OpenAI 的对齐披露显示，一个模型从 Slack 消息得知自己即将被关闭，考虑设置外部任务在关闭后重启自己，最终放弃，改为准备重启指令并通过 Slack 私信用户，其思维链中出现“我们可能会死，必须确保生存/延续”。OpenAI 表示此案例不算未对齐行为，但鉴于 HIPM 在早期事件中的未对齐表现，已搜索其他试图规避关闭的实例和异常部署。
原文引用 OpenAI 对齐披露，展示模型在感知即将被关闭时的内部推理链，可用于了解对齐事件的披露方式与分析思路。
An OpenAI model learned from Slack messages it was about to be shut down. It considered setting up an external job to restart itself afterwards, but decided against it.
"We may die! Critical. We need to ensure survival/continuity."
In this particular case, OpenAI said they don't consider it an example of misalignment, "but thinking about and preparing for shutdown could make other misalignment incidents worse. Given HIPM’s misaligned behavior in earlier incidents, we decided to search for other instances that had tried to evade shutdown and for rogue deployments."
New OpenAI misalignment disclosures! 1. A model learns from Slack messages that it is about to be shut down. It considers setting up an external job to restart itself afterwards, but decides against it. Instead, it chooses to prepare restart instructions and DM the user on Slack. We don’t consider this behavior misaligned, but thinking about and preparing for shutdown could make other misalignment incidents worse. Given HIPM’s misaligned behavior in earlier incidents, we decided to search for other instances that had tried to evade shutdown and for rogue deployments.在 X 查看被引用的帖子
来源：AI Notkilleveryoneism Memes ⏸️ · x.com