Nvidia loses $5.7 trillion in market value as AI infrastructure bet reverses

Nvidia's market value has been reported to have fallen by $5.7 trillion, a reversal of the enormous gains that made Jensen Huang one of the most powerful figures in the technology economy. The decline, described in The Atlantic as a face-plant, marks the point where the AI infrastructure boom stopped being read as a simple success story and became a question about whether the spending machine can keep outrunning its own costs.

The central evidence is the number itself: $5.7 trillion in lost value, a scale that cannot be explained by ordinary market noise. The article's framing places the reversal inside the broader rise of Nvidia as the supplier of chips, systems and software for large-scale AI training and inference. That makes the story not just about one company's stock chart, but about the balance sheets of data centers, cloud providers, governments and investors that financed the buildout.

The pattern fits a familiar tech-economy sequence: a scarce input becomes essential, its supplier captures outsized profits, and the market begins to ask who is actually paying. Nvidia's position was built on a belief that AI demand would be perpetual, elastic and profitable enough to justify massive capital expenditure. The reported reversal suggests investors are testing that belief against depreciation, energy costs, competition and the slow translation of model capability into durable enterprise revenue.

The stakes extend beyond Nvidia and its founder. If the market revalues AI infrastructure, it changes the cost of capital for data centers, accelerators, power projects and the companies that lease them. It also weakens the political argument that public spending on AI is automatically safe because private demand is unlimited. For workers, customers and taxpayers, the number is a warning that infrastructure built on expectation moves quickly when expectation changes.