人工智慧趨勢下教育沙盒實驗環境：推測設計的觀點

Educational Sandbox Experimental Environments in the Era of Artificial Intelligence Trends–A Speculative Design Perspective

隨著人工智慧（AI）技術的快速進展和教育領域的持續創新，推測設計（speculative design）方法學已成為探索和想像未來教育環境可能性的關鍵方式。在軟體開發中，「沙盒」提供一個隔離的環境來安全測試新程式碼，而「教育沙盒」可說是一個被保護的實驗空間，允許教師和學生自由嘗試和開發創新的教學法和學習方式，而不會干擾現有的教學架構。本文應用推測設計的方法學來論述AI如何與區塊鏈、元宇宙中的VR/AR、大數據科技相結合，在不斷進化的教育沙盒創新環境中扮演關鍵角色。研究方法包括文獻探討和綜述國際發展現況，以具體化推測性概念及其運作機制，並從學習科學的角度探究教育沙盒中的學習方法和相關的學習素養型態。教育沙盒實驗環境構想是一個基於AI知識庫的專案學習環境框架（AI-based project-based learning），本文進一步探討虛擬及增強現實在教育沙盒中的應用，並提出實施步驟作為實踐的參考。全文旨在揭示一種創新的合作探索方式及其互動原型，期望能應對未來學習的需求並開拓新的學習方向。希望本文所提供的觀點能為教育者、學習者、政策制定者及科技開發者提供一個具前瞻性的參考模型，促進AI在教育技術中的整合與發展，為未來學習開拓新的可能性。

With the rapid advancement of Artificial Intelligence (AI) technologies and continuous innovation in the education sector, Speculative Design methodology has become a crucial approach for exploring and envisioning the possibilities of future educational environments. In software development, a “sandbox” provides an isolated environment for safely testing new code. Similarly, an “educational sandbox” can be considered a protected experimental space that allows teachers and students to freely experiment with and develop innovative teaching methods and learning approaches without disrupting existing educational frameworks. This paper applies Speculative Design methodology to discuss how AI, in conjunction with blockchain, VR/AR technologies in the metaverse, and big data, plays a key role in the evolving innovation environment of educational sandboxes. The research methods include a literature review and a synthesis of the current international developments to concretize speculative concepts and their operational mechanisms, as well as an exploration of learning methods and related literacy types in educational sandboxes from a learning science perspective. The concept of the educational sandbox experimental environment is framed as an AI-based, project-based learning framework. This paper further explores the application of virtual and augmented reality within educational sandboxes and proposes implementation steps as a practical reference. The paper aims to reveal an innovative collaborative exploration approach and its interactive prototypes, with the expectation of addressing future learning needs and opening up new directions for learning. The perspectives provided in this paper are intended to offer a forward-looking reference model for educators, learners, policymakers, and technology developers, promoting the integration and development of AI in educational technology and unlocking new possibilities for future learning.