> **来源：[研报客](https://pc.yanbaoke.cn)** # AI and the Economic Transition Summary ## Core Content This report from Morgan Stanley analyzes the economic impact of AI, focusing on investment spending, labor markets, and productivity growth. It highlights that AI is a significant macroeconomic force, with its adoption and spending continuing to drive growth despite external challenges such as energy price shocks. ## Main Points - **AI-Related Investment Spending**: AI-related capital expenditures (capex) are expected to remain a dominant driver of growth in nonresidential fixed investment (NFI) for 2026 and 2027. Hyperscaler spending is projected to exceed \$1 trillion in 2027, and AI-related spending shows structural trends rather than cyclical ones. - **Productivity Growth**: AI is likely to raise output per worker, especially when combined with organizational changes. Industries with high AI exposure have seen a faster acceleration in productivity, indicating that AI's impact is more about increasing output than replacing labor. - **Labor Market Impact**: The report suggests that AI disruption is more micro than macro, with early and narrowly concentrated displacement effects. Unemployment rates have risen modestly, with the highest increase observed in high-exposure occupations, but the overall impact remains manageable. Young workers are more affected by AI displacement, but the magnitude is still small. - **Task Reallocation**: Task reallocation is evident, with AI exposure leading to changes in the nature of tasks rather than widespread job loss. The productivity gains are attributed to faster output growth rather than labor displacement. - **AI Diffusion and Economic Transition**: The speed of AI adoption significantly affects labor displacement risks. If AI diffuses at a moderate pace, labor displacement is manageable. However, if feedback effects from task creation and indirect wealth effects are strong, even fast diffusion results in a smoother transition. The report compares AI diffusion speeds to previous innovation waves, indicating that AI is adopting faster than the internet but slower than some hypothetical scenarios. ## Key Information ### AI Spending Trends - **AI-Related Spending Growth**: AI-related spending is projected to grow at 7.0% in 2026 and 8.0% in 2027. - **Structural Nature of AI Spending**: AI-related expenditures show less cyclical volatility and have become a structural component of GDP growth. - **Contribution to GDP Growth**: AI-related investment has contributed significantly to GDP growth, with estimates showing a contribution of 0.4-0.5pp in 2026 and 2027. ### AI Exposure and Labor Market - **AI Exposure Index**: Occupations are classified based on their exposure to AI, with high-exposure occupations typically having higher median incomes and a greater percentage of employees with a bachelor's degree. - **Unemployment Trends**: Unemployment rates for high-exposure occupations have risen more than for low-exposure ones, but the overall impact is still modest. - **Young Workers**: Young workers show slightly more disruption from AI, though the magnitude remains small. It takes longer for them to find more AI-exposed jobs. ### Task Changes and Productivity - **Task Reallocation**: AI-exposed occupations are experiencing more task changes, which may lead to new roles rather than job loss. - **Productivity Acceleration**: Industries with high AI exposure have shown faster productivity growth, particularly in 2025, due to faster output growth rather than labor displacement. ### AI Adoption Speeds and Economic Impacts - **Diffusion Speeds**: The speed of AI adoption is a critical factor in labor displacement risks. Faster diffusion can lead to more significant displacement unless counterbalanced by task creation and policy support. - **Feedback Effects**: Strong feedback effects from task creation and indirect wealth effects can mitigate the negative impacts of AI on the labor market. ### AI Usage in S&P 500 Firms - **AI Adoption in S&P 500**: About 25% of S&P 500 firms have identified quantifiable benefits from AI usage. - **AI Projects in Production**: A majority of CIOs expect to have AI projects in production by the end of 2026, showing increasing adoption across sectors. ### AI Applications Across Sectors - **Digital vs. Physical AI Applications**: The report categorizes AI applications as either digital or physical, with examples including visual search, smart kitchens, and autonomous delivery in consumer sectors, and predictive maintenance, smart grids, and carbon tracking in energy and materials. - **AI in Financials**: AI is being used for fraud detection, alternative credit scoring, and automated back office tasks. ## Conclusion The report concludes that AI is a transformative force in the economy, with its impact more micro than macro, and that the economic transition driven by AI is likely to be manageable with appropriate task creation and policy support. It emphasizes the importance of understanding the speed of adoption and the role of feedback mechanisms in mitigating potential labor displacement risks.