AI-Powered Instructional Design at ASU
How ASU's Collaboration with OpenAI is Reshaping the Role of Instructional Designers
Hey Folks 👋
Back in January, Arizona State University (ASU) - one of the world’s most innovative universities - announced a new partnership with OpenAI.
As part of the announcement, ASU invited faculty, staff and students to propose ways to integrate ChatGPT on campus. This week - eight months and over 400 submissions later - ASU released an update on what has happened since.
The developments and experiments at ASU provide a fascinating window into two things:
How the world is reimagining learning in the age of AI;
How the role of the instructional designer is changing in the age of AI.
In this week’s blog post, I’ll provide a summary of how faculty, staff and students at ASU are starting to reimagine education in the age of AI, and explore what this means for the instructions designers who work there.
Let’s go!
ASU’s “Crowdsourcing” Approach to AI
As ASU’s Chief Innovation Officer, Lev Gonick, said this week:
“Everything we do at ASU is focused in on impact. If you want to focus in on impact the impact of the technology start by actually asking the community what they want to solve for.”
In response to its call to action back in January, ASU received 400 proposals for AI use cases from across 80% of its colleges, showing widespread interest and engagement with AI across a variety of disciplines.
These proposals fall into three categories:
AI use cases to support teaching and learning
AI use cases to advance research for public good
AI use cases to enhance the future of work
Since then, ASU have turned 50% of the submitted ideas (200 proposals) into active, projects. Let’s take a look at what they’re doing.
ASU’s AI Use Cases
Let's delve into some specific projects that showcase the varied use cases of AI at ASU:
AI as a Writing Companion
Students often struggle with scholarly writing and need immediate feedback. This project uses ChatGPT to provide real-time feedback to help students to improve their writing skills more quickly and efficiently than traditional methods.
Virtual Patient Interactions
Health students need practice in patient-provider interactions but opportunities for real-world practice are limited. A chatbot named 'Sam,' powered by ChatGPT, helps students in the College of Health Solutions practice patient-provider interactions through role-based conversations. The primary purpose of this tool is to allow students to authentically practice their motivational skills, but it also drives teacher efficacy by providing transcripts for faculty for easier grading.
Research Participant Recruitment
Recruiting participants to take part in research project often involves complex, jargon-heavy communication that can deter potential participants. Academics at ASU are exploring how ChatGPT can be used to support effective and ethical participant recruitment for research studies. The goal is to use AI to ensure that communication with potential participants is comprehensible and highlights opportunities without excessive scientific jargon.
AI Tutor for Psychology
Psychology students expressed a need for more personalised support outside of class hours. This project creates a virtual study buddy to help psychology students explore topics in depth, ask questions and test their understanding.
Immersive Newsroom Project
Journalism education needs to keep pace with rapidly evolving technology in news production. The Walter Cronkite School of Journalism and Mass Communication is exploring AI's potential in news production, preparing students for the future of journalism. By exploring AI's potential in news production, this project prepares students for the future of journalism, giving them hands-on experience with cutting-edge tools they'll likely encounter in their careers.
Language Buddy
Traditional language learning methods often lack personalisation and accessibility. An AI-powered language tutor which provides personalised, accessible language learning experiences. This AI-powered tutor can adapt on the fly to individual learning needs, with the goal of making language acquisition more efficient and effective.
Three key themes I noticed in ASU's prioritised projects:
Student-focused: Most projects are designed to directly benefit student learning.
Personalisation: Many projects are designed to offer personalised learning experiences, typically through the provision of 1:1 support using AI.
Real-world skill development: Several projects focus on practical skills needed in professional settings.
Implications for Instructional Design at ASU
As ever, the question top of my mind in all of this is: what does all of this mean for instructional designers?
How does the introduction of AI into the process of teaching and learning impact the day-to-day work of Instructional Designers (IDs) at ASU? And what does this tell us about the “post AI” ID role?
Based on the experiments that are happening at ASU, I put together a list of the impact on the role and skills of instructional designers. This provides an insight into the day to day work of IDs at ASU right now and a reliable glimpse into the very-near “post AI” future of Instructional Design.
Let’s look at how ID is changing through the lens of the ADDIE process:
Analysis:
Before: IDs conducted in-person interviews with Subject Matter Experts (SMEs) to gather content and understand learning needs.
After: IDs use AI-assisted knowledge extraction tools to analyse SME-created content, conversations, and even published works before they collaborate. This process is more efficient and often uncovers connections that might have been missed in traditional interviews.
