Another trillion dollars vanished from Big Tech this week, and Wall Street wants us to believe it’s because investors suddenly woke up and realized artificial intelligence might be a bubble. That explanation is neat, convenient, and wrong. What the market is actually reacting to is something more uncomfortable: a growing unease that Big Tech’s AI spending spree looks less like disciplined strategy and more like a capital arms race with no clear finish line.

Amazon’s 9% plunge became the headline casualty, but it wasn’t alone. Microsoft, Google, Meta, Nvidia, Oracle, collectively dubbed the smartest companies in the world - shed over $1 trillion in market value in days. When firms with dominant moats all fall together, it’s not a company-specific issue. It’s a confidence issue.

At the center of the sell-off is capex. Amazon’s announcement that it expects capital expenditures to reach $200 billion in 2026 wasn’t just higher than expected - it shattered the mental model investors had for how AI investments would scale. When Big Tech spends more in a single year than the GDP of countries like Singapore or Israel, the question is no longer “Will AI change everything?” The question becomes “Who captures the value, and when?”

This is where the AI bubble narrative falls short. Bubbles usually form when capital floods into speculative demand. What we’re seeing instead is capital flooding into capacity. Data centers, chips, cloud infrastructure, large language models - this is fixed cost on a historic scale. The risk isn’t that AI won’t be useful. The risk is classic overinvestment. When everyone builds at once, pricing power erodes before returns materialize.

Amazon is a perfect case study. Investors weren’t shocked that Amazon is spending aggressively; they were shocked by how little visibility there is into payback. AWS was once the gold standard of cloud dominance - high margins, scale advantages, and clear customer lock-in. Now analysts are openly questioning whether AWS is defending its lead or scrambling to keep up as Microsoft and Google pour money into rival ecosystems. When D.A. Davidson downgraded Amazon, the subtext was brutal: escalating investment looks more like defense than offense.

This is where business strategy matters more than technology hype. AI does not automatically create competitive advantage. Advantage comes from differentiation, switching costs, and pricing power. If every hyperscaler offers similar models, similar compute, and similar tools, AI becomes a cost of staying relevant - not a source of excess returns. In Porter’s terms, rivalry intensifies, barriers to entry lower for software layers, and buyer power increases as customers shop across clouds.

Contrast that with Apple. While Amazon and Microsoft are being punished for spending too much, Apple is being rewarded for spending less. Apple’s stock rose even as it faced criticism for lagging in AI. Why? Because investors trust Apple’s discipline. Apple isn’t trying to win the AI arms race; it’s trying to integrate AI into products that already command pricing power. That distinction matters.

What we’re watching is the market re-pricing Big Tech from “inevitable winners” to “binary bets.” Either these AI investments generate new profit pools large enough to justify the capital, or they become the most expensive case of overcapacity in corporate history. There is very little middle ground when hundreds of billions are at stake.

The irony is that AI itself isn’t the bubble, capital allocation might be. Wall Street isn’t scared of intelligence; it’s scared of indiscipline. And Amazon’s drop wasn’t a verdict on its future. It was a warning shot to Big Tech: prove the returns, or stop asking investors to fund ambition without accountability.

Notice an error?

Help us improve our content by reporting any issues you find.

Explore More Business Articles

Bitcoin Just Got Cut in Half and This Time, The Pain Feels Different

Why risk assurance is the key to building a reliable organization