A Framework for institutional change in the age of AI
This paper proposes a new framework for institutional change in STEM higher education that addresses the unique challenges of generative AI by shifting from traditional adoption models to a flexible approach centered on six dimensions—covering tools and people—and emphasizing local inquiry, pedagogical focus, and collaborative partnership to navigate genuine uncertainty.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
The Big Picture: A New Kind of Storm
Imagine universities as ships that have spent decades learning how to navigate a specific ocean. They have excellent maps (evidence-based teaching methods) and a proven strategy: find a good map, show it to the crew (faculty), and sail together toward a known destination (better student learning). This worked great for tools like clickers or peer discussion groups.
But then, Generative AI arrived. It didn't arrive like a new map; it arrived like a sudden, massive storm that changed the ocean itself while the ships were already sailing. The authors argue that the old maps and navigation strategies no longer work because the "storm" (AI) is changing too fast, we don't fully understand it yet, and the students are already swimming in it without a guide.
The paper proposes a new framework to help universities navigate this uncertainty. Instead of trying to force a perfect solution, they suggest a more humble, flexible approach based on six key areas.
The Six Dimensions of the New Framework
The authors break down the problem into two groups: the Tools (the technology) and the People (the humans involved).
Part 1: The Tools (The "What")
1. The Evidence Base: "The Recipe vs. The Mystery Ingredient"
- Old Way: Before, if a school wanted to use a new teaching tool, they waited for scientists to test it, write a recipe, prove it worked, and then share it.
- AI Reality: AI arrived before anyone wrote the recipe. It's like a mystery ingredient dropped into the kitchen. We don't have a proven recipe yet, but the students are already eating it.
- New Strategy: Humble Inquiry. Instead of pretending we have the perfect answer, schools should admit, "We don't know everything yet." They should let teachers try small experiments locally, share what works, and build a shared understanding slowly, rather than forcing a "best practice" that might be wrong.
2. The Rate of Change: "The Shifting Sand"
- Old Way: Educational tools were like sturdy furniture. Once you bought a desk, it stayed a desk for ten years. You could buy 100 of them and be safe.
- AI Reality: AI is like a shape-shifting sandcastle. The version of the tool you use on Monday might be completely different by Friday. If you build your whole school system around a specific AI tool today, it might be obsolete next month.
- New Strategy: Focus on Approaches, Not Tools. Don't buy a specific AI platform. Instead, teach teachers how to think about learning and assessment. If the tool changes, the teaching method (the approach) should still hold up.
3. The Scope: "The Swiss Army Knife vs. The Specialized Tool"
- Old Way: Old educational tools were like specialized screwdrivers made just for school. They were designed by teachers to help students learn.
- AI Reality: AI is a giant, general-purpose Swiss Army Knife made by tech companies for the whole world. It wasn't built for schools; it just wandered in. It's used for everything from writing emails to solving physics problems, and it comes with a lot of hype and ethical baggage.
- New Strategy: Broad Conversations. Teachers can't just talk about "how to use the tool." They have to talk about ethics, privacy, and how AI fits into a student's future career. The tool is now part of the lesson, not just a helper.
Part 2: The People (The "Who")
4. Faculty Agency: "The Uninvited Guest"
- Old Way: Teachers usually chose to join reform projects. If they wanted to try a new teaching method, they raised their hand. It was voluntary.
- AI Reality: AI didn't wait for a raise of the hand. It walked into the classroom uninvited. Students are using it whether the teacher likes it or not. Many teachers feel anxious or forced to deal with it.
- New Strategy: Map the Needs. Don't treat all teachers the same. Some are terrified, some are experimenting, and some don't feel affected yet. Support needs to be tailored to where each teacher is starting from.
5. The Role of Change Agents: "The Guide vs. The Expert"
- Old Way: "Change agents" (experts who help schools reform) acted like brokers. They had a bag of proven, working tools and handed them out to teachers. "Here is the solution."
- AI Reality: The experts don't have the "solution" bag anymore because the solution doesn't exist yet.
- New Strategy: Facilitators of Inquiry. The experts should stop acting like gurus with all the answers. Instead, they should act as guides who help teachers ask good questions, run experiments, and figure things out together. They need to be experts in learning, not necessarily experts in the latest AI software.
6. The Student Role: "The Passengers vs. The Co-Pilots"
- Old Way: Students were usually the "passengers" or the "test subjects." Teachers changed the course, and students experienced the result.
- AI Reality: Students are already the co-pilots. They are using AI to learn, cheat, or create before the teachers even know it. They are the ones driving the change.
- New Strategy: Students as Partners. Schools need to ask students: "How are you using this? What do you think is fair?" Students should help design the rules and policies because they are the ones living in this new reality.
A Real-World Example: The Physics Workshop
The authors tested this framework in a physics department at the University of Colorado.
- What they did: Instead of hiring an AI expert to lecture on "How to use ChatGPT," they held workshops where teachers shared their own local problems and tried small experiments.
- How it worked: They looked at student surveys to see how students were actually using AI. They focused on teaching methods (like how to grade homework) rather than specific software. They treated the teachers as a team figuring it out together, rather than students waiting for a lecture.
The Bottom Line
The paper argues that we cannot wait for a perfect, proven plan for AI in schools because the technology moves too fast. Instead, we need to:
- Be humble and admit we are learning.
- Focus on teaching methods, not specific software.
- Treat teachers and students as partners in figuring this out, rather than just following orders from above.
It's about navigating a storm together, rather than trying to sail with a map that was drawn for a calm sea.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.