On Thursday afternoon in Washington, American Compass co-hosted a Summit on AI and Labor. Daniel has thoughts:
This week, a 7,000-word thought experiment published by Citrini Research ricocheted across the political and financial press. The research note described a hypothetical future in which AI advances so rapidly that mass white-collar layoffs destabilize the consumer economy, trigger a credit crisis, and send the S&P 500 down 38% from its highs. The Wall Street Journal called it a “doomsday report” that laid bare “Wall Street’s deep anxiety” about the future of AI.
The speculation commanded so much attention because it pinpointed a mechanism that investors, executives, and journalists intuitively understand. In the 1930s, economist Michal Kalecki identified what has since become known as Kalecki’s Paradox of Costs. His insight is simple: What is rational for any individual firm—cutting labor costs to boost margins—is catastrophic when every firm does it simultaneously, because workers are also consumers. The wages that firms suppress are the same income streams that generate the demand on which their revenues depend. A firm that replaces ten workers with a new technology saves money. But an economy in which every firm does the same destroys the consumer base that provides the revenue to keep all firms solvent, unless entrepreneurs seamlessly redeploy the labor in new and more productive pursuit—a phenomenon that economists like to assume with waving hands, but that the real world does not accommodate.
The Citrini scenario applies this logic to AI-driven headcount reduction. Companies cut staff and invest in AI, AI improves, workers become less necessary, displaced workers spend less, companies cut more: a feedback loop that doesn’t have a natural brake (unless government steps in).
Suffice to say, Citrini’s predictions may not materialize. The pace of AI progress may be slower. The buildout of physical infrastructure to support widespread deployment faces real resource constraints. Effective integration into enterprises often takes longer than technologists expect. And political backlash against a sector with a trust deficit could throw sand in the gears at every level of government. But whether or not Citrini’s scenario unfolds, the dynamic it describes is well worth contemplating, and being prepared to address.
A few days after the Citrini note was published, American Compass hosted a Summit on Labor and AI in partnership with the New American Industrial Alliance and Palantir Technologies. The program was both high-powered and eclectic:
Compact editor Matthew Schmitz interviewed Senator Josh Hawley
I moderated a conversation between Julius Krein and Edward Luttwak on industrial policy
Compass policy director Chris Griswold moderated a conversation between Catholic University’s Father Aquinas Gilbeau, Valar Atomics CEO Isaiah Taylor, and Cluny Institute director Luke Burgis on how religious faith informs these issues
Oren moderated a conversation between Teamsters President Sean O’Brien and Palantir CEO Alex Karp that covered, well, a lot
The Citrini note’s underlying prediction—that the AI revolution will replace the worker rather than augment his abilities, and that it will hit white-collar workers hardest—was a throughline in each of the panel discussions. O’Brien, in his conversation with Karp, put it plainly: For once in his lifetime, the people who have never been economically vulnerable are about to experience what most working people have experienced for decades.
O’Brien’s observation is worth dwelling on. For 30 years, the organizing logic of the American economy has been a flight from the physical: from atoms to bits, from factories to platforms, from making things to managing things. The professional class built its security on the assumption that the future would always reward credentialed, cognitive symbol manipulation and that the people who worked with their hands were on the wrong side of history. It was a reasonable bet that paid handsomely for a long time—and may now be wrong.
The same qualities that made white-collar work seem secure—its abstraction, its scalability, its distance from physical constraint—are precisely what make it vulnerable to a technology that processes information more quickly and cheaply. The professional class watched the nation’s industrial base atrophy and drew the wrong conclusion: that cognitive work was insulated from the logic of substitution. AI is correcting that assumption.
That logic has a history that animates much of our work at American Compass. Stagnant wages and productivity growth that didn’t accrue to the typical worker were the consequence of a market deliberately unbound from any recognition of mutual obligation and from a misalignment between that which generates profit and that which advances the national interest. Globalization made cheap foreign labor abundant and sent capital abroad. Mass immigration suppressed wages and discouraged investment at home. Financialization enabled profit from arbitrage rather than expansion of domestic capacity.
AI arrives in that same unbound system. If the gains flow the same way—to capital and away from the workers whose labor it displaces—the Kaleckian feedback loop Citrini describes becomes the default.
Julius Krein, editor of American Affairs and one of the sharper analysts on matters of political economy, closed the panel I moderated on an optimistic note. The professional class, he argued, shouldn’t be understood as AI’s inevitable victim. In a political moment defined by reindustrialization, AI, and competition with China, there will be ample opportunities for credentialed talent to redirect itself toward building the physical economy rather than managing the financial one. The engineers, lawyers, and financiers who spent three decades optimizing a globalized system could spend the next three building a national one. And the capital markets that exploited arbitrage opportunities could instead finance a reinvigorated industrial base.
Blue-collar workers were the losers of the last 30 years. AI threatens to make the professional class the losers of the next 30. Inversion of their fortunes is the wrong goal, as is managing the decline of one class while celebrating the revival of another. What we should strive for instead, and aim to have well-structured and well-supported markets deliver, is convergence: the talents of the professional class and the resources of our capital markets redeployed in the service of all workers. The last 30 years unbound the market from the nation. The next 30 must reconstitute and celebrate the ties that bind us together. — Daniel
SPEAKING OF LABOR AND AI
We’ve got the reliably awful: Sam Altman, foot reliably in mouth, one-upped himself in India this week, arguing:
One of the things that is always unfair in this comparison is people talk about how much energy it takes to train an AI model relative to how much it costs a human to do one inference query. But it also takes a lot of energy to train a human. It takes, like, 20 years of life and all of the food you eat during that time before you get smart.
The audience laughs awkwardly, but Altman is not joking and plows ahead. Regular reminder: If Nobody Trusts You, They’re Not the Problem.
