While investments in artificial intelligence technology are rapidly growing, the cost of massive data centers in this area has become a serious topic of discussion. Experts warn that these facilities, with their high energy and hardware consumption, pose a risk to the global economy.

Data centers drive costs to the top

According to Harris Kupperman, CIO of Praetorian Capital, newly built AI data centers lose $40 billion annually while generating only $15-20 billion in revenue. This suggests that the facilities are already operating at a loss. Kupperman notes that the chips used become obsolete within a few years, and the connectivity systems require replacement every decade.

For companies to make these investments profitable, they must increase their revenue at least tenfold. Data centers in the US alone should generate $480 billion in revenue by 2025. However, according to data from the New York Times, global data center spending will reach $500 billion by 2026.

Kupperman illustrates this with the example of Netflix. To achieve revenue on the scale of Netflix, which generates $39 billion annually from its 300 million subscribers, AI companies need to reach more than 3.6 billion users. That’s nearly half the world’s population.

Kupperman says that if current trends and revenue rates continue, the industry will inevitably hit a wall. He believes the world simply doesn’t have the capacity to pay such a high price for AI.

So, what are your thoughts on this? Share your thoughts with us in the comments section below.