Practical example: For a new course on emerging technologies, IDs use AI to analyse the SME's published papers, recorded lectures, and social media posts to create an initial knowledge map. The ID and SME then refine this map together.
Design:
Before: IDs created linear course outlines with limited personalisation options.
After: IDs design adaptive learning paths and frameworks for AI to generate personalised study materials, quizzes, and discussion prompts.
Practical example: In the AI tutor for psychology project, IDs design a system where AI analyses a student's performance, learning patterns and goals to create and continuously adjust a personalised study plan to hit a goal defined by the tutor.
Development:
Before: IDs manually created all course content and assessments.
After: IDs use AI tools for initial content generation and curation, focusing on higher-order learning activities and creating guidelines for AI-generated content.
Practical example: For the writing companion project, IDs design rubrics and guidelines for AI to provide real-time, personalised feedback on student writing.
Implementation:
Before: IDs focused on training faculty to use the Learning Management System and troubleshooting technical issues.
After: IDs oversee the integration of AI tools, ensure they align with learning objectives, and train faculty on effective use of these tools.
Practical example: In the virtual patient interactions project, IDs help implement the 'Sam' chatbot, designing realistic scenarios and training the AI to respond appropriately to student inputs, with input from the SME.
Evaluation:
Before: IDs relied on end-of-course surveys and basic completion rate analytics.
After: IDs use advanced learning analytics for continuous, real-time evaluation, allowing for dynamic course improvements.
Practical example: IDs design dashboards that aggregate data from various AI-powered learning tools, providing insights into student engagement, performance trends, and areas needing additional support.
Emerging Key Skills for Instructional Designers
As I discussed in last week’s post, as AI reshapes education, it also reshapes instructional design. As a result, several new skills are becoming crucial:
AI Literacy
We are seeing a shift from IDs needing basic digital literacy skills to them requiring more in-depth understanding of AI capabilities and limitations.
In practice, IDs need to understand different types of AI (e.g., natural language processing, machine learning), their applications in education, and how to effectively integrate them into course design.
Data Analysis
We are also seeing an increasing need for IDs to be able to interpret complex, real-time data. IDs must be able to interpret AI-generated learning analytics, identify meaningful patterns, and use these insights to inform course design and student support strategies.
Modular Learning & Logic Design
IDs increasingly need to think about instructional design as multi-dimensional, shifting from the creation of static course structures to the design of dynamic, personalised learning paths.
In practice, this will mean designing modular content and “logic” for decision trees that allow AI to assemble personalised learning experiences based on individual student needs and progress.
Interdisciplinary Focus & Facilitation
In a post-AI world, IDs will shift from discipline-specific content to more cross-disciplinary design focused on developing key skills like creative thinking, research skills and problem-solving. In practice, IDs use AI-identified connections to facilitate collaboration between SMEs from different fields, creating more competency-based learning experiences which asses the ability to apply skills rather than recall knowledge.
Prompt Engineering
At least in the immediate term, IDs need to hone their prompt engineering skills to ensure that they are able to guide AI in generating relevant, accurate, and pedagogically sound content and assessments. Check out this post on how to 10X your prompting skills in 10 mins.
Ethical AI Use and Governance
We are also seeing a shift from a focus on general academic integrity to AI-specific ethical considerations. In practice, IDs need to understand and address issues like AI bias, data privacy, and responsible AI use. They must design courses that not only use AI ethically but also teach students about these important considerations.
Closing Thoughts
As more and more “content to course” AI tools attempt to automate instructional design and replace instructional designers, ASU's pioneering work is a welcome reminder of an alternative narrative: one where AI will play a key part in elevating the significance of the Instructional Designer, shifting their focus from functional “output” tasks to more strategic and higher-order tasks.
Of course, just as Canva made graphic design skills available to those who couldn’t afford a graphic designer, “content to course” AI tools will open up access to standardised and automated instructional design skills. But just as Canva did not “kill” graphic design, these tools will not “kill” instructional design.
ASU’s experiments provide a glimpse into an alternative and very real post-AI world - one which positions IDs in a more critical role than ever in the design, development and evaluation of post-AI education.
As I have concluded before and will definitely conclude again: the implications of AI for our skills & day-to-day work are significant but positive, elevating rather eliminating the role of the ID.
Happy innovating!
Phil 👋
P.S. If you want to get hands-on and experiment with how to integrate AI into your instructional design process, check out my AI Learning Design Bootcamp.