THE LONG VIEW

In 1911, the U.S. government broke up Standard Oil, not because it was inefficient, but because its control over refining capacity had become a chokehold on the entire economy—gasoline prices, transportation, even industrial innovation were all bent to the will of one company’s balance sheet. A century later, we’re seeing a subtler but equally powerful monopoly: TSMC, not through anticompetitive behavior, but through unmatched competence and timing, has become the Standard Oil of silicon. The irony is that TSMC’s excellence—the clean fabs, the customer obsession, the process leadership—has made it indispensable, and therefore dangerous. Just as Rockefeller’s efficiency masked the systemic risk of centralization, TSMC’s reliability hides the fragility of depending on a single source for the brain of the AI age. The lesson from history isn’t that monopolies are evil, but that they distort evolution. When one node controls the rate of progress, innovation slows to its pace, not the pace of imagination. The hyperscalers, for all their power, are now in the position of early automakers begging Standard Oil for more fuel—except the fuel takes three years to refine. The real breakthrough won’t come from begging TSMC to build faster, but from realizing that competition isn’t a risk—it’s the only insurance against stagnation. And like the trust-busters of old, the market may need to force that competition into existence, not for fairness, but for survival. —Sir Edward Pemberton