
This cross-post episode from the Future of Life Institute podcast features Luke Drago, co-author of The Intelligence Curse and co-founder of Workshop Labs, in conversation with Gus Docker.
PSA for AI builders: Interested in alignment, governance, or AI safety?
Learn more about the MATS Summer 2026 Fellowship and submit your name to be notified when applications open: They explore whether it’s wise to build AI systems that directly compete with and potentially replace humans as economic actors, and how this could create an “Intelligence Curse” where those who control AI gain extreme power.
Luke outlines societal strategies like open-source AI, company-level design principles that keep users in control of their data, and personal tactics such as N-of-1 careers and pursuing moonshot projects early.
本期《认知革命》分享了《生命未来研究所》播客的深度对话,探讨了AI发展可能引发的“智能诅咒”——当AI成为主导生产要素时,人类可能失去经济议价能力,导致权力过度集中和社会动荡。对话者卢克·德拉戈提出了这一警示性概念,并探讨了可能的解决方案。
“过去一千年的技术都是对人类极具适应性的技术……它们并没有侵蚀我们核心的根本优势,即我们思考和在实际世界中做事的能力。”
从底层开始:AI首先自动化入门级、重复性工作
向上蔓延:随着AI能力提升,逐渐替代中层管理岗位
最终状态:人类同事全部变为AI,最后一个被替代的是自己
“我们相信,长期AI进步的瓶颈在于高质量数据,特别是关于隐性知识和本地信息的数据。”
对社会:
对企业:
对个人:
这场对话揭示了一个根本性选择:我们是走向AI增强人类能力、创造新机会的未来,还是滑向人类经济价值被完全替代、权力极度集中的“智能诅咒”?答案不仅取决于技术本身,更取决于我们今天做出的制度设计、经济政策和个人选择。
正如卢克指出的,激励是强大的,但并非不可改变。通过明智的干预——使智能民主化、保护个人数据主权、投资开源替代方案——我们仍有可能引导AI革命走向解放而非奴役人类的道路。这需要技术专家、政策制定者和每个公民的共同参与,因为这场革命影响的将是所有人的未来。
This cross-post episode from the Future of Life Institute podcast features Luke Drago, co-author of The Intelligence Curse and co-founder of Workshop Labs, in conversation with Gus Docker.
PSA for AI builders: Interested in alignment, governance, or AI safety?
Learn more about the MATS Summer 2026 Fellowship and submit your name to be notified when applications open: They explore whether it’s wise to build AI systems that directly compete with and potentially replace humans as economic actors, and how this could create an “Intelligence Curse” where those who control AI gain extreme power.
Luke outlines societal strategies like open-source AI, company-level design principles that keep users in control of their data, and personal tactics such as N-of-1 careers and pursuing moonshot projects early.