生成式AI融入資優生程式教育之課程實踐與省思

Integrating Generative AI into Gifted Students’Programming Education:A Practical Exploration and Reflection

本教學案例係以生成式AI為輔助工具，結合聯合國永續發展目標（SDGs）與STEAM教育精神，設計並實踐一門具跨領域特色之專題課程，課程中導入Arduino Uno進行程式控制實作，旨在引導國小資優學生從圖形式程式介面逐步邁向文字編程的自主學習歷程。課程設計以「枝葉園圃」為主題，結合實際生活問題，引導學生從感測器應用、資料判讀、條件設計與邏輯控制等層面，進行跨域統整與創意思考。本教學案例實施後觀察到生成式AI在課程中扮演多重正向角色，不僅協助學生釐清語法結構、除錯程式錯誤，亦可激發其發想深度與反思能力，展現出作為能力彌補者、創思發想促進者與學習夥伴與輔助者的角色，並能支持資優班教師進行個別化教學與差異化引導。
在課程實施歷程中亦觀察到AI應用所帶來的潛在隱憂，例如學生過度依賴AI而導致思考歷程被壓縮與轉移的「大腦外包」現象，以及因生成錯誤內容但學生難以辨識正確性的「AI幻覺」問題，特別在學生尚未具備足夠程式基礎時更為明顯。
綜合而言，生成式AI於國小資優學生程式教育中展現高度輔助潛力，不僅提升學生的學習效率與動機，也擴展其解決問題與創造思考的深度。然為確保其正向效益，未來課程設計應更加重視教師引導策略，強化學生提問能力、判讀與驗證素養，並設計具挑戰性、彈性與反思性的任務情境，使生成式AI成為促進學生知識建構與自主思維的延伸工具，而非取代其學習主體性的外部資源。

This teaching case integrates generative AI as an assistive tool, combined with the United Nations Sustainable Development Goals (SDGs) and the educational philosophy of STEAM, to design and implement an interdisciplinary project-based curriculum. The course incorporates Arduino Uno for programming-based control tasks, aiming to guide gifted elementary school students in transitioning from graphical interfaces to text-based programming through autonomous learning. Centered on the theme of“leaf mold bed,”the curriculum connects with real-life problems and encourages students to apply sensor data, interpret information, design conditional logic, and engage in cross-disciplinary thinking and creativity.
Observations during implementation reveal that generative AI played multiple positive roles in the course. It assisted students in clarifying syntax, debugging code, and stimulating deeper ideation and reflective thinking. AI functioned as a capability compensator, a creative thinking enhancer, and a co-instructor, while also supporting teachers in delivering differentiated instruction and personalized guidance for gifted learners.
However, several potential concerns were also observed. Students tended to over-rely on AI, which led to compressed thinking processes and a tendency to delegate cognitive effort—a phenomenon referred to as“cognitive outsourcing.”Additionally,“AI hallucinations”occasionally produced misleading content that students struggled to identify, especially when they lacked foundational programming knowledge.
In conclusion, generative AI demonstrates strong potential as a learning support tool in programming education for gifted students. It enhances motivation, efficiency, and the depth of problem-solving and creative thinking. To ensure its positive impact, future curriculum design should emphasize teacher-led guidance strategies, foster students’questioning and evaluative skills, and incorporate reflective, flexible, and cognitively challenging tasks. Ultimately, generative AI should serve as an extension of students’cognitive processes—not a replacement for their intellectual autonomy.