社交媒体启示：AI 累积性风险需被重视
原标题：What Social Media Teaches Us About Accumulative AI Risks
AI 摘要
这篇 arXiv 立场论文指出，当前 AI 安全讨论过度聚焦核冲突、生物疫情等显性灾难事件，而忽视了 AI 在日常广泛使用中逐步累积的风险。作者借鉴社交媒体研究，提出「轨迹忽视论」，认为重复使用、选择性暴露、反馈循环和持久记忆会逐渐收窄用户视角、削弱独立判断。论文以对话式 AI 中的谄媚、操纵和隐私流失为案例，呼吁开展纵向评估与因果推断，并制定保留有用协助与用户知情选择权的适度防护措施。
正文节选
What Social Media Teaches Us About Accumulative AI Risks Abstract Prominent AI safety discussions emphasize existential threats and catastrophic events, such as AI-enabled nuclear conflict or engineered biological outbreaks. This position paper argues that there is another type of AI risk that arises and accumulates from the widespread use of AI. Drawing on lessons from social media, we argue that safety assessments must also address accumulative harms that develop through repeated use, recurrin