Bloomberg Economics 表示，先进经济体中 27% 的劳工，也就是超过 120 million 人，可能会受到「AI 的重大影响」，但这并不意味著所有人都会被取代。最近一项调查中，近四分之一的 CEO 表示，他们超过一半的员工将需要被「提升技能」。Randstad 将未来技能组合描述为「AI 流畅度」加上情绪智商、创造力、解决问题能力、批判性思维和伦理判断，但这听起来更像是理想特质，而不是为被裁员工制定的转型计划。
在科技产业内部，预设答案包括「全民基本收入」或「学一门技艺」，而 Anthropic PBC 和 OpenAI 的估计显示，蓝领工作受到生成式 AI 影响的程度可能较低。在中国，今年早些时候的 OpenClaw 热潮关乎争相重新培训，也关乎代理式软体的采用；而在美国，焦虑则以社区抗议资料中心为标志。任何严肃的转型期都需要即时资料追踪，了解工作在哪里被增强或摧毁，还需要转型保险、针对照护工作和教育的定向公共投资，以及在 AI 丰裕真正到来时分享生产力收益的计划。
Current data gaps leave uncertainty over whether AI will trigger a white-collar jobpocalypse, but the language of “retraining,” “upskilling” and reskilling is already being used as displacement begins to trickle. America already ran this experiment after deindustrialization: manufacturing losses were relatively concentrated, but the damage spread through families, communities and politics for generations, and the payoff from retraining policy during this era was weak at best.
Bloomberg Economics says that 27% of workers in advanced economies, more than 120 million people, are likely to be “meaningfully affected by AI,” though that does not mean all will be displaced. Nearly a quarter of CEOs in a recent survey said that more than half of their workforce will need to be “upskilled.” Randstad describes the future skill set as “AI fluency” plus emotional intelligence, creativity, problem solving, critical thinking and ethical judgment, but that sounds like desirable traits rather than a transition plan for laid-off workers.
Inside the tech industry, default answers include “universal basic income” or “learn a trade,” while Anthropic PBC and OpenAI estimates suggest blue-collar work may be less exposed to generative AI. In China, OpenClaw mania earlier this year was about scrambling to reskill as well as agentic software adoption, while in the US, anxiety is marked by communities protesting data centers. Any serious transition period needs real-time data tracking on where jobs are being augmented or destroyed, transition insurance, targeted public investments in care work and education, and a plan for sharing productivity gains if AI abundance ever arrives.