人对AI敞开心扉，暴露隐私却更信任，矛盾背后的心理机制引关注。
Self-Disclosure to AI: The Paradox of Trust and Vulnerability in Human-Machine Interactions
- 用社交渗透理论解析人机互动中的信任形成过程。
- 简单机器人比人类更易获取个人秘密，体现信任悖论。
- 适合研究人机关系、隐私伦理的学者与技术设计者阅读。
本文探讨人机交互中信任与脆弱性的悖论，受亚历山大·雷本的BlabDroid项目启发。该项目使用小型朴素机器人主动与人互动，成功从个体处获取个人想法或秘密，往往比人类更有效。这一现象引发关于机器是否能激发信任、为何愿意向机器倾诉的深层思考。通过分析心理机制，结合社会渗透理论（Social Penetration Theory）与传播隐私管理理论（Communication Privacy Management Theory），揭示在分享个人信息时，感知安全与暴露风险之间的微妙平衡。同时引入后人类主义与现象学等哲学视角，探讨数字时代信任、隐私与脆弱性的深层议题。人工智能快速渗入私人领域，迫使我们重新审视并定义伦理责任。
原文摘要 · Abstract (English)
In this paper, we explore the paradox of trust and vulnerability in human-machine interactions, inspired by Alexander Reben's BlabDroid project. This project used small, unassuming robots that actively engaged with people, successfully eliciting personal thoughts or secrets from individuals, often more effectively than human counterparts. This phenomenon raises intriguing questions about how trust and self-disclosure operate in interactions with machines, even in their simplest forms. We study the change of trust in technology through analyzing the psychological processes behind such encounters. The analysis applies theories like Social Penetration Theory and Communication Privacy Management Theory to understand the balance between perceived security and the risk of exposure when personal information and secrets are shared with machines or AI. Additionally, we draw on philosophical perspectives, such as posthumanism and phenomenology, to engage with broader questions about trust, privacy, and vulnerability in the digital age. Rapid incorporation of AI into our most private areas challenges us to rethink and redefine our ethical responsibilities.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。