Peter Smith
● Ross Hendricks recently posted a chart that's getting attention for all the right reasons: it shows just how wildly AI spending has outrun actual money coming back in. The numbers paint a picture of an industry pouring cash into infrastructure while revenues struggle to keep up.
● The data tells a stark story. Cumulative AI investment jumped from $6 billion in 2013 to a projected $889 billion by 2025. The big four—Meta, Microsoft, Amazon, and Google—are dropping around $400 billion on data centers, GPUs, and training infrastructure. Meanwhile, AI-native revenues are expected to hit just $18.5 billion by 2025, with total AI software revenues at $175 billion. Hendricks points out that this massive gap suggests spending has gotten way ahead of actual business returns.
● The fifty-to-one ratio between investment and revenue has some analysts drawing comparisons to the dot-com bubble. The concern? Delayed returns, shrinking margins, and growing investor doubt. Some experts think the tech giants need to shift focus toward making money—enterprise tools, API subscriptions, industry-specific apps—instead of just building bigger and bigger compute farms.
● Here's the reality: AI technology is advancing exponentially, but the financial payoff is still moving in a straight line. If this imbalance continues, we could see a correction like past tech booms. Then again, the early cloud era looked similar—years of overspending before profits kicked in—so maybe AI just needs more time to catch up.
Peter Smith
