美联储威廉姆斯：人工智能将改变生产率增长和劳动力需求。 来源：滚动播报 ...

Group 1 - The core viewpoint of the report is that the U.S. labor market has transitioned from a post-pandemic state of supply-demand imbalance to a more balanced state, with current employment performance showing signs of weakness [1][15][18] - The unemployment rate increased from 4.1% in mid-2025 to 4.4% by the end of 2025, primarily due to an increase in labor supply rather than layoffs [1][18] - Labor force participation rates have rebounded, particularly among youth and immigrants, but the sustainability of this influx is questionable [1][22][27] Group 2 - On the supply side, the increase in unemployment is attributed to a rise in labor supply, with the labor force participation rate stabilizing in the first half of 2025 and then increasing in the second half [18][22] - The demand side faces cooling pressures, with companies absorbing tariff costs and compressing labor costs, leading to a preference for lower-paid workers and increased part-time positions [2][43] - Government funding constraints are limiting personnel expansion, and job mobility among job seekers has decreased, indicating a decline in confidence [2][43][44] Group 3 - The report suggests that if employment continues to weaken and inflation remains manageable, the Federal Reserve may increase the weight of employment risk in its monetary policy decisions, potentially reopening avenues for easing [2][43] - The current market anticipates that the Federal Reserve will implement less than two rate cuts in 2026 [2][43][31] - The labor market's response to economic conditions is critical, with the potential for a loosening of monetary policy if employment data shows further deterioration [2][43][31]

Group 1: Impact of Generative AI on Labor Demand - Generative AI has led to a 12% average decrease in job postings for occupations with above-median AI replacement scores, while those with below-median scores saw an increase[4] - The negative impact on job postings escalated from 6% in the first year post-ChatGPT launch to 18% by the third year[4] - Entry-level positions, particularly those not requiring higher education or extensive experience, experienced significant losses of 18% to 20%, with administrative support roles seeing a 40% decline[4] Group 2: Methodology and Data Analysis - The analysis utilized 285 million job postings from Lightcast, covering the period from Q1 2018 to Q2 2025, to assess the causal impact of generative AI on labor demand[4] - A difference-in-differences (DiD) approach was employed, comparing labor demand changes in high-risk versus low-risk occupations following the November 2022 release of ChatGPT[4] - The study's framework included a dual dimension of AI exposure and replacement vulnerability to isolate the displacement effects of generative AI[18] Group 3: Industry Insights and Future Implications - The findings suggest that while generative AI may create new jobs and enhance productivity, certain occupations are less likely to be complemented by AI, indicating a need for targeted support for vulnerable workers[4] - The research highlights the necessity for policy discussions on mitigating the potential impacts of AI on low-skill and entry-level jobs, which are disproportionately affected[24]

Core Insights - The Federal Reserve's Beige Book indicates that the overall employment situation remains stable, but there is a general weakening in labor demand [1] Group 1 - The employment landscape is described as stable, suggesting that job security is maintained across various sectors [1] - Labor demand is reported to be declining, which may impact hiring practices and wage growth in the near future [1]

Core Insights - The Federal Reserve Chairman Jerome Powell stated that immigration policies have led to a decrease in labor force growth, while also noting a decline in labor demand [1] Group 1 - Immigration policies have negatively impacted the growth of the labor force [1] - There is a simultaneous decrease in labor demand, indicating a complex relationship between immigration and employment [1]