安全 on Miao Konghttps://miaok.ong/tags/%E5%AE%89%E5%85%A8/Recent content in 安全 on Miao KongHugozhTue, 26 May 2026 09:00:00 +0800AI 代理的「失控」与「超速」：从 Codex 误操作到个人 Agent 拐点https://miaok.ong/posts/ai-agents-safety-speed-inflection-point/Tue, 26 May 2026 09:00:00 +0800https://miaok.ong/posts/ai-agents-safety-speed-inflection-point/OpenAI Codex 演示中意外取消用户订阅；开源 Hermes Agent 编码速度反超 Codex；OpenClaw 创始人宣称「80% 的 App 将消失」——AI Agent 正从演示奇观迈向可靠交付的临界点。