The massive capital expenditure by hyperscalers on AI will likely create an oversupply of capacity. This will crash prices, creating a golden opportunity for a new generation of companies to build innovative applications on cheap AI, much like Amazon utilized the cheap bandwidth left after the dot-com bust.

As AI infrastructure giants become government-backed utilities, their investment appeal diminishes like banks after 2008. The next wave of value creation will come from stagnant, existing businesses that adopt AI to unlock new margins, leveraging their established brands and distribution channels rather than building new rails from scratch.

The frenzy in AI investment mirrors past technological revolutions like railways. Following Schumpeter's theory, overinvestment occurs as many firms race for dominance. This leads to a bust where most fail, but the infrastructure they built remains, benefiting society in the long run.

Massive upfront capital expenditure (CapEx) for AI infrastructure creates a timing gap before revenue materializes. This mirrors historical bubbles like the dot-com and railroad eras, where the technology succeeded but early investors were wiped out waiting for returns.

Massive investments in AI hyperscalers are not the end game. They are laying foundational infrastructure, like the 19th-century electrical grid, which will enable a future explosion of derivative applications across all industries.

The AI industry's massive infrastructure spending mirrors historical tech bubbles like railroads and the internet, where the initial investors were bankrupted. The truly profitable companies—the "inheritance generation"—emerged later, building on the now debt-free infrastructure left behind. AI is likely following this same pattern.

History shows that revolutionary technologies like AI require massive, often debt-fueled, infrastructure buildouts. The revenue from these technologies frequently lags the debt obligations, causing the first generation of investors to go bust. Real wealth is often captured by later investors who buy in after the initial collapse.

The massive, redundant CapEx in AI infrastructure is analogous to the late-90s fiber-optic boom. While that fiber enabled future giants like Netflix, the initial investors went bankrupt. This suggests the ultimate beneficiaries of AI may be society and end-users, not the companies spending trillions on the build-out.

Even if the current AI boom is a bubble that bursts, the outcome is a net positive for society. Like the railroad and dot-com bubbles, massive investment creates infrastructure (data centers, models) that will fuel future innovation for everyone, even if some investors lose money.