Nvidia’s AI surge is starting to look like 1980s Japan, a former New York Times Tokyo bureau chief warns

Nvidia’s latest results have once again underscored how dominant the company has become in artificial intelligence infrastructure, and one high-profile observer is warning that the current excitement could mirror an older cycle from Japan. Howard W. French, a former New York Times Tokyo bureau chief, cautioned that Nvidia’s AI boom is beginning to resemble the kind of overheating that marked Japan’s 1980s era of exuberance, according to commentary carried by Yahoo Finance.

The company reported another record quarter with revenue rising 106% year over year to $96.2 billion. Profit more than doubled to $59.7 billion, reinforcing that Nvidia’s supply of AI chips and related systems has been translating quickly into earnings power rather than staying confined to early-stage adoption.

French’s broader point, as framed in the discussion, is not that technology progress is fake, but that markets can lose perspective about durability when expectations run ahead of what will ultimately translate into sustainable, economy-wide returns. In his view, the visible signs of expansion can create a feedback loop in which investors focus on momentum and scale, while the transition from hype to long-term value becomes harder to distinguish.

The parallel to 1980s Japan, in that telling, is a reminder of how rapidly sentiment and capital spending can surge when a new era is believed to be unstoppable. Japan’s boom period is widely discussed as one in which asset prices and corporate ambition rose quickly, only for the long-term economic path to diverge from early optimism. French’s warning is essentially that modern markets can be vulnerable to the same kind of mismatch between story and eventual payoff.

Nvidia’s role in today’s AI buildout, however, is closer to an enabling platform than a single product fad. The company supplies GPUs, or graphics processing units, and related AI compute building blocks that data centers use to train and run machine-learning models. That matters because AI workloads require massive parallel computation, and the chip ecosystem becomes a practical bottleneck or facilitator for large customers.

Even so, the market-news commentary does not lay out any new Nvidia-specific operational weakness or regulatory issue. Instead, it focuses on the risk side of the cycle, emphasizing how quickly conditions can shift when a market starts treating current performance as a permanent baseline rather than as the peak of a particular adoption phase.

A key uncertainty is how much of French’s comparison should be read as a general macro warning versus a forecast of what will happen to Nvidia’s financial trajectory. The discussion does not provide additional data on competitors, contract terms, demand visibility, or how Nvidia’s growth rate might change, nor does it specify what concrete indicators would confirm the “1980s Japan” analogy in the coming quarters.

For investors and industry watchers, the immediate thing to watch is whether Nvidia’s results continue to sustain unusually high growth and margins as the AI buildout matures, and whether customers’ spending shifts from early infrastructure buildouts to more repeatable, long-cycle deployments. In a market defined by fast-moving expectations, the durability test will likely be less about whether AI compute remains necessary, and more about how quickly financial results converge toward more normalized patterns.