AI and Productivity – Stripe Economics (www.stripeeconomics.com)

Recent research underscores the significant impact of AI tools on productivity, revealing that large language models (LLMs) can enhance the efficiency of customer service agents by 14% and writers by 40%. AI-driven transformations in workplace dynamics show a trend of micro productivity gains, such as consultants completing 12% more tasks and reduced time spent on emails, leading to an overall rise in U.S. labor productivity to approximately 2.5%. However, there remains a disconnect between these task-level improvements and broader aggregate productivity, particularly as total factor productivity (TFP) growth has stagnated, raising questions about the scalability of AI's benefits. The nuanced analysis indicates that while AI enhances specific tasks, these efficiencies have not yet translated into significant aggregate productivity gains due to persistent human bottlenecks and existing workflow structures. Surprisingly, the current productivity surge seems more tied to increased capital utilization—companies are maximizing their existing resources rather than fully realizing the potential of AI technologies. This revelation suggests a crucial path forward: as businesses adapt and reconfigure their operations around AI, the anticipated macro-level productivity enhancements may eventually materialize, supporting the ongoing evolution of the AI/ML landscape.