Goldman Sachs released an updated research note covering the massive acceleration in capital expenditures. They project that the four primary hyperscalers — Alphabet, Microsoft, Amazon, and Meta — will spend a combined $5.3 trillion in capital expenditures between fiscal 2025 and 2030.

If you expand the lens over the next five years to include the broader tech industry ecosystem (adding Apple, Oracle, Tesla/xAI, private equity infrastructure funds, and global utility grid upgrades), Goldman estimates total baseline AI infrastructure spending will hit roughly $7.6 trillion . To put that into perspective, $5.3 trillion is larger than the entire annual gross domestic product (GDP) of economic powerhouses like Japan or India.

I recently read "1873: The Rothschilds, the First Great Depression, and the Making of the Modern World" By Liaquat Ahamed, which was a fascinating look at the end of the railroad boom, which was financed by credit.

To give some perspective, the railroad boom of the 1870s–1880s was massive by any measure. Spending peaked in the early 1870s at nearly 5% of GDP, and railroads consumed 15–20% of total national investment through the 1870s and 1880s.

Today's AI buildout is accelerating but hasn't reached those levels. Global AI infrastructure spending totaled $318 billion in 2025, more than double 2024's $153 billion, with the U.S. accounting for 77% of that. As a share of GDP, estimates range widely: Goldman Sachs puts hyperscaler AI capex at roughly 0.8% of GDP, well below the 1.5%+ peaks of prior tech booms. So we may have a way to go yet.

abromber