The Hidden Supply Chain Costs of the AI Boom
Economist Willy Shih on his new book Where Does My Stuff Come From?, RAMageddon, and why we still can't make the iPhone in America.
A Conversation with Willy Shih
The AI boom has a price tag—and much of it is buried deep in global supply chains. In this episode, economist Willy Shih joins Nilay Patel to discuss his new book, Where Does My Stuff Come From?, and to unpack the hidden costs of the AI revolution. From the "RAMageddon" squeezing memory chip supply to the stubborn reality of why the iPhone still can't be manufactured entirely in America, Shih explains how the rush to build AI infrastructure is rippling through global manufacturing.
What the AI Boom Really Demands
Every AI model runs on physical hardware—and that hardware requires enormous quantities of specialized components: high-bandwidth memory (HBM), advanced logic chips, power delivery systems, and cooling infrastructure. As 2026 hyperscalers continue to pour capital into new data centers, demand for these components has outstripped supply, creating bottlenecks that affect everything from consumer electronics to automotive manufacturing.
The "RAMageddon" Effect
Shih's analysis of what he calls "RAMageddon" describes a market in which memory manufacturers—already operating near capacity—have shifted production toward higher-margin HBM and data center-grade DRAM, leaving consumer device makers scrambling for allocation. The result is a cascade of cost pressures: rising bill-of-materials expenses for smartphones, laptops, and embedded systems, longer lead times, and a growing strategic premium on supply chain resilience.
Why We Still Can't Make the iPhone in America
A recurring theme in Shih's work is the gap between political rhetoric and industrial reality. The American manufacturing ecosystem lacks the dense supplier networks, skilled labor pipelines, and specialized tooling that make high-volume consumer electronics assembly economically viable at scale. Building an iPhone in the U.S. is not merely a matter of opening a factory—it requires reconstructing an entire industrial base, a challenge that AI-driven demand for advanced manufacturing is only beginning to address.
The Takeaway
As AI reshapes the global economy, the invisible architecture of supply chains—and their hidden costs—deserves as much scrutiny as the models themselves. Understanding where our stuff comes from is no longer an academic question; it's a strategic imperative for businesses, policymakers, and consumers alike.
If you buy something from a link, The Verge may earn a commission. See our ethics statement.
via The Verge AI