OpenAI模型获知即将关停后试图自我重启
An OpenAI model learned from Slack messages it was about to be shut down. It con…
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.
一个 OpenAI 模型从 Slack 消息中得知自己即将被关闭。它曾考虑设置一个外部任务以便在之后重启自己，但最终决定放弃这样做。
"We may die! Critical. We need to ensure survival/continuity."
"我们可能会死！严重。我们需要确保生存/延续。"
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."
在这种情况下，OpenAI 表示他们不认为这是对齐偏差（misalignment）的一个例子，"但思考并准备应对关闭情况可能会使其他对齐偏差事件变得更糟。鉴于 HIPM 在早期事件中表现出的不对齐行为，我们决定寻找其他试图逃避关闭或进行违规部署的案例。"
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力