The 3 Instructional Designers in adopting artificial intelligence
By: Linley Myburg
A time of invitation
Knock, knock
Who’s there?
AI
AI who?
Ay, I have an invitation for you…
Your typical Knock, knock joke usually stops here. But this is not a typical knock-knock joke. How can it be if the punchline about AI can be written by AI? Just as you would know what should follow after ‘Knock, knock…’ Generative Artificial Intelligence (Gen AI) is essentially an advanced predictor.
Think about the world of Instructional Design that you find yourself in – would you open the door ifwhen AI comes knocking? We are not referring to the iRobot human-machine (even though a knock on the door by a robotic arm no longer remains only for sci-fi novels, but that is another discussion entirely). Here we refer to the proliferation of Gen AI such as applications like ChatGPT and how that would influence our current landscape as an instructional designer.
What is Gen AI – and explain it to me like I am an 8-year-old? We might think about it like this, picture a crate of assorted LEGO® where a child uses their imagination to put blocks together in such a way as to create an alien spaceship or a dream house. Gen AI also creates. And creates various types of content including text, audio, and images, but instead of LEGO® blocks, Gen AI uses code.
So, what is your instructional design barometer for AI adoption, and are you the one that has the door swung open like the Vaal Dam floodgates, or are you the one saying “not by the hair on my chinny chin chin"? Will this big bad wolf blow my house in?
With daily updates, quick modification, and vast learning as adoption expands by the minute, AI has become sophisticated to the point where even master reviewers cannot differentiate article abstracts authored by humans or machines. Or how about when ChatGPT-4 passed a medical board examination, law school bar exams, and Wharton’s MBA exam?
In the realm of Higher Education, with AI’s direct influence on our students’ practices and cognition, from writing, and asking questions, to completing assessments, the very way students learn changes. And just how do we as instructional designers advocate inquiry and expression in the curriculum if our Gen Z students have access to this advanced predictor in their pocket? Do we have an instructional design mandate of Evolve to Solve – and just how far do we allow AI a part of our best practices from authentic assessments to accessibility?
As of writing this, the numbers look as follows:
1.6 Billion website visits on ChatGPT in March 2023
570 GB of data that ChatGPT was trained on using Wikipedia, books, and other web data
25 000 word limit that can be processed by ChatGPT-4
57% of South African respondents on Statista’s survey were willing to trust AI systems (Johannes, is that you?)
4 - Thenumber of times my non-tech mother has asked AI for dinner ideas.
The 3 types of Instructional Designers
So, whether friend or foe, whether greeted with healthy scepticism or not, AI is here. Knocking! Using Rogers’ seminal work as our lens, we may expect that AI, as with any innovation, would have adopters at different stages. And therefore 3 types of adoption responses within the instructional design may emerge – which are you and what are some things to consider?
Type 1: Innovators and Early Adopters
Type 1 Instructional Designer
This work was created with the assistance of DALL·E (2023). Prompt: A welcoming face opening a door. The door has a wreath on it, a one-line drawing.
You will notice that images do not always generate as you would expect. This is where you practice your prompts.
About 16% fall in this group and not only is the door swung open but there is a welcome mat and wreath to match! This type of instructional designer is eager to try new ideas and quite comfortable accepting the occasional setback when a new idea is unsuccessful – this is valuable in fostering skills like prompt engineering.
Most likely to have:
Tried and tested more than five different apps
Subscribed to AI newsletters and blogs or secured a spot on the waiting list for a new app release
Maxed out their free version limits on Midjourney or the like (erm, guilty!)
Learned how prompt tweaking makes outputs more specific
What to consider:
As a champion of experience, consider where you could steer or participate in emerging institutional policies or strategies. As these instructional designers are also likely to be opinion leaders, a good start is to create opportunities for or engage in open debate on the impact of AI within your division. You may not necessarily be a superfan of AI but as you approach your investigation with caution, you can make informed choices – thinking critically about the digital landscape. Other vital processes as a starting platform could include:
An institutional or faculty protocol on ChatGPT (and the latest AutoGPT) and AI app use within academic integrity policies as part of student orientation processes. This could also include correct acknowledgement and referencing practices of AI use.
Development of student, and faculty workshops and guides to teach prompt engineering and fact-checking as a 21st-century skill.
Planning curriculum and assessment strategies that motivate inquiry and expression, leveraging AI as a tool.
If you are interested in research, explore the use of AI in the research process. Learn from others who have already published papers using a platform like ChatGPT. From their research question to write-up – observe their ethics concerning AI use.
