Discussing Early Efforts Integrating ChatGPT into Mathematics Teacher Preparation Courses

Michele Cudd (Morehead State University)

This paper shares examples of initial efforts at integrating ChatGPT into undergraduate mathematics teacher preparation courses. Examples focused on lesson design experiences include building from one specific ChatGPT-generated lesson plan and a comparative analysis of four different lesson plans. Other examples involve real-time prompts such as consulting for examples of specific types of goals and refining assessment questions. The affordances of these experiences include the option of modifying tailored examples, although caution should still be heeded as preservice teachers may not yet have the knowledge base to be critical consumers of ChatGPT-generated products.

Ethical Reflection on AI in the Development of a Teacher Dashboard

James Drimalla (University of Virginia)

Developers of AI tools are responsible for building trust with teacher educators and teachers. Thus, mathematics teacher educators (MTEs) and mathematics education researchers contributing to the development of AI tools need to explicate the ethical considerations of their work for other MTEs and mathematics teachers. In this manuscript, I describe (a) an AI-powered teacher dashboard in development and (b) the ethical considerations for creating this tool. I outline the values that guided the design of our tool and our investigation of potential bias. The manuscript is a move towards an honest conversation among MTEs about the potential benefits and limitations of AI tools.

Black Feminist Thought as a Guide for Ethical Integration of Artificial Intelligence in Mathematics Classroom

Eunhye Flavin and Matthew T. Flavin (Georgia Institute of Technology)

The advancement of artificial intelligence (AI) underscores the need for its ethical integration into the mathematics classroom. Despite the potential benefits, AI tools present new challenges in how they perpetuate racialized and gender-based violence. Given how U.S. mathematics classrooms reproduce systemic effects of racism and patriarchy, we argue that discussions must extend beyond the limitations of AI alone. Rather, we should explicitly address the power dynamics at play. Drawing on Collins and Bilge’s matrix of domination, this study identifies harms that AI tools can enact in mathematics classrooms with respect to structural, cultural, disciplinary, and interpersonal domains of power

The "Task" at Hand: LLMs and Low-Floor, High-Ceiling Math Tasks

Anna Gustaveson and Margaret Ann Donnelly (University of North Carolina)

Large Language Models (LLMs) have potential for supporting creation of low-floor, high-ceiling math tasks. We asked ChatGPT-3.5 and Gemini to do just that for adding fractions with unlike denominators. The tasks produced by these LLMs interpreted low-floor, high-ceiling as separate tasks or questions rather than elements of one rich task, and task components varied in the degree of potential for student thinking and strategy selection. Additionally, mathematical and language errors were present. This points to a need for Teacher Educators to support teachers in prompting, evaluating, and implementing tasks created by LLMs and open-access tasks more generally.

Scaffolding a Critical Lens of Generative AI for Lesson Planning

Alesia Mickle Moldavan and Bailey Nafziger (Georgia Southern University)

This article reports on two teacher educators supporting preservice teachers using an AI tool (MagicSchoolAI) in their elementary mathematics and science methods courses to guide lesson planning. Preservice teachers used the AI tool to learn more about content standards and students’ misconceptions. They also assessed the AI-generated output with critical questions and justification from outside, reputable sources. Exposing preservice teachers to AI in their methods courses allowed them to develop their understanding of AI tools and their criticality of AI-generated output. Lessons learned are shared to guide considerations for ethical implications and recommendations for use in teacher education.

The Professional Knowledge Required for High-Quality AI-Generated Mathematics Lesson Planning

Drew Nucci (WestEd), Alex Liu (Univ. of Washington), Min Sun (Univ. of Washington), and Lorraine M. Males (Univ. of Nebraska)

What do mathematics teachers need to know to use generative AI tools for lesson planning? In this brief, a team of mathematics teacher educators and AI designers argue for the importance of human professional knowledge in the design and use of generative AI tools. The same professional knowledge required for ambitious mathematics teaching was used in the development of a pilot AI lesson planning system. We describe our design inputs and provide guidance for iterative processes that support mathematics teachers’ use of generative AI for planning ambitious mathematics instruction.

Teacher Perspectives on Harnessing AI in Mathematics Classrooms

Byungeun Pak (Utah Tech University)

Mathematics teacher educators (MTEs) around the world are expected to address the emerging need and potential for teachers to use AI in mathematics instruction. To offer insights into the US mathematics teacher education, the authors share the result of an anonymous survey targeting 131 teachers in South Korea to gauge their perspectives on the benefits and challenges of the use of AI in their mathematics instruction. Through findings, the authors highlighted some benefits (e.g., personalized instruction and assessment) and challenges (e.g., implementation and students’ thinking skills). They offer implications for MTEs to support teachers in integrating AI into mathematics instruction.

Developing Secondary Mathematics Prospective Teachers' Confidence in Using Programming in Mathematics Classrooms: Benefits of the Use of Large Language Models

Hyejin Park and Eric Manley (Drake University)

In this paper, we share our work with secondary prospective teachers (PSTs) to support their learning in integrating programming in mathematics classrooms. Our study shows the potential use of ChatGPT as an assistant tool to generate computer code, which could contribute to increasing PSTs’ openness to and confidence in incorporating programming into teaching mathematics. However, to support teachers’ judicious integration of ChatGPT-generated code in mathematics lessons, it is necessary to provide learning opportunities through which teachers can experience interpreting, revising, designing, and evaluating programs in mathematics learning contexts so they can build knowledge for teaching mathematics with programming

Exploring the Potential of ChatGPT as Codesigners for Culturally Relevant and Inquiry-Based Mathematics Tasks

Jennifer Suh, Maureen Vora, and Deanna Moreau (George Mason University)

This study examines the role of ChatGPT as a co-designer for culturally relevant and inquiry-based math tasks. With the increase in culturally and linguistically diverse student populations in the U.S., teachers require support to create lessons that resonate with varied student lived experiences.  The research engaged twenty-one mathematics teacher leaders, who utilized ChatGPT to design culturally responsive lessons.  Findings suggest that while ChatGPT generates many cultural contexts, it often lacks depth, risking cultural appropriation or stereotype reinforcement.  The effectiveness of ChatGPT as a tool hinges on the teacher's knowledge to adapt AI suggestions to specific classroom standards and the diversity of students. Thus educators must critically engage with AI, applying their expertise to refine and align generated content with pedagogical objectives, cultural competence, and ethical considerations. 

Modeling AI as a Thought Partner for Mathematics Instruction

Linda Venenciano and Karren Timmermans (Pacific University)

In this article we focus on the growing role of Artificial Intelligence (AI) in K-12 education, particularly in our teacher candidates' planning and teaching of math concepts and their students’ learning experiences. We address the importance of informing teachers about technological trends, specifically on the integration of generative AI tools as thought partners in their teaching and learning. Although AI cannot replace the unique qualities of a classroom teacher, our goal is to ensure teacher candidates develop the skills needed to effectively utilize AI within a math context.