and consumption as follows:
These 5 categories cover different dimensions of learning preferences and scenarios. Each can cater to the diverse needs of learners and instructors, allowing for a more effective and engaging learning experience.
But many LMS’ also have limitations.
Wonda’s AI features offer a completely new dimension that can boost the efficacy of all 5 categories above, creating a holistic ecosystem of powerful learning interactions.
You already have a good understanding of this new dimension’s possibilities from Module 2, but what we will explore more deeply here is how to maximize Wonda’s AI features within your learning design.
Wonda’s platform already has LMS connectors set up within Canvas, Moodle and Blackboard as well as other LMS’. You can learn more about setting this up by visiting Wonda’s documentation on LMS Integrations.
You can also find us on Instructure's Emerging AI Marketplace.
Even if you’ve done the intensive learning design work of identifying your learning personas’ needs and mapping out a learning journey, implementing new technologies in learning environments often involves significant cultural challenges and institutional barriers. Misunderstandings of AI can lead to colleagues or other stakeholders not valuing or appreciating the use of AI to achieve learning outcomes.
The most important way to overcome institutional barriers is to identify like-minded stakeholders and co-design with them. When you work with such people who “get it”, you are more likely to create something compelling and scale it enough to generate interest and momentum.
This will often require creating AI-infused experiences that can be easily scaled amongst a certain learning discipline or domain. For example, crafting an AI experience like Case Study 1 and sharing it amongst other Spanish instructors is a great way to partner with the right stakeholders and generate momentum.
On the other hand, you should also identify people who will be hard to convince – that is, those who represent a significant institutional barrier. When you listen attentively to their concerns, you can then create and iterate upon a variety of prototypes in Wonda, which is entirely possible since it is so fast and easy to design and develop on this platform.
It’s one thing to address the concerns of your detractors using words and arguments – it's entirely another to show them what could be possible by allowing them to experience it for themselves.
💡 TIP: Start with a small targeted activity (to overcome the novelty effect described below) and to sanity check anything you may have missed, such as identifying time constraints and technological challenges you may not have considered.
Avoiding the Novelty Effect
The novelty effect is a psychological phenomenon where individuals show heightened interest and engagement when exposed to something new or unfamiliar. This increased curiosity and attention are typically temporary and fade as people become more familiar with the novelty. It is often leveraged in marketing, product design, and education to capture attention and encourage exploration, but it requires a balance with familiarity to be effective in the long term.
When things are new they can be exciting but they could lead to undesirable interactions with the AI, which may work against your learning goals. One way to get around this is to create several touch points with an AI (ideally visualized on a learner journey map!).
For example, ensure all your students have done some similar interaction in ChatGPT first so that they get their “curiosity” to push the AI to its limits out of the way as soon as possible. Or you could re-create yourself in Wonda’s platform and introduce it to your students in a playful way to see how they react before designing something that targets pedagogical learning needs.
Playtesting
You have probably heard of the term “user testing”, but may have not heard of “playtesting”.
Playtesting and user testing, though they might seem different, share a common goal of improving experiences with products by involving real people.
In playtesting, the focus is on enhancing typically game-based experiences by observing players and gathering feedback on gameplay elements. In contrast, user testing applies to a broader spectrum of products, like websites and software, with the aim of enhancing usability and overall user satisfaction.
Both methods rely on real people to provide valuable insights that drive iterative design and ensure that products better meet the needs and expectations of their intended audiences. So, while the specifics vary, the underlying principles are the same.
In another sense, Wonda’s AI features contain the best of both worlds: they encourage and invite others to play (i.e., playtest), but the experiences are often delivered through the web as software (i.e., user test).
When in an educational context, however, we want to ensure that we also think about our users or playtesters as learners, too.
This is a natural extension of identifying our learning personas above. We not only want to observe their user behavior and player behavior but also ensure that we are solving learning problems and addressing learning challenges.
We can do this by engaging in a cycle of playtesting small, targeted prototypes and iterating upon it.
But how do you know what to iterate on?
Every playtest should be accompanied by a short, anonymous feedback form. You can create such feedback forms easily using google forms, airtable or even within your LMS.
Be sure to ask for honest feedback. Without honest feedback, you won’t be able to prioritize and iterate on the right things. Students and co-workers will often be more kind than usual, so be sure to reiterate that the purpose of asking for honest feedback is to improve the learning experience.
It’s worth stressing that much of the initial work on defining your learner personas and mapping a learning journey should have put you in the best possible position to develop a great prototype. If you skipped or neglected the importance of these steps, you will likely find your feedback to be more negative than you expect.
Once you have your feedback, strategize by prioritizing what is more important given your context and then iterate, iterate, and iterate again!
Lastly, have fun! If you are having fun or deeply engaged in what you are creating, it is very likely that your learners will, too.
Also, thank you in advance for sharing your feedback about this course which is still in its early version!