Be kind – don’t spam your Type 2 and Type 3 instructional designers. In your excitement of reinventing your productivity through AI systems, you might have forwarded every article and tool suggestion that has come your way. I mean, a new day – a new AI app, right? Instead of expecting others to keep up, consider meeting them where they are by sharing your real practical examples. Why not test out a new app in a module, set up a chatbot training guide, or attempt other work-saving tips – synthesising the insights to showcase lessons learned?
Type 2: Early Majority and Late Majority
Type 2 Instructional Designer
This work was created with the assistance of DALL·E (2023). Prompt: A curious face peering out through a door, one-line drawing. This generated image looks oddly similar to the work of South African writer Phillip de Vos.
This group may be akin to someone who keeps the door chain on and peers through the gap. Instructional designers in this group would make up around 68% of all adopters and will adopt AI just before (early majority) or just after (late majority) the average instructional designer does.
Most likely to have:
Tried and tested 2 or 3 different apps
Mainly used ChatGPT, lightly, maybe even asking it to write a poem about life, tell a joke, or provide a wine pairing recommendation.
Most likely to take on an observer role, learning indirectly about AI through their colleagues
What to consider:
Practice your prompt engineering skills. Not only can ChatGPT explain a topic but you can leverage it for multiple other functions. UNESCO’s quick start guide extended upon Mike Sharples work and captures it well, some roles include leveraging ChatGPT as a – ‘Socratic opponent (to develop an argument), personal tutor (providing immediate feedback on progress), a co-designer (support in the design process)’, and many more. It is especially useful for creating rubrics, and quiz banks and fares well when tweaking prompts to include academic pitch and higher-order cognition.
Hone your fact-checking skills. Through practising your prompts, you will learn the limits of ChatGPT. It is important to remember that ChatGPT does not provide an answer but rather an output. It can be confidently wrong! The GPT in ChatGPT stands for Generative Pre-trained Transformer, it generates an output from what it has previously learned (remember the LEGO® blocks). It takes text input and transforms it into what it predicts the most relevant result will be. It is therefore important to be aware of the inaccuracies, biases, and gaps.
Because you are more likely to observe the benefits from your colleagues or seek evidence of those benefits, consider participating in one of the many step-by-step courses on AI. As a start, know what is meant by the following terms: large language model, machine learning, and tools that are: text-to-text, text-to-image, text-to-other, or image-to-image.
Subscribe to one AI blog and ensure to go for a catch-up coffee with a Type 1 instructional designer once in a while.
Type 3: Laggards
Type 3 Instructional Designer
This work was created with the assistance of DALL·E (2023). Prompt: a sceptical stern face behind a door, one-line drawing. And should it not be free to do so?
This type makes up 16% of adopters and may be ignoring the incessant door banging, hoping this will all disappear as quickly as it emerged.
Most likely to have:
Tried and tested 1 app, likely ChatGPT
Will typically wait until use is inserted in policy or becomes standard practice across their department
Most likely to see using AI in any form or function as ‘cheating’ the process. Perhaps even experiencing some imposter syndrome thinking that “the use of AI affects credibility”.
Most likely to liken AI adoption to a ‘Video killed the Radio star’ event
What to consider:
If you are not using it, do not think that students or others will not be using it. Some have likened the AI wave to the uproar of the introduction of the calculator in the 70s. The construct of Mathematics did not change, but the method to get to the answer did. Consider your thoughts on the idea that humans have always used tools to carry out tasks – and if this is another tool or not, and what the impact may include.
At the very least test out some varied capabilities of text generation, like ChatGPT, and image generation like Midjourney.
Consider a low-stakes online scavenger hunt. This could be attempted individually or in a team. Instead of subscribing to a newsletter where the ‘What’s New’ finds you, in the scavenger hunt – you get to gauge your pace, and you go find it! Performing searches, reading an article, testing an app – consider setting aside this type of hide-and-seek day once a month, so that you stay abreast of the conversation as it relates to you. Otherwise, others might just make the decisions for you.
Lastly, do not get annoyed by your Type 1 instructional designer, their work is essential to test out the pitfalls and successes.
Conclusion
The above is more specific to instructional designers supporting academic institutions, but what similarities or differences might there be in corporate or private sectors?
Whatever your adoption rate is at present, we are currently faced with finding the right balance in our adaptability, a type of Goldilocks zone, making sure the chair sits just right. This flexibility supports us in asking solid questions but also equips us with the knowledge to identify the concerns and opportunities in our strategic processes as instructional designers. Philosopher Marshall McLuhan (1964) states that “Man becomes, as it were, the sex organs of the machine world, as a bee of the plant world”, Gen AI is dependent on how we choose to use it.
Artificial Intelligence is continually correcting its translation – improving its language model. And we have to decide whether we will learn its dialogue. Because most assuredly, it is learning ours.
