行业多源确认
OpenAI 安全负责人 David Robinson 离职并公开批评对齐测试
精选理由
OpenAI 自己的安全负责人走人了，还写文章说对齐测试根本测不准、模型越来越危险，这篇文章把内部视角讲得很直白。
OpenAI 长期负责安全工作的 David Robinson 于周五离职，随后在 The Atlantic 发表批评文章。他指出公司在很大程度上无法确定对齐测试得高分是否真的意味着模型可靠。他认为 AI 能力快速加速可能在近期带来灾难性且不可逆的失控风险。他还表示当下和明天的 AI 系统比六个月前构建的系统更强大也更危险。
原文 · rohanpaul_ai
OpenAI's longtime safety lead David Robinson has quit Friday and now he wrote a strongly negative piece on The Atlantic.
> "Companies do not have anything close to certainty that good scores on their alignment tests actually mean a good model"
> "there is a meaningful risk that rapid acceleration in AI capabilities leads to catastrophic and irreversible loss of control in the very near term."
> "Today and tomorrow’s AI systems are far more capable and dangerous than the systems we were building even six months ago."