这种通膨压力主要由数据中心对记忆体晶片的暴增需求所推动。根据彭博经济研究公司的计算，在情况缓解之前，这种记忆体紧缩预计将使整体通膨率增加 0.4 个百分点。这一趋势与中国的情况形成对比，在中国，由于制造商难以将更高的成本转嫁给消费者，消费者通膨已陷入停滞。
尽管一些经济学家指出精确衡量 AI 驱动通膨的挑战，并预期未来的生产力提升将减轻这些压力，但目前的影响是确实存在的。联邦准备理事会决策者，包括理事 Lisa Cook 和圣路易联邦准备银行总裁 Alberto Musalem，都强调了 AI 热潮如何透过晶片和电力推高价格。因此，专家警告，能降低通膨的生产力繁荣仍需一段时间才会到来。
The massive buildout of artificial intelligence (AI) data centers has emerged as a novel source of inflation in the United States, driving up the costs of critical hardware used in consumer staples such as phones, computers, and cars. Official inflation data indicates a significant rise in the prices of software and computer accessories, which surged by 14.5% in May compared to the previous year, defying the historical trend of technological components becoming cheaper over time.
This inflationary pressure is largely driven by the surging demand for memory chips required by data centers. According to calculations by Bloomberg Economics, this memory squeeze is projected to add 0.4 percentage point to the headline inflation rate before it begins to ease. This trend contrasts with the situation in China, where consumer inflation has stalled because manufacturers face difficulties passing higher production costs onto consumers.
While some economists note the challenge of precisely measuring AI-driven inflation and expect future productivity gains to mitigate these pressures, current effects are actively felt. Federal Reserve policymakers, including Governor Lisa Cook and St. Louis Fed President Alberto Musalem, have highlighted how the AI boom is pushing up prices through chips and electricity. Consequently, experts warn that a disinflationary productivity boom remains a distant prospect.