Anthropic 能否单方面停下?三位 AI 安全人士重估囚徒困境叙事

ShakeelHashim · x · 2026-10-27

Vox journalist Sigal Samuel talks with Daniel Kokotajlo, Nathan Calvin, and Geoffrey Irving about whether Anthropic could unilaterally halt frontier development. The piece argues against framing AI development as a prisoner's dilemma — labs race because they assume rivals won't stop — and for simply doing the right thing, even if it means slowing down alone.

所属事件:前英国AI安全所首席科学家吁单方面暂停前沿AI(3 条相关)→

原文链接 →

「漫话AGI」频道最新

更多「漫话AGI」频道 AI 资讯 →