And now we come to the really hard question: is hyperscaling a bubble? Will it make money? The Economist has this to say about the question:
By our reckoning, total revenues from the tech accruing to the West's leading AI firms are now $50bn a year. Although such revenues are growing fast, they are still a tiny fraction of the $2.9trn cumulative investment in new data centres globally that Morgan Stanley, another bank, forecasts between 2025 and 2028--a figure which excludes energy costs.
Today's Daily Planet is a fantastic photo essay in the Wall Street Journal on this very topic.
The artificial intelligence boom has triggered an unprecedented global construction surge, with tech giants investing over $15 billion in a single AI data center in Ellendale, North Dakota - a sum equal to a quarter of the state's annual economic output.
Most of that money goes into buying chips, energy and cooling. Very few jobs. So despite spending a quarter of the state's annual output in one location, it will not have much of an effect on the local economy. Data Center economics disembeds the installation from its surroundings.
This massive spending spree, surpassing even the cost of building the U.S. interstate highway system when adjusted for inflation, reflects a speculative bet that AI will revolutionize the economy and generate substantial profits.
However, the path to profitability remains uncertain. Companies like Microsoft, Meta, and OpenAI are pouring hundreds of billions into AI infrastructure, yet many organizations report little to no return on their investments. CoreWeave, a key player leasing data centers and chips, has amassed over $42 billion in contracts and $15 billion in debt, raising fears of a bubble reminiscent of the dot-com bust.
Despite rapid user adoption of AI tools like ChatGPT, most consumers use free versions, and businesses have been cautious in spending. Analysts estimate that $2 trillion in annual AI revenue by 2030 is needed to justify current investments, a figure exceeding the combined revenue of major tech firms.
While some argue AI's potential to boost global GDP and transform industries justifies the spending, concerns about overbuilding, high debt, and slower-than-expected technological progress persist, leaving the future of AI's economic payoff uncertain.