The global race to build artificial intelligence is attracting unprecedented levels of investment, with companies pouring money into data centres, chips and computing infrastructure. But alongside the excitement over AI’s potential, economists are questioning whether future revenues and productivity gains will arrive quickly enough to justify the spending.
AI Spending Reaches New Scale
According to projections cited by Reuters, global spending on data centres could exceed $30 trillion by 2050. The scale of investment is already being compared with earlier technology and infrastructure booms such as railways and the internet.
Anthropic alone plans to spend about $518 billion in coming years, according to its IPO prospectus, an amount far exceeding its 2025 revenue.
Companies Need Bigger Returns
The growing investment is based on expectations that AI will create new markets and significantly improve productivity. However, researchers say current AI applications may not generate enough revenue to support the infrastructure being built.
A Bain & Company study cited by Reuters estimated that AI infrastructure companies could need more than $4.2 trillion in additional revenue over the next five years to support the planned expansion.
Productivity Gains Remain a Key Question
The economic impact of AI may take longer to appear than investors expect. Economist Diane Coyle noted that productivity benefits from major technologies have historically taken years or even decades to spread through economies.
AI is already affecting employment patterns, particularly among younger workers in white-collar roles exposed to automation. A Stanford study cited by Reuters found lower employment among 22-to-25-year-olds in some AI-exposed industries compared with less AI-exposed occupations.
What Happens If Growth Takes Time?
The AI boom is being driven by expectations of breakthroughs in areas ranging from scientific research to robotics and workplace automation. At the same time, heavy borrowing and infrastructure commitments could increase financial risks if demand or revenue growth falls short.
Historical examples suggest that even when technology investment bubbles burst, useful infrastructure can remain and support future economic growth. The debate over AI, therefore, may depend not only on how powerful the technology becomes, but also on how quickly its economic benefits can catch up with the enormous investment being made today.