ARTICLE · 1128871
AI对中美两国在就业、资本到政策的影响异同 the Similarities and Differences in 's Impact in China and the US
AI对中美两国在就业、资本到政策的影响异同 the Similarities and Differences in 's Impact in China and the US今晚我们聊聊AI对中美二国在就业、资本到政策的影响异同。 Tonight, let's talk about the similarities and differences in AI's impact on employment, capital, and policy in China and the US. AI革命，让普通人首先感受到生存压力增大。原因是硅基与碳基生产对劳动力要求和需求的变化，使得劳动力由旧能力向新能力转变。这里需要说明的是，硅基指生产劳动的基础材料是以Si为基础，而碳基指生产劳动的基础材料是以C为基础的。 The AI revolution has made ordinary people feel the growing pressure to survive. The reason is that changes in labor requirements and demands for silicon-based and carbon-based production have shifted the workforce from old capabilities to new capabilities. It should be noted here that silicon-based means the basic material for production labor is based on Si, while carbon-based means the basic material for production labor is based on C. 技术革命本质上是对人的再定价，硅基与碳基对人力需求发生变化时，就业市场会通过岗位和工资变化寻找新的均衡，最终促使劳动力由旧能力向新能力转变。我国就业端表现较为明显，硅基扩张，碳基收缩。近 4 年 AI 相关的硅基就业年复合增长 3.2%，而建筑和公用事业等典型碳基行业分别下降 11.0% 和 1.5%。但是现在硅基体量小，其就业扩张难以弥补更大规模的碳基就业下降。2023-2025 年，我国 AI 相关就业增加约 270 万人，而非AI相关就业减少约 980 万人。 The technological revolution is essentially a repricing of people. When the demand for labor between silicon and carbon-based companies changes, the job market seeks a new equilibrium through shifts in positions and wages, ultimately driving the shift of labor from old capabilities to new ones. China's employment side is particularly evident: silicon-based jobs expand while carbon-based industries shrink. Over the past four years, AI-related silicon-based employment has grown at a compound annual rate of 3.2%, while typical carbon-based industries such as construction and utilities have declined by 11.0% and 1.5%, respectively. However, silicon-based jobs are currently small, and their job expansion cannot offset the larger-scale decline in carbon-based employment. From 2023 to 2025, China's AI-related jobs will increase by about 2.7 million, while non-AI-related jobs will decrease by about 9.8 million. 而美国碳基就业仍在增长，但硅基就业反而收缩，近4年美国信息技术就业年复合下降约3.0%，而建筑业和公用事业分别增长1.7%和2.1%。这是因为，我国主要是劳动要素内部的重构，即硅基对碳基的挤出。美国则发生在要素之间，即资本对人力的挤出。这可以从企业注册数得到印证，2022年至今，美国信息技术业雇员人数下降11.9%，而新企业注册数从7920增至1.32万，增长66.4%，创历史新高。 In contrast, carbon-based employment in the U.S. continues to grow, while silicon-based employment has contracted. Over the past four years, U.S. information technology employment has declined at a compound annual rate of about 3.0%, while construction and utilities have grown by 1.7% and 2.1%, respectively. This is because China's main issue is the internal restructuring of labor factors, i.e., silicon-based jobs being squeezed out of carbon-based ones. In the U.S., this happens between factors, i.e., capital squeezing out human resources. This is confirmed by the number of registered companies: since 2022, the number of employees in the U.S. information technology industry has dropped by 11.9%, while the number of new company registrations has increased from 7,920 to 13,200, a 66.4% increase, setting a new record. 劳动报酬是变化的另一重要因素。中美均出现AI领域工资有更大增幅的特征。2024 年，我国AI相关行业工资总额同比增速由5.2%升至8.5%，而其他行业仅由3.3%升至3.5%，两者增速差由1.9个百分点扩大至4.9个百分点。同样，美国截至2026年7月，AI相关行业周薪增幅同比为4.8%，高于整体的3.5%。 Labor compensation is another important factor in the change. Both China and the U.S. have shown greater wage growth in the AI sector. In 2024, the year-on-year growth rate of total wages in China's AI-related industries rose from 5.2% to 8.5%, while other industries only increased from 3.3% to 3.5%, widening the growth gap from 1.9 percentage points to 4.9 percentage points. Similarly, as of July 2026, weekly wages in the U.S. AI-related industries grew by 4.8% year-on-year, higher than the overall 3.5%. 工资增长信号已影响教育选择。我国985高校AI专业录取持续扩大，高考排名的领先位次由2022年的782位扩大至2025年的2600位，代表考生愿为AI专业付出更高的“成绩溢价”。同样，美国2014-2024 的10年间，计算机与信息科学学位授予人数从5.6万翻番至12.2万人，同期社会科学则降至16.2万人。 Wage growth signals have already influenced educational choices. AI major admissions at China's 985 universities continue to expand, with the leading ranking in the college entrance exam rising from 782nd in 2022 to 2,600th in 2025, indicating that candidates are willing to pay a higher "grade premium" for AI majors. Similarly, in the U.S., the number of computer and information science degree recipients doubled from 56,000 to 122,000 in 2014-2024, while social sciences dropped to 162,000 during the same period. 