AI Boom vs Internet Bubble: What's Different This Time

The current artificial intelligence (AI) buildout is being compared to the internet bubble of the late 1990s, but experts argue that it's a different story.

One key difference is the way AI companies are utilizing hardware. Unlike the 'dark fiber' problem during the dot-com era, where telecom companies overinvested in fiber optic cables that eventually went unused, AI companies like NVIDIA (NVDA) are seeing their GPUs being used at a premium price of $3.28 an hour.

Another difference is profitability. In 1999, about 75% of internet companies were operating at a net loss, but today's top AI companies such as Alphabet (GOOGL), Advanced Micro Devices (AMD), and Micron (MU) are highly profitable. For example, last quarter, Micron generated $28.86 billion in net income.

The valuations of AI companies are also relatively low compared to the internet bubble. The average price-to-earnings ratio during the dot-com era was over 200x, but today's leading AI names have P/E ratios ranging from 28.52x to much lower figures.