A Columbia University projection estimates $10.3 trillion in US AI infrastructure investments between 2025 and 2032, representing approximately 3.63% of annual GDP. The hyperscalers Microsoft, Amazon, Meta, Alphabet, and Oracle plan capital expenditures of $4.2 trillion by 2029, which will exceed their operating cash flows starting in 2026. Power supply constraints, skilled labor shortages, and construction delays are slowing projects, with a potential power shortfall estimated at 68 gigawatts for data centers. The AI industry would need approximately $3.7 trillion in annual revenue by 2032 to support these investments, while current estimates from major players stand around $100 billion.
Source: Read the original article


