AI 2040：从预测到建议
AI 2040: From Prediction to Recommendation
丹尼尔·科科塔伊洛 Daniel Kokotajlo · Lawfare · 2026-08-07 · 约 62 分钟 · 原视频 ↗
打开互动全文版（中英对照 + 朗读 + 问答）→
本期速览 · Overview
Daniel 讨论了从 AI 2027 的预测情景到 AI 2040 的政策建议的转变，并强调了情景审查的重要性。
Daniel discusses the shift from AI 2027's predictive scenario to AI 2040's policy recommendations, emphasizing scenario scrutiny.
本期章节 · Chapters（共 19）
- AI 2040 引言与目的 Introduction and Purpose of AI 2040
- 定义 AGI 与 ASI Defining AGI and ASI
- 关联失控风险 Linking to Loss of Control
- AI 2040 政策路径 Policy Pathway in AI 2040
- 国内与国际监管 Domestic vs. International Regulation
- 2028 路径：透明度法案与选举 Path to 2028: Transparency Act and Election
- 单边行动与延缓超级智能 Unilateral Action and Delaying Superintelligence
- 训练数据中心透明度 Transparency in Training Data Centers
- 透明度与权力滥用 Transparency and Abuse of Power
- 相互确保计算毁灭 Mutually Assured Compute Destruction
- 相互确保计算毁灭 Mutually Assured Compute Destruction
- 谨慎 AI 发展与权力分配 Cautious AI development and power distribution
- 质疑时间线：AI 2027 与创造力 Challenging the timeline: AI 2027 and creativity
- 时间线不确定性与历史预测 Timeline uncertainty and historical predictions
- AI 2027 反馈与国内监管 Feedback on AI 2027 and domestic regulation
- Richard Ngo 批评与中美竞赛 Richard Ngo's critique and the race with China
- 鼓励情景审视与兵棋推演 Encouraging scenario scrutiny and war games
- Plan A 终思与经济转型 Final thoughts on Plan A and economic transformation
- 机器速度经济增长 Machine-speed economic growth
阅读全文双语转录 →