生成式人工智慧角色扮演於輔助設計思考實務訓練之應用模式

Applying a Generative Artificial Intelligence-based Roleplaying Learning Model to Support Design Thinking Practices

設計思考實務訓練強調實踐以人為本的思考方式，但一般設計課程很難找到真實的目標對象，配合授課時程實務練習。因此，本研究應用生成式人工智慧角色扮演，輔助設計思考實務訓練，透過生成式人工智慧進行角色扮演，讓學生有機會可練習訪談技巧，深入探索並了解目標客群。本研究透過質化分析，探究學生與生成式人工智慧的互動模式，並分析其學習歷程成果。研究結果發現，應用生成式人工智慧角色扮演可幫助學生探究問題，促進創意發想及評估設計方案；然而，在互動過程中亦發現一些潛藏的問題，例如生成式人工智慧角色易發生混淆、重像人物之謬誤等。建議未來在設計思考相關課程中，教師宜引導學生，進行多次修正與調整對話，方能得到較理想的回饋與建議。

Design thinking practice emphasizes a human-centered approach. However, recruiting real target customers that align with the course schedule for practical exercises is difficult in traditional design courses. Therefore, this study applied generative artificial intelligence (GAI) role-playing to assist design thinking practice. By role-playing practices with GAI, students could practice interview skills and explore their understanding of target customers. Qualitative analysis was used to examine the student interactions with GAI and their learning outcomes. The results show that GAI role-playing can help students explore problems, foster creative ideas, and evaluate design solutions. However, some latent issues during the interaction process were noted, such as the confused roles of GAI and Persona twins. For future design thinking-related courses, it is recommended that instructors guide students through multiple revisions and adjustments in dialogues to obtain ideal feedback and suggestions from GAI.