Teaching in the Age of AI: What I Learned from the Governor’s Teaching Fellows Program 

I went into the Governor’s Teaching Fellows program expecting to pick up a few new teaching ideas, learn about some useful tools and come back with things I could try in my own classes.

I did get all of that, but I also came away thinking much more deeply about what teaching will look like as AI becomes a normal part of both education and the workplace.

One of the big questions we kept coming back to during the May 2026 symposium was pretty simple: How can AI actually make teaching better and help prepare students for what comes next? We also talked about whether it makes more sense to think of AI as “augmented intelligence” instead of artificial intelligence — something that can support human thinking rather than replace it.

That idea stuck with me.

A lot of the conversation around AI in higher education has focused on cheating and how to stop students from using it. The program pushed us to think about the issue a little differently. Instead of only asking, “How do we keep students from using AI?” we were encouraged to ask, “What do students need to know in order to use AI well?”

That does not mean we should just accept everything AI produces. Actually, some of the most useful parts of the program focused on the problems with AI, including hallucinations, bias, privacy concerns, academic integrity, overreliance and the limits of AI detection tools.

One of the takeaways was that simply banning AI probably is not going to solve much in the long run. AI detectors are not always reliable, and blanket bans do not necessarily stop students from using the technology. A better approach is to be clear about expectations, talk openly about academic integrity and help students understand when AI can be useful and when it can get in the way of real learning.

One idea I especially liked was the difference between an AI-resistant assignment and an AI-resilient assignment.

An AI-resistant assignment is designed to limit or prevent the use of AI. Think in-person exams, oral presentations or proctored work. There are definitely times when that makes sense.

An AI-resilient assignment is a little different. It is designed so that student thinking is still visible and important even if AI is available.

That really changed how I thought about assignment design.

Instead of trying to create an assignment that AI cannot touch, maybe the better question is whether the assignment still requires the student to think. Can they explain their reasoning? Can they critique an AI response? Can they compare alternatives, make a decision, defend a position or reflect on what they learned?

The goal is for AI to be one part of the process, not a shortcut around the process.

We also spent time looking at actual AI tools and classroom applications. We talked about tools for research, literature reviews, course design, presentations, data analysis and instructional support. We saw examples of NotebookLM being used with course materials and custom chatbots being used for role-playing and stakeholder exercises.

What I liked about those examples was that the AI itself was not really the point. The learning came from what students did with the AI output afterward.

They had to question it. Evaluate it. Decide whether it was accurate. Think about what it missed. That puts the student back in the center of the learning process.

That leads to probably my biggest takeaway from the whole experience: the human still has to stay in the loop.

Students still need knowledge. They still need judgment. They still need to know when something does not sound right. They need to recognize weak evidence, bad assumptions and questionable conclusions.

In some ways, AI may actually make those skills more important.

The same thing is true for faculty. AI can help us brainstorm, design assignments, create examples, organize information, analyze data and save time on certain tasks. But the output still has to be reviewed and checked. Good use of AI requires context, clear instructions, follow-up questions and human judgment.

I left the Governor’s Teaching Fellows program more interested in AI, but also more cautious about it. I do not think those two things conflict.

AI has a lot of potential in higher education, but using it well is going to require faculty who are willing to experiment while still keeping student learning at the center of what they do.

That was probably the biggest value of the program for me. It gave me time to think about teaching in a different way, hear from faculty outside my own discipline and consider what students are going to need from us in the years ahead.