《技术革命与金融资本》作者Carlota Perez认为，AI仍处于"狂热部署阶段"，历史上每次技术革命都需要经历泡沫破裂才能进入黄金时代。与此前革命 不同，AI革命首次由软件主导，网络效应放大了机遇与风险，同时AI公司有望直接受益于其释放的广泛经济价值。 人工智能基础设施正迎来史上规模最大、速度最快的技术浪潮。然而，历史规律表明，每一轮伟大的技术革命将周期性地从疯狂投资到泡沫破裂，再到 黄金时代的到来。 《技术革命与金融资本》作者Carlota Perez认为，AI仍处于"狂热部署阶段"（installation phase），历史上每次技术革命都需要经历泡沫破裂才能进入 黄金时代。与此前革命不同，AI革命首次由软件主导，网络效应放大了机遇与风险，同时AI公司有望直接受益于其释放的广泛经济价值。 当前，这股投资狂潮正以前所未有的规模席卷全球。仅在今明两年，谷歌、亚马逊、微软和Meta四大科技巨头就将在数据中心领域投入高达7500亿美 元，以支持其AI模型。摩根士丹利预测，到2029年，全球在该领域的总支出将达到3万亿美元。 在最初的"部署阶段"，新技术颠覆旧有产业和地区，引发大量的"创造性破坏"和社会动荡。 ...

Group 1 - The core viewpoint is that the current AI revolution is in a "frenzied deployment phase," which historically precedes a bubble burst before entering a golden age [1][4] - Major tech companies like Google, Amazon, Microsoft, and Meta are projected to invest up to $750 billion in data centers to support AI models over the next two years, with global spending in this area expected to reach $3 trillion by 2029 [1][2] - A report from MIT indicates that 95% of surveyed companies have not seen any returns from their generative AI investments, raising concerns about the sustainability of current investments [1][4] Group 2 - Historical patterns show that technological revolutions, including the current AI wave, follow a predictable cycle characterized by initial over-investment and market bubbles [2][3] - The AI revolution is unique as it is driven by both software and hardware, altering financial dynamics and enabling rapid global expansion through network effects [5][6][7] - The competitive landscape is intensified by digital globalization, which increases both opportunities and risks, as seen with the impact of cheaper AI models on investor confidence [8] Group 3 - To transition into a golden age, society must actively shape the revolution to serve its purposes, similar to historical efforts to regulate powerful companies and address labor market disruptions [9] - Current challenges such as dysfunctional financial markets, concentrated corporate power, and climate change pose significant hurdles that need to be addressed for the AI revolution to fulfill its potential [9]