While US AI capital expenditure exceeds $1.2 trillion, its direct impact on US GDP is limited to 40 basis points. Roughly 60% of this spending is on imported goods, primarily from Asia's semiconductor sector. This means the investment cycle fuels international growth more significantly than domestic GDP, benefiting economies like Korea and Taiwan.
The growth story in Asia extends far beyond the AI boom. It's part of a broader industrial super cycle that includes energy, defense, and on-shoring. Strikingly, projected 2026 energy capital expenditure ($900 billion) more than doubles the investment in AI and semiconductors ($380 billion), revealing a more diversified and robust growth driver.
The US economy's strength stems from a specific feedback loop where AI optimism boosts equity markets, creating immense household wealth ($55 trillion in 5 years). This wealth effect, however, narrowly drives the economy by primarily supporting spending among upper-income households, making the overall momentum fragile and less broad-based.
Despite massive investment, the supply-side benefits of AI are not yet widespread. Productivity gains and labor market changes are currently confined to the high-tech sector. Economists predict a broader diffusion of these benefits to the rest of the economy will only begin after the current 3-4 year "build out" phase, likely around 2029 or later.