投入人工智慧的庞大资金需要广泛的经济整合才能证明其合理性。虽然大型科技公司的宣传论述强调广泛的产业采用，并举例说明如微软的 AI 模型市场拥有 100,000 名客户——包含 Levi Strauss 等传统企业——但潜在的现实却揭示了一个高度集中的产业版图。

对微软 AI 营收的分析显示，其严重依赖少数科技实体，而非多元化的经济基础。最显著的是，微软大约 70% 的 AI 业务（高达 241 亿美元）在最近一个会计年度直接由 OpenAI 产生。除了此主要合作伙伴关系之外，Azure AI 的最高支出皆来自其他大型软体与 AI 新创公司。例如，截至六月，ByteDance 预计每年支出超过 10 亿美元，而 Meta 每年贡献数亿美元，每周消耗数兆个运算权杖。

这种高度的营收集中引发了人们对科技产业内部商业交易循环性质的重大担忧。尽管标志性的科技公司成为新兴数位工具的主要早期采用者是合乎逻辑的，但目前的财务指标表明，AI 尚未在更广泛的全球经济中实现真正的普及。因此，如 Jensen Huang 与 Satya Nadella 等产业领袖持续积极提倡在传统领域整合 AI，他们体认到，虽然该技术目前对大型软体企业具有变革性，但其整体的经济可行性仍未得到证实。

The massive investments poured into artificial intelligence require broad economic integration to be justifiable. While promotional narratives from major technology companies emphasize widespread industrial adoption, citing examples like Microsoft's 100,000 customers for its AI model marketplace—including traditional businesses like Levi Strauss—the underlying reality reveals a highly concentrated landscape.

An analysis of Microsoft's AI revenue demonstrates profound dependence on a small cluster of technology entities rather than a diversified economic base. Most notably, approximately 70% of Microsoft's AI business, translating to a staggering $24.1 billion, was generated directly by OpenAI in the most recent fiscal year. Beyond this primary partnership, the highest expenditures on Azure AI stem from other major software and AI startups. For instance, ByteDance was projected to spend over $1 billion annually as of June, while Meta contributes hundreds of millions of dollars each year, consuming trillions of computational tokens every week.

This intense revenue concentration raises significant concerns about the circular nature of business dealings within the technology sector. Although it is logical that iconic technology firms are the primary early adopters of emerging digital tools, the current financial metrics suggest that AI has not yet achieved genuine mass adoption across the broader global economy. Consequently, industry leaders like Jensen Huang and Satya Nadella continue to aggressively advocate for AI integration across traditional sectors, recognizing that while the technology is currently transformative for massive software enterprises, its overarching economic viability remains unproven.