Historian Walter Russell Mead on civilizational rot, expert distrust, and why AI's deflationary effect on services may be the story everyone's missing

Civilizational rot, expert failure, and AI's deflationary upside

Walter Russell Mead, a Hudson Institute fellow and Wall Street Journal contributor, frames the dominant story of our moment as the slow unraveling of the mid-century Western model — mass middle-class society, democratic market capitalism, a functioning social safety net — that seemed, briefly around 1990, to have solved the central problems of modern civilization.

The Soviet Union fell without a nuclear war. The alliance system held. And then, bit by bit, the thing started to rot.

The cocoon metaphor

Mead reaches for an analogy that is worth taking seriously. The industrial revolution was the caterpillar. The mid-century liberal order was the cocoon — once necessary, now decaying. The question is whether we claw out of it into something better, or suffocate inside it. He sees both the pessimistic vibe and the genuine excitement around technology as symptoms of the same transition, not contradictions.

On the recurring argument that 1971 was the pivot point, Mead is skeptical. Monetary systems, social capital, institutional trust — these erode across decades, not calendar years. The gold standard, he notes, was already a "hypocritical pretense" long before Nixon ended it formally.

“The cocoon is starting to rot. And if we don't get out of the cocoon, we're going to be in terrible trouble. But the fundamental underlying thing is good. Because if we can manage to claw our way out of this sticky, decaying mess — once very functional and necessary, but now just clogging everything — then we have a new and better life.”

China's view of AI

Mead's read on Beijing is that CCP leadership sees AI not as an existential risk but as the long-awaited vindicator of centrally planned economies. The market system provides feedback through prices and democratic anger; the Chinese technocratic instinct is to replace that messy signal with a national panopticon that sees everything and optimizes accordingly.

The threat they actually worry about is more mundane: that an AI optimizing for problem-solving will simply route around Xi Jinping's ideological constraints, because those constraints are a sandbox, not fundamental principles of the universe.

Expert collapse

The expert class held its authority, Mead argues, because the mid-century model was stable enough that conventional wisdom actually worked. Read the right chapter of Keynes, apply the policy, move forward. That era is over, and the experts trained for it — selected for methodological fluency and conventional wisdom navigation rather than creative thinking under uncertainty — are poorly equipped for a world where nobody has the rule book.

The Internet accelerated the visibility of that failure, but the failure was already structural.

AI's deflationary upside

The part of the AI story Mead thinks is underweighted is the democratization of services. The industrial revolution made manufactured goods abundant and cheap; the next wave does the same for professional services that have historically been rationed by price. Legal advice, financial planning, nutritionists, personal stylists — Mead argues these are not frivolous examples. Wedding planners, once a Rockefeller-tier luxury, are now a standard middle-class fixture. The same trajectory is plausible across the entire professional services sector.

His working definition of economic progress is direct: things only the super-rich could access eventually percolate down to everyone. AI compresses that timeline.

Adaptability as the underrated variable

On five-year forecasts, Mead is honest about the limit of the exercise — too many variables, too many unknowns. But his structural bet is that humanity systematically underestimates its own adaptability. We are bad at averting foreseeable disasters in advance. We are very good at improvising once the train arrives. In 1850, over half of Americans worked the land; today, less than 1% do, and the economy did not collapse — it invented trucking, aviation, and, eventually, emotional support animal certification letters.

The policy implication he draws is less about regulating AI and more about lowering the friction for adaptation: easier business formation, lighter labor bureaucracy for small firms, education systems retooled around emerging technologies.

The open variable is timing. How fast jobs disappear, and how fast new ones form, determines whether the transition is manageable or brutal. Truck drivers and white-collar bureaucrats — what Mead calls "a primitive form of AI in which human beings apply algorithms to data" — are the near-term pressure points. Whether the safety net and the entrepreneurial culture move fast enough to meet them is the question nobody can answer cleanly yet.