Nathan Moyer, 2 May 2025
I was recently able to attend the 2025 International Conference on Technology in Collegiate Mathematics (ICTCM) for the first time. Excited to hear about new teaching technologies to enhancing student learning, I soon found out that nearly all of them—no surprise—were about Artificial Intelligence. The agenda was packed with hands-on workshops on the everyday applications of AI, from generating assignments and simplifying rubrics to building classroom activities and providing personalized remediation. There were even enlightening presentations on how AI is revolutionizing math education and altering student learning outcomes. At this conference, I saw for myself how AI is revolutionizing the way we learn and teach mathematics.
Every conversation I had with some of the 250+ attendees touched on AI in a meaningful way. Views ranged from hopeful enthusiasm about its potential to wariness, concern, and even dread of the unknown. But two things were universally acknowledged: AI is here to stay, and it is already transforming higher education. Two older professors even conceded that they were thankful to be near retirement age, as they would not have to face the full impact of these innovations. Their perspective made me pause – what does this hold for those who are mid-way or just beginning their careers?
My own experience with AI in the classroom had been fairly limited before this conference. I had used ChatGPT to create humorous tales and poems of Real Analysis and experimented with NotebookLM to craft math philosophy podcasts for my students. At the time, I saw AI as nothing but a novelty—a fun classroom addition that piqued interest before we resumed “real learning” with regular lectures, practice problems, and discussion. Nevertheless, after ICTCM, I’ve realized that the applications of AI in education are far deeper than what I’d imagined. When reflecting on everything that I’d learned at ICTCM, what came to my mind was the realization that AI is not just a tool, it is a seismic shift for education. In this post, I would like to summarize some key insights as well as share some practical entry points into teaching with AI.
The Scope of AI’s Impact
Ever since ChatGPT was launched less than three years ago, the world of large language models has come a long way. AI tools are now commonplace, with the big names such as Microsoft, Google, and OpenAI offering free products. What excited me the most about ICTCM was discovering the many practical uses of integrating these AI tools into my classes. AI can make instruction both more effective and efficient, automating some tasks and performing them faster. Here are just a few examples:
- Rubrics: AI can generate detailed grading rubrics for assignments, clearly outlining expectations for different grade levels (A, B, etc.).
- Grading Aid: Imagine a class where every student gives you an Excel sheet with their solutions to some statistical problems. With AI, you can develop R code that extracts specific data from every student’s sheet and compiles it in a single scoring sheet. You can simply copy, paste, and run the code even without R expertise for quick grading.
- Letters of Recommendation: Put together a list of a student’s accomplishments, character traits, and anecdotes, and AI can produce a well-structured, tailored letter for scholarships, research assistantships, or graduate schools.
- Student Evaluations: AI can sort through all written student comments at the end of semester and summarize them into a concise report with significant highlights and important takeaways.
Of course, it’s important to note that most large language model platforms are not FERPA-compliant, so student data must always be anonymized before being processed by AI.
Beyond efficiency, AI can significantly enhance student learning.
- Assessments: Whether it’s a quiz, exam, or worksheet, AI can create relevant problems that go beyond simple practice. It’s particularly good at introducing creative variations and stimulating deeper student reflection.
- Classroom Activities: AI can create games, interactive puzzles, and other activities to support students in learning specific topics, offering scaffolding as students make their way through the content.
- Personalized Remediation: By analyzing a student’s graded exam, AI can identify areas of difficulty and generate customized remediation exercises, allowing students to focus on improving their understanding of the concepts that gave them trouble.
- Real-World Applications: AI can design exercises based on each student’s individual interests, making the content more engaging and relevant by connecting it to real-world applications.
These applications are just the beginning—as AI continues to develop, its role in education will only expand. I’ve only just begun to tap into AI’s potential in my classroom, and I’m eager to keep exploring and putting new ideas into practice.
Effective Prompting
AI-created artifacts don’t result from a simple, one-line prompt. Effective prompting takes strategy and refinement. One useful tactic is to have the AI to adopt a specific role. For example, you might say: “You are a university calculus instructor introducing your students to the chain rule for the first time…” or “You are a sophomore math major taking Linear Algebra with little proof-writing experience…” Framing requests this way makes AI generate more relevant and customized responses.
