.png)
Todd Zakrajsek, University of North Carolina at Chapel Hill
Keywords: Generative AI, Learning, Teaching
Key Statement: Amid widespread nervousness about generative AI, we can draw on self-efficacy to establish a foundation from which to move forward in creating new learning opportunities for our students with this groundbreaking tool.
Introduction
I’ll be honest—generative artificial intelligence (genAI) makes me a little uneasy. It’s something I’ve been thinking about a lot lately, probably like pretty much everyone in higher education (and beyond). Nervous or not, it is imperative to recognize that education is undergoing a profound transformation right now. In the past month alone, I’ve heard numerous people say, “AI isn’t going anywhere,” or “It looks like AI is here to stay.” Agreed.
I’ve been teaching for nearly 40 years. Across that time, I have continuously worked to come up with new ways to engage learners. For the past 30 years, I’ve been running faculty and educational development programs, always adapting to new trends and helping educators create better learning experiences for their students. I’ve authored or coauthored six books over the last four years, all written without the assistance of AI. Experimenting with new approaches isn’t foreign to me when it comes to keeping up with cutting-edge research and using evidence-based teaching practices. That said, I believe it’s perfectly natural to feel some ambivalence or even outright nervousness about genAI. What effects will it have on our lives, on our jobs? Although that question is unanswerable right now, you have agency. You have self-efficacy.
Our Selves
Changing anything, like the impact of genAI on our jobs, is highly dependent on our individual levels of self-efficacy. Self-efficacy is a person’s belief in their ability to achieve a goal or complete a task and is extremely important in education, both for students and faculty. How do we establish our self-efficacy? Bandura (1977) identified four sources. The first is personal success—when we accomplish something, we’re more likely to believe we can do similar things in the future. The second is vicarious experience—seeing someone else succeed can bolster our own confidence. The third source is verbal encouragement—if I’m watching my daughter play soccer and yell, “You’ve got this!” it is likely to boost her confidence. Finally, self-efficacy can come from how we interpret our physical responses. For example, if you’re nervous before a big presentation and a friend says, “I bet you’re excited to give this talk,” you might start seeing your nerves as a sign of energy, rather than fear, which can raise your self-efficacy.
How does self-efficacy relate to genAI? Recently, I was the closing keynote speaker for a conference and Ethan Mollick (2024), author of Co-Intelligence: Living and Working With AI was the opening keynote speaker. He showed us what AI can do—and it was a lot to process, from excitement to terror. As he presented the latest genAI possibilities, I found myself feeling uneasy, even questioning what my professional work could possibly look like going forward. How do I teach when students can use genAI to generate answers faster than I can explain a concept? If I continue to write books, will anyone bother to read them? Why would anyone read my explanation of metacognition when they could just have a “personal” chat with ChatGPT?
And then, in a flash, my perspective shifted. Without even realizing it, I was reinterpreting my reaction (the fourth source noted above relative to self-efficacy). I thought, “Wow, this is a fascinating time to be a faculty member. The last time the field of education shifted this dramatically was when we transitioned from focusing on teaching to focusing on learning (though the current shift is happening much faster).” Yes, I still had some anxiety, but it was mixed with the excitement of creating something new. My self-efficacy got a bit of a boost. Watching Mollick navigate AI so skillfully made me think, “With a bit of work, I can do this, or an approximation of it.” Before the day was over, I used my laptop to create an image based on what I’d just learned from Mollick’s presentation. That small success raised my self-efficacy just a bit more.
Our Concerns
Of course, there are serious concerns we need to address with genAI. Issues like academic integrity and plagiarism have already pushed us to rethink assessment strategies and the concept of a learning community. Equity and access have long been issues in education, and with the rapid pace of technological change, those without access to genAI tools, or the resources to use them, will fall even further behind (OpenAI, personal communication, October 18, 2024). And there’s also the risk that overreliance on AI could erode students’ ability to build foundational knowledge, which is crucial for critical thinking (Zakrajsek, 2025). (Imagine looking up half the words in this text just because you didn’t think it was necessary to memorize them.) Of course, genAI is also creating a significant negative impact on our environment that deserves serious attention.
Our Opportunities
With respect to education, if we focus only on what genAI can’t do or how it could be used to bypass learning, we’ll miss a huge opportunity (Alasadi & Biaz, 2023). Similar concerns have been aired about calculators, computers, the internet, and even books. Now, it’s AI’s turn. Yes, there are things to be concerned about, but we can—and should—push in another direction. We, as educators, are the ones who come up with solutions to pressing social issues. We conduct research that informs a vast network of intelligent, professional individuals. We’re in higher education for a reason, and it’s not just to deliver content. We’re here to guide learners, using the best tools we have.
If you’re already using genAI, take the opportunity to help your colleagues, especially those feeling paralyzed by fear or indifference, to move forward. Show them how genAI can create personalized learning experiences or how students can use ChatGPT to brainstorm ideas for papers. Start small if needed. When others see you rethinking aspects of education with genAI, they’ll realize they can do it, too. Their growth will accelerate as they experiment with new teaching concepts and find success in areas that have long been challenging. Imagine asking an open-ended question in class about a topic of discussion, having students input their answers into a short Google Form that connects to ChatGPT, and having AI analyze the responses and generate key themes within seconds. When else have we been able to gather qualitative feedback from 400 students in class and share the results just seconds later?
Sure, there will be moments when you or your colleagues hit snags or doubt the future. That’s when encouragement is needed. We know verbal support can boost self-efficacy: We can make sure to maintain our professional relationships on a human level, or set up small work groups to support one another.
Our Futures
Yes, I’m nervous about what genAI means for the educational system and how I will best use the decades of professional knowledge I’ve accumulated. But that nervousness isn’t going to stop me from rethinking how I can help learners grow more quickly and comprehensively than ever before. We’re in the midst of an educational revolution—I think not in what it means to be an educator, but certainly what it means with respect to how we teach.
This is an exciting time. I’ve tried a few things that have pushed my concept of teaching in new directions. Every day, my genAI self-efficacy and skill grows. Just a few years ago, I was seriously considering retirement, but I’m now so glad I didn’t take that step. I wouldn’t have wanted to miss this transformation.
It’s scary, sure. But that scariness is also mixed with excitement. You’ve got this.
Discussion Questions:
What level of self-efficacy do you have with respect to teaching with genAI? How might you expect your self-efficacy to change with future use?
From your perspective at your institution, what are the top three potential risks and top three rewards of using genAI in higher education, both for students and faculty?
In what ways can faculty actively support each other in embracing genAI as a tool for enhancing learning? How might this collaboration transform more traditional teaching methods?
References
Alasadi, E.A., & Biaz, C.R. (2023). Generative AI in education and research: Opportunities, concerns, and solutions. Journal of Chemical Education, 100(8), 2965–2971.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191–215.
Mollick, E. (2024). Co-intelligence: Living and working with AI. Penguin.
Zakrajsek, T. (2025). Essentials of the new science of learning: The power of learning in harmony with your brain. Routledge.
About the Author
GenAI Disclosure Statement: I used ChatGPT-4 to generate the three common concerns about genAI in higher education and also to provide a suggestion for rewriting a paragraph that wasn’t quite coming together. With ChatGPT’s input, I revised the paragraph again—better than my initial attempt. I also used ChatGPT to draft the discussion questions. Of the three items suggested by ChatGPT, one I liked and made only a minor change, one I grinned at and tossed, and one gave me an idea for a question that I then wrote all on my own.