资金是支撑AI产业扩张的关键要素之一。从股权融资看，我国硅基的IPO募资占比大幅上升，今年已达 53.5%的历史高位，美股更高达79.3%，资金向硅基领域聚焦。同时，两国硅基IPO数量占比并未大幅上升，说明硅基企业单项目融资规模更大，资金正加速向头部集中。 Capital is one of the key elements supporting the expansion of the AI industry. From the perspective of equity financing, the proportion of IPO fundraising for silicon-based companies in China has risen sharply, reaching a historic high of 53.5% this year, and as high as 79.3% in US stocks, with funds focusing on the silicon-based sector. At the same time, the proportion of silicon-based IPOs in both countries has not increased significantly, indicating that the scale of single-project financing for silicon-based companies is larger, and funds are accelerating their concentration at the top. 从债务融资看，中美硅基公司债发行额占非金融发行总额的比重均自2024年低位逐步上升，但我国硅基企业对债券融资依赖比美国低。 From the perspective of debt financing, the proportion of issuance of silicon-based corporate bonds in total non-financial issuance in China and the US has gradually increased from lows in 2024, but Chinese silicon-based companies rely less on bond financing than in the US. 间接融资方面，我国硅基企业银行贷款余额保持两位数增长，显著快于基础设施、服务业和工业等领域。美国自2025年起，硅基企业大型银团信贷授信额度也显著提速，领先于医疗 健康、制造和材料等其他关键领域。 In indirect financing, the balance of bank loans for silicon-based enterprises in China has maintained double-digit growth, significantly outpacing sectors such as infrastructure, services, and industry. Starting in 2025, large syndicated credit lines for silicon-based companies in the U.S. will also accelerate significantly, leading other key sectors such as healthcare, manufacturing, and materials. 政策立法的重心在向AI倾斜。2017-2025 年，我国涉AI的中央法规由4条增至189条，地方性法规由46条增至599条。但全部中央及地方法规总量却由约35.9万条降至21.2万条，下降约41%，AI相关法规占比从约0.14‰升至约3.7‰。在美国特朗普上台后，自2025年至今已有9项行政令，直接加大对AI及其基础设施的支持力度 The focus of policy and legislation is shifting toward AI. From 2017 to 2025, China's central regulations related to AI increased from 4 to 189, and local regulations from 46 to 599. However, the total number of central and local regulations dropped from about 359,000 to 212,000, a decrease of about 41%, with AI-related regulations accounting for about 0.14‰ to about 3.7‰. Since Trump took office in the US, nine executive orders have been issued since 2025, directly increasing support for AI and its infrastructure. 中美都在用“真金白银”争夺AI和先进制造。我国支持科技创新和高端制造业发展的税费补贴规模由2024年的2.63万亿元进一步升至2025年的超3.51万亿元。美国也在持续加大税收激励，联邦研发税收抵免由2025年的约361亿美元升至2027年预计388亿美元。美国政府还通过“创世计划”“星际之门”“AI行动计划”等法案，进一步整合联邦科研设施、算力、数据与私人资本， 将政府资源与企业投资共同导向AI和芯片等关键领域。 Both China and the U.S. are competing with real money to compete for AI and advanced manufacturing. China's tax subsidies supporting technological innovation and high-end manufacturing will rise from 2.63 trillion yuan in 2024 to over 3.51 trillion yuan in 2025. The U.S. is also continuously increasing tax incentives, with federal R&D tax credits rising from about $36.1 billion in 2025 to an estimated $38.8 billion in 2027. The U.S. government has also passed bills such as the "Genesis Project," "Stargate," and "AI Action Plan," further integrating federal research facilities, computing power, data, and private capital, channeling government resources and corporate investments toward key areas like AI and chips. 从就业、资本到政策，目前社会资源更偏向硅基领域，是宏观数据和微观体感背离的核心原因。AI必将从虚拟世界走向物理世界，技术红利将从少数硅基部门的自循环，辐射进汽车、制造和服务等更广泛的行业。硅、碳也有望从“替代”走向“共赢”，届时中国的人力、算力优势将会凸显。 From employment and capital to policy, current social resources are more concentrated in silicon-based fields, which is the core reason for the divergence between macro data and micro-level sensory perception. AI will inevitably move from the virtual world to the physical world, and technological dividends will spread from a few silicon-based sectors to broader industries such as automotive, manufacturing, and services. Silicon and carbon are also expected to shift from "replacement" to "win-win," at which point China's advantages in human and computing power will become more prominent.