US AI Compute Spending Hits $219 Billion by 2025 as Inference Prices Drop 94%
Updated
Updated · Financial Times · Jun 1
US AI Compute Spending Hits $219 Billion by 2025 as Inference Prices Drop 94%
2 articles · Updated · Financial Times · Jun 1
$219 billion in US AI compute spending by 2025, up from $37 billion in 2023, coincided with a 94% quality-adjusted drop in inference prices, according to economists Anton Korinek and Patrick McKelvey.
That combination implies AI output may be rising far faster than official GDP captures, because standard price-adjustment methods struggle to reflect rapid quality gains in AI services.
Their upper-bound estimate suggests properly measuring those price declines could lift 2025 real GDP by 4 percentage points, though the FT report says that figure likely overstates the visible economic impact.
The debate comes as first-quarter GDP data show a sharp rise in computer and software investment from early 2025, while imported data-centre components and uncertain productive use of AI output still cloud the true growth effect.
Is the AI boom creating real economic value or just an inflationary bubble for a few tech giants?
How do we balance AI's invisible economic gains against its very real and massive energy consumption?
If our economy is creating trillions in 'dark output,' what other massive trends are we completely missing?
The $900 Billion AI Compute Boom: Spending, Pricing Paradox, and the Race for Infrastructure Dominance
Overview
The AI compute market is booming, driven by massive investment and rapid infrastructure expansion as tech giants race to build out their capabilities. Investors are closely watching which companies will benefit first from this surge in spending, expecting it to boost economic productivity in the future. Global spending on AI infrastructure is set to nearly triple by 2029, with leading firms expanding data center capacities to support advanced AI models. This unprecedented growth highlights how strategic investments and infrastructure development are shaping the future of AI and its impact on the broader economy.