AI 导读
一位 OpenAI 安全研究员表示，竞相推进 RSI 是世界史上前所未有的愚蠢行为；自 9 年前 RLHF 以来对齐没有根本进展，能力远超对齐的预测已成真，需要很快显著放慢前沿研究。引用内容指出这是 OpenAI 首席科学家 Jakub Pachocki，其写道目前没有实验室将对齐和监控解决到足以继续全速扩展的程度，期望自愿放慢成为常态，并认为未来 AI 发展的国际协调应成为各国政府的首要任务。
正文
OpenAI safety researcher: "racing into RSI is a folly that is unprecedented in world history."
"there have not been fundamental advancements in alignment since RLHF [9 years ago]"
"the prediction that capabilities would race far ahead of alignment has come true"
"we will need to significantly slow the frontier of research very soon"
I should have led with this but here is the chief scientist of openai: not a crank or a newcomer, but rather among the top most respected scientists in the field, someone who has solved innumerable capabilities problems on the way here在 X 查看被引用的帖子
来源：AI Notkilleveryoneism Memes ⏸️ · x.com