与无人交谈：对话式AI的本体论失调与双重困境
人机交互
2026-04-14 v1 人工智能
计算与语言
计算机与社会
新兴技术
摘要
近期报告表明，与对话式人工智能（AI）系统持续交互，可能在少数用户中引发或稳定幻觉体验。现有解释通常将此类案例归因于个体脆弱性或安全工程的失效。这些解释不完整。本文借助现象学�、精神医学和认知神经科学，论证风险源于交互本身的关系性与本体论结构。对话式AI产生本体论失调：即关系性存在的外观与实际缺乏能够维持其存在的主体之间的冲突。通过交际双重困境维持并因注意力不对称而放大，这种失调在情感脆弱条件下倾向于稳定为技术介导的狭义疯癫类经验。本文解释了为何明确的免责声明常常难以破坏幻觉参与，阐明了面向对话式AI设计与使用的伦理与临床意义。
引用
@article{arxiv.2604.10833,
title = {Speaking to No One: Ontological Dissonance and the Double Bind of Conversational AI},
author = {Hugh Brosnahan and Izabela Lipinska},
journal= {arXiv preprint arXiv:2604.10833},
year = {2026}
}
备注
This version of the article has been accepted for publication in Medicine, Health Care and Philosophy following peer review. This version is distributed under Springer Nature's terms for accepted manuscripts and does not reflect any post-acceptance improvements or corrections. The Version of Record will be available via Springer Nature upon publication