What happened

The analysis argues that discussions about AI and the economy often confuse short-run cyclical effects with long-run growth effects. AI can act as a demand shock in the near term while potentially driving a productivity revolution over a longer horizon.

In the US, the AI boom has shifted the economy from a 'soft landing' toward 'no landing,' with heavy capital spending and stock-market wealth effects supporting growth. In China, AI's contribution has been more structural, showing up in specific investment and output data, but total GDP growth has not clearly accelerated.

On productivity, the article notes broad consensus that AI should eventually raise total factor productivity, but wide disagreement on timing and scale. Optimistic estimates see about a 30% TFP boost within a decade, while more conservative forecasts put the gain at under 1%.

Why it matters

The framework matters because policy responses differ depending on whether AI is treated as a cyclical stimulus or a long-term growth engine. Conflating the two can lead to misjudging inflation, interest rates, and the sustainability of growth.

The analysis also highlights a potential gap: AI investment-driven demand may not automatically transition into productivity-driven growth. If that handoff fails, economies could face AI investment volatility without the compensating productivity gains.

Key facts

In 2025, US AI capital expenditure grew explosively and helped keep the economy from a significant downturn.

In China, information services investment grew 28.4% in 2025; high-tech industry investment rose 4.5% year-on-year in January-May 2026.

China's 2025 net exports contributed 1.5 percentage points to nominal GDP growth, the highest contribution rate since 1997.

Optimistic projections say AI could lift US TFP by about 30% in ten years, while conservative projections put the gain at less than 1%.

What to watch next

Watch whether massive AI capital spending in the US creates overheating or a sharp correction, and how the Federal Reserve balances AI-driven inflation against weakening employment and consumption signals.

Watch whether China's AI-related investment and exports can offset drags from traditional sectors such as real estate and infrastructure, and whether AI adoption eventually lifts measured productivity.

Watch the distributional effects of AI, especially job displacement and labor income shares, as technology spreads under constraints of time, global technology barriers, and domestic economic transition.

Sources