利用道德掛鉤於人機協作腦力激盪提升當責意識效果探討

Abstract

本論文探討在人工智慧協助腦力激盪過程中，道德掛鉤如何提升當責意識。隨著生成式人工智慧（GAI）的廣泛應用，本研究根據道德脫鉤理論探討如何能夠減輕減少當責意識和對AI生成內容的被動依賴等風險。透過一系列實驗設計，涵蓋不同的道德掛勾設計，本研究評估了這些道德掛鉤對參與者貢獻程度與腦力激盪點子創新度的影響。研究發現指出，特定的道德掛鉤策略，尤其是提高貢獻透明度，顯著提高了參與者的當責意識和腦力激盪成果的多樣性。本研究不僅擴展了道德脫鉤的理論框架，也為人機協作腦力激盪AI介面設計提供了建議。

This thesis explores the role of moral engagement in enhancing accountability during human-AI brainstorming. With the proliferation of Generative Artificial Intelligence in problem-solving, this research investigates how moral mechanisms can mitigate risks such as reduced accountability and passive reliance on AI-generated content. Through a series of experimental designs involving two moral engagement interventions, the study assesses the impact on participants' perceived contribution and uniqueness of brainstorming ideas. The findings suggest that specific moral engagement strategies, particularly transparency creation, significantly improve participants’ accountability and the diversity of brainstorming outcomes. This research contributes to the understanding of effective human-AI collaboration, offering insights for designing AI systems that foster ethical interaction and enhanced creative participation.