AI may be more than a tech story. That is the core argument Jared Franz makes in Fortune, where he compares the current AI buildout with China's post-2001 investment boom and says the scale now underway could reshape global markets for years.
Franz writes that when China joined the World Trade Organization in 2001, it set off the largest capital spending boom the world had ever seen. He says investments in factories, machinery, housing and other long-term assets rose from roughly $360 billion in 2000 to $3.2 trillion by 2010, with cumulative Chinese fixed investment over that decade totaling about $20 trillion after inflation.
He argues a similar surge is now underway around artificial intelligence. U.S. hyperscalers alone — Microsoft, Amazon, Alphabet, Meta and Oracle — are projected to spend roughly $800 billion in capital expenditures in 2026, up 83% year over year. Gartner estimates global AI spending will surpass $2 trillion this year. For hard AI capital investment, including chips, data centers, power, cooling and networking, Franz says estimates cluster around $10 trillion to $15 trillion globally over the next decade. For broader AI spending, including software, services and AI-enabled products, he says the 10-year total could approach $30 trillion.
That scale matters because it reaches beyond Silicon Valley. Franz says AI is lowering the cost of cognitive work the way Chinese manufacturing once lowered the cost of physical goods. Tasks involving pattern recognition, language processing or data synthesis are now facing the same kind of deflationary pressure that Chinese goods imposed on developed economies two decades ago.
The physical footprint is also growing fast. The International Energy Agency projects that global data center electricity use will more than double by 2030, reaching 945 terawatt hours, roughly equivalent to Japan’s entire current power consumption.
Franz also draws a distinction that will matter to investors and policymakers. China’s fixed asset investment reached roughly 50% of GDP at its peak and was fueled by state-subsidized capital. AI investment, he says, is still perhaps 2% of global GDP and is driven overwhelmingly by market forces. He argues that could allow diffusion to happen faster.
But the labor-market risks are real. Franz says the political economy consequences could be severe if AI substitutes for cognitive labor rather than complements it. He writes that in the U.S., job categories with high exposure to AI are deteriorating faster than low-exposure categories by about a quarter of a percentage point per month, and he says that trend appears to be accelerating.
The market has already voted on the size of this buildout. Capital is flowing into chips, data centers and electricity at a pace that looks less like a passing theme and more like a structural shift. That is good news for productivity if the gains spread through the economy.
It is also a warning for the taxpayer and for workers whose jobs sit closest to automation. Big investment booms create winners fast. They also expose who pays when the rules, wages and returns do not keep up.
