Infusing AI into one of my classes
A fall semester experiment
How might university instructors include new AI technologies in our classes?
This semester I’m teaching three seminars for Georgetown University’s Learning, Design, and Technology (LDT) program. I wanted to write about one of those classes here, because it’s an experiment in infusing AI into a preexisting course.
The class is on the future of higher education, a topic dear to my heart. I designed it from scratch and have developed it over several years. Students get to study how futuring works, to think about the future of higher ed, and to learn more about researching in an interdisciplinary way.
This year I added AI, as I think it potentially plays a major role in higher education’s future. I did so with a few principles in mind:
To mix practical, hands-on work with critical study and discussion.
To align AI exercises with non-AI syllabus content. i.e., a week on scenario creation will lead us to using AI to generate scenario content.
To meet students where they are at the start of the class, and to help them gradually accumulate knowledge as we go.
Combining instruction (learning how to use and think about the tools) with getting meta- about them (how did we use them in this class? how can we apply this elsewhere?).
There are some challenges involved in doing so. I chose to rely on web-hosted services, rather than open source tools, because I didn’t have the time to prepare the latter and because learning to use such applications would have taken up a significant chunk of class time. Therefore we’re bound by whatever these web services choose to do: adding or substracting guardrails, charging new fees, changing interfaces, or going offline.
Other things might change over the next four months which could alter the class design. The university where I’m teaching might issue policies or make enterprise decisions, anything from restricting access to certain platforms to rendering access to new ones. And the cultural politics around AI can shift, which influence the world my students inhabit as well as their attitudes.
Speaking of which, I’m prepared for the possibility that some students might not want to use specific platforms, categories of AI, or generative AI at all. If one objects to (say) Google politically and won’t use Bard, we have Bing and ChatGPT on desk. I’m ready to allow students to opt out of a category, allowing them to use another (i.e., images instead of text), if they make such a case. If they don’t want to use AI at all, they don’t have to, but should observe the practice going on around them. And naturally any such arguments fit into our overall discussion about the tech.
Back to the syllabus: I’m hoping to build up students’ AI capacity over the weeks and months we work together. I’m not sure I scaffolded this correctly, though, as each exercise depends on the non-AI content of each session. So I’ll have to keep tabs on how each student develops over time and be ready to alter class plans as we go.
I have not written policies about academic integrity for this class. Instead, I hope the students and I can develop practices and guidance as we go.
Here’s the syllabus. You can see the AI exercises marked out at quotations, amidst the syllabus context:
Thursday, August 24, 2023 - Introductions
Designing the class: technologies, community, practices, pathways
Forecasting methods: introduction to futuring
Exercise: introduction Canvas thread
Stewart Brand, “Pace Layering: How Complex Systems Learn and Keep Learning”
AI exercise: introduction to tools
Thursday, August 31, 2023 - Futures and Systems
Gidley, The Future: A Very Short Introduction, chapters 1-4
How To Run a College, chapters 1-6
Exercise: assemble academic systems
AI exercise: how AI imagines higher education
Thursday, September 7, 2023 - Systems and scanning
Forecasting methods: horizon scanning, higher education
How To Run a College, chapters 7-9
Exercise: set up digital scanning practice
AI exercise: how can AI help with horizon scanning?
Thursday, September 14, 2023 - Trend Analysis
Exercise: horizon scanning
Forecasting methods: trend analysis, STEEP
Exercise: turn this week's horizon scanning into trends
Academia Next, chapters 1-6
AI exercise: turn horizon scan results into trends
Thursday, September 21, 2023 - Scenarios
Exercise: horizon scanning
Academia Next, chapters 7-15
Forecasting methods: scenarios
AI exercise: use AI to create scenario text and images (some extra thoughts here)
Thursday, September 28, 2023 - Speculative Fiction as Futures Tool
Exercise: horizon scanning
Forecasting methods: fiction
Stories:
Hernan Ortiz, “The Punishment Room”
Padgett, "Mimsy Were The Borogoves."
Suzette Haden Elgin, "For The Sake Of Grace.”
Saxey, “Not Smart, Not Clever”
Wagner, "University, Speaking"
AI exercise: ask AIs to create future stories
Friday, September 29, 2022 - TRENDS ANALYSIS DUE
Thursday, October 5, 2023 - Delphi Method: Educational Technology
Exercise: horizon scanning
Forecasting methods: the Delphi Method
AI exercise: explore for Delphi process
Thursday, October 12, 2023 - Educational Technology
Exercise: horizon scanning
Readings: selected by students
(items here)
AI exercise: TBD
Thursday, October 19, 2023 - Gaming the Future
Forecasting methods: simulation gaming
Exercise: horizon scanning
Readings:
Alexander, “A Web Game for Predicting Some Futures: Exploring the Wisdom of Crowds”
Practice games
Matrix University game materials (link tk)
Thursday, October 26, 2023 - Student Determined Topics
Exercise: horizon scanning (share notes here)
Matrix game, 2
AI exercise: getting an AI to manage a simulation game (some thoughts here)
Friday, October 27, 2022 - STRATEGY MEMO DUE
Thursday, November 2, 2023 - Decolonizing the University
Exercise: horizon scanning (share notes here)
la paperson, A Third University Is Possible
AI exercise: to what extent does the technology reflect colonialism?
Thursday, November 9, 2023 - Higher Education and the Climate Crisis, I
Exercise: horizon scanning (share notes here)
Universities on Fire (to 111)
AI exercise: use text and image generators to imagine academic futures in the Anthropocene
Thursday, November 16, 2023 - Higher Education and the Climate Crisis, II
Exercise: horizon scanning (share notes here)
Universities on Fire (115-to end)
AI exercise: consider AI's academic role in the climate crisis
(no class November 23, 2022 - Fall Recess)
Thursday, November 30, 2023 - Student Futures
Presentations
AI exercise: envisioning the rest of the century
Tuesday, December 12, 2023 - FINAL PROJECT DUE
CLASS READINGS
Bryan Alexander, Academia Next.
____, Universities on Fire.
Jennifer Gidley, The Future: A Very Short Introduction.
Brian C. Mitchell and W. Joseph King, How to Run a College A Practical Guide for Trustees, Faculty, Administrators, and Policymakers.
la paperson, A Third University Is Possible.
That’s all for now. I have another class to finish preparing, one on technology and innovation, and I’m also including AI work there. I’ll share that soon.
I love your syllabus Bryan. What a great way to introduce, use, and challenge students with evolving AI trends and issue!
This looks like a wonderful class, Bryan! Your students are in for a fascinating semester.
One thing I particularly appreciate about your approach is your incorporation of the "meta." This seems hard to avoid, given that you're focusing on a topic that may itself have a direct impact on the course during the semester, and I think it's so smart to embrace this uncertainty as part of the challenge and excitement of futuring.
I've also been thinking a lot about LLMs require meta-level thinking in the way both we and LLMs use language as the medium to talk, write, and think about the way LLMs talk, write, and, well, not quite think (yet). This is one way I see the popular calculator analogy falling short.
I hope the course is a great learning experience for both your students and you!