NVIDIA's AI Surge: A Structural Shift or Just Another Bubble?

At 45, I have seen many bubbles inflate and burst. A common pattern is that basic business fundamentals stop making sense for a while, followed by a sudden collapse.

Recently, opinions from Sequoia Capital, Barclays Investment Bank, and others suggest that NVIDIA’s customer spending by companies like Microsoft, OpenAI, Tesla, Oracle, and others is unsustainable. At first glance, it’s easy to call this a bubble and think that we are at the start of the end of this period of rapid growth and innovation.

However, I believe this situation is different. Sequoia’s analysis points to a $600 billion gap in revenue needed to justify the current spending. They estimate $100 billion in potential AI-related revenue, leaving a $500 billion shortfall.

My intuition is that NVIDIA’s numbers represent tech companies prioritizing investment in silicon over people to provide more shareholder value. Their investment is not just for creating AI-specific revenue with a Field of Dreams, “build it and they will come” mindset; it’s to transform their organizations to be AI-centric. They are integrating AI capabilities into every offering and absorbing AI costs into traditional cost centers instead of those dedicated to innovation.

To keep affording these large AI investments, companies are using automation and low-cost labor as defaults for new organizational design. They are making fundamental changes to enable AI investments across their entire operations.

I agree that if tech companies don’t make structural changes, the current investment in AI chips will prove to be a bubble that deflates soon. However, I believe there is a greater chance that these changes will be made to support the investment in AI.

This likely means that more jobs and roles at big tech companies will become obsolete, requiring rapid reskilling of the domestic tech labor force in the US.