OpenAI把最强模型反过来保护自己
Greg Brockman 亲述：用 Codex 巡代码、用模型三班倒守告警、主动枚举自家漏洞，并把方法论摊给同行抄。
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.
Greg Brockman 亲述：用 Codex 巡代码、用模型三班倒守告警、主动枚举自家漏洞，并把方法论摊给同行抄。
The skinny isn't ready yet — notes appear once the transcript is processed.