289: From Curves To Concepts: A Framework For Generative Ai As A Catalyst For Parametric Design Education

This paper presents a pedagogical framework that integrates Generative Artificial Intelligence (GenAI) into parametric architectural design education. Developed through a Design-Based Research (DBR) methodology, the framework unfolds in three iterative phases: (1) AI-assisted geometric referencing and speculative ideation; (2) visual algorithmic modeling with Grasshopper and GenAI visual interpretation; and (3) design and fabrication of context-aware pavilions for dystopic climate scenarios. Multimodal GenAI tools scaffold visual reasoning and bridge abstract ideation with rule-based logic. Rather than serving automation, GenAI functions as a cognitive scaffold fostering conceptual fluency and design agency. Results indicate that pattern-based AI outputs support reverse reasoning and enhance student engagement with computational logic, especially in multi-model workflows. Final outputs—parametric scripts, prototypes, and narrative visualizations—reflect a reflective, iterative learning process. Future developments include natural-language scripting with LLMs, ChatGPT-assisted geometric strategies, and simulation plug-ins. The study offers a replicable model for aligning GenAI experimentation with computational design pedagogy.