Achieving high-quality results also involves iterating on prompts to refine output. Asking of AI, “Is there more you need to know before finalizing this task?” can prompt it to request clarifications, leading to better results. Entire courses are now begin developed to study the art of training AI prompting, such as Vanderbilt University’s Prompt Engineering for ChatGPT. Like any other tool, AI requires time, practice, and experience to be used effectively. Investing in that training is vital to maximize its impact on our students.
Shifting Student Learning Outcomes
The significant shift that AI is bringing to higher education isn’t just about what tools educators have access to —it’s about how students are using those same tools. With virtually unlimited knowledge now available at their fingertips, we must ask: What is the actual value of a college degree?
Bloom’s Taxonomy is a familiar model that helps educators structure learning objectives in terms of their complexity and depth. In today’s AI-driven world, how might we employ AI tools to support and enhance learning at each level of this hierarchy? Oregon State University has produced a helpful chart for reimagining AI’s usefulness within this model. This prompts us to see these tools as ways to support or enrich human learning, rather than substitute for it.
For decades, collegiate mathematics instruction has been shifting from a focus on procedural skills to an emphasis on mathematical thinking. AI will necessarily accelerate this transformation. In the midst of a world where factoring, taking derivatives, solving equations, and memorizing formulas are increasingly automated, prioritizing these skills alone means training students to be nothing more than inferior human versions of ChatGPT.
Instead, we must broaden our definition of student learning outcomes to encompass intellectual virtues like persistence and creativity. We must foster our students’ ability to interpret, quantify, abstract, and strategize. More than that, we must instill persistence, humility, creativity, and courage—habits that make us as human and set us apart from technology.
School AI Policies
During my time at ICTCM, I was interested in seeing how different schools handle AI in coursework. Unsurprisingly, AI policies are either nonexistent or all over the place. Expectations vary by discipline, course, and even assignment. This patchwork approach makes it difficult for students to navigate AI use effectively.
At the conference, I was introduced to some helpful work done by Illinois State University’s Center for Integrated Professional Development. They have developed a structured framework that categorizes the use of AI in the classroom into seven various levels, ranging from complete prohibition to full, unrestricted use. This guide provides a standardized way of discussing and implementing AI policies between and within academic departments.
As AI continues to evolve, universities must provide more targeted guidance to both faculty and students. A well-defined framework can help ensure that AI is used ethically and efficiently, enhancing teaching and learning. However, that is still not enough. Real investment in faculty development is also essential. Instructors need to remain ahead of students in the understanding of AI capabilities and its implications for coursework. Discipline-specific training has to be strongly encouraged and supported because effective AI integration will significantly vary across academic fields.
Looking Back, Moving Forward
For me, ICTCM 2025 underscored a clear and urgent message: AI is not a passing trend—it is a transformative force redefining the rules of higher education. But this kind of disruption has happened before. In fact, the history of math education has always been a story of adaptation. The introduction of personal calculators in the 1970s revolutionized classroom pedagogy, rendering slide rules and log tables obsolete. The 1980s and 90s saw the introduction of graphing calculators, and teachers were forced to rethink assessment strategies in the face of symbolic manipulation technology. In more recent times, the prevalence of tools such as Desmos and Photomath has continued to push the limits of what it means to “do math” in a digital age. The 2025 decision to allow the built-in Desmos calculator on the AP Calculus exam signals a broader shift: we’re no longer fighting these tools—we’re integrating them.
While previous tools enhanced calculation, AI can create, analyze, and even replicate aspects of human reasoning. That revolution demands more than tweaks to policy or embrace of tools; it demands that we reexamine our pedagogical frameworks, learning goals, and professional duties. We must teach students to work with AI, but more importantly, think with it—develop the wisdom, judgment, and moral ground that our digital tools lack.
What will mathematical education be in three years? In five years? We don’t know yet—but we can be a part of building that future. Let us walk into this challenge with purpose, humility, and creativity. If we lean into this technological shift, we can enhance not only what and how we teach, but also who our students are empowered to become.