A Human-AI Hybrid Co-Design Model: Planning Effective Instruction with ChatGPT

Abstract

This article explores the integration of ChatGPT into instructional design, utilizing a backward planning framework to align learning objectives, assessments, and activities for enhanced educational outcomes. Grounded in the principles of Understanding by Design (UbD) the paper introduces a three-step human-AI hybrid co-design model that enables educators to streamline planning processes while maintaining pedagogical rigor. This manuscript demonstrates how generative AI can support the creation of transfer goals, performance tasks, and authentic learning experiences, providing detailed prompts and examples applicable to both K-12 and higher education settings. Findings from five preliminary evaluation studies with over 500 participants highlight the model’s adaptability across diverse educational contexts, with teachers reporting improved alignment, creativity, and efficiency in lesson planning. Participants also emphasized the importance of iterative feedback and foundational AI knowledge for maximizing the benefits of the co-design process. The model empowers teachers to implement evidence-based practices by focusing on alignment and real-world relevance, fostering meaningful student engagement and knowledge transfer. Additionally, the article addresses potential limitations, such as the need for foundational AI literacy and instructional planning skills, and proposes strategies for effective application across diverse educational contexts. This work contributes to the growing field of AI-assisted pedagogy by offering a practical guide that balances the scalability of AI with the critical insights of human expertise.