A new standard in awful: Not to be outdone in the dehumanization derby, Secretary of Agriculture Brooke Rollins took to Fox Business to lament the rise in fertilizer, fuel, interest rates, and labor costs, then proudly tweeted the clip. One of these things is not like the others! Oren often laments the economist’s tendency to treat labor as just another input, like lumber or electricity. Equating workers instead with fertilizer and debt manages to be even more demeaning.
Some interesting progress: In 2018, Annie Lowery wrote Give People Money, which was supposed to show how Universal Basic Income “has the potential to solve some of our most intractable economic problems, while offering a new vision of citizenship and a firmer foundation for our society in this age of turbulence and marvels.”
Eight years later, contemplating the prospect of widespread unemployment for white-collar workers that Daniel discusses above, she has come to a very different conclusion: “UBI is a dystopian outcome, not a utopian one.” She explains:
For families to thrive in this new post-work paradigm, the government would need to redistribute a lot more than $1,500 per person per month, necessitating confiscatory taxes on corporations—taxes they would fight tooth and nail. The bigger problem would be that Americans would hate a world without work, where the jobless rate floats at 30 percent instead of 4 percent. Many Americans like working. They like having somewhere to go during the day. They like trading watercooler stories with their colleagues. They like getting promoted and starting their own firms. They are proud of earning a living. Long-term unemployment destroys people’s mental and physical health, and generates toxic societal unrest. Politicians love saying that Americans want a hand up, not a handout—and they aren’t wrong.
Well said.
And a more reasoned take: In the Wall Street Journal, Greg Ip takes a look at the evidence on how AI has actually affected employment and how it most likely might (given the typical trajectory of revolutionary technologies), concluding “Tech Has Never Caused a Job Apocalypse. Don’t Bet on It Now.” Even in software development, he notes, “if such a revolution were upon us, we should see some sign of it. We don’t, at least not yet. The ranks of software developers, widely assumed to be acutely vulnerable to AI, are up 5% in January from a year earlier, a pace largely consistent with the past 23 years.”
GOOD WEEKEND READS
Snaps for our friends at American Affairs, who just published an especially interesting issue. Start with:
Rebuilding America’s Sea Power, by Senator Todd Young
The Global Industrial Development Toolkit: Unpacking Trump’s Investment Deals with Japan and South Korea, by Peter Harrell
Between Hype and History: Conversations with the AI Elite, by Sophia Brown-Heidenreich
Friend of the ’stack Ruy Teixeira has an excellent essay on “No Populism Without Cultural Populism”:
This is a bitter pill for most Democratic elites to swallow. In today’s America, they are the “Establishment” even if in their imaginations they are sticking it to the “Man” and fighting nobly for social justice. The failure to understand that they themselves are targets of populist anger is a central reason their populist pitch fails—and will fail—to get traction among the working class.
Panka Bencsik and Tyler Giles have posted a fascinating new paper, “Local Prosecutors and Public Health,” that finds “narrow election of a Republican prosecutor reduces all-cause mortality rates among young men ages 20 to 29 by 6.6%. This decline is driven predominantly by reductions in firearm-related deaths, including a large reduction in firearm homicide among Black men…” (h/t Santi Ruiz)
Related: In case you missed it, be sure to read our own Drew Holden’s essay from last week, “Cities Really Can Just Enforce the Law.”
And in Fortune, a devastating deep-dive into technology in schools: “The U.S. spent $30 billion to ditch textbooks for laptops and tablets: The result is the first generation less cognitively capable than their parents.” (h/t Jonathan Haidt, who observes, “putting computers and tablets on students desks in K-12 may turn out to be among the costliest mistakes in the history of education.”)
KEEPING CALM AND TARIFFING ON
Checking in on the data: At the American Enterprise Institute, Derek Scissors takes a very fair look at the data on the first year of the global economic reset, acknowledging, “12 months of financial success worth trillions and nine months of trade success, at least on Trump’s view.” But, he says, “that’s not the end of the story. The last three quarters of the 2025 also saw manufacturing employment decline each month.”
As we emphasize repeatedly, manufacturing job growth is not the right barometer of reindustrialization’s first year. Even if one wanted to look there, the sector’s job decline slowed by half from the prior year. Scissors acknowledges as much, and then notes, “it can take time for trade and financial changes to affect employment.”
To summarize his analysis, then: the economic results from the administration’s trade policies look promising, except for the job numbers, which also improved and are not yet relevant. OK, then! (Perhaps, “2025 Trade and Investment Didn’t Yield Manufacturing Jobs” was not the right title for the piece?)
Speaking of data, Bloomberg highlights the growing gap between what China says it exports to the United States and what the United States says it imports from China, suggesting substantial underreporting by importers to evade tariffs. As one importer who does obey the law puts it, “the tariffs are just the cost of doing business, but the tariff cheats are the ones that are very, very damaging.” The problem is parallel in many ways to the employment of illegal immigrants, where law-abiding employers get undercut.
Solution: Fund the trade police. More detailed solution: Build an AI model that filed reports by importers with local data on the originating source of the products.
But what about the Supreme Court? Well, be sure to listen to this week’s episode of the American Compass podcast, in which senior political economist Mark DiPlacido and I go through the details of the ruling, the alternative authorities on which the administration is now relying, and what to expect next.
If reading is more your thing, Peter Harrell’s analysis at Lawfare is excellent.
Or, if you prefer badly uninformed commentary, there’s always economist Justin Wolfers, who seems confused by the Court’s decision regarding the interpretation of the regulatory power under IEEPA and thinks, instead, that the tariffs were declared unconstitutional.
Now underway: The Wall Street Journal explains how the administration will use new Section 232 tariffs to target sectors vital to national security. And U.S. Trade Representative Jamieson Greer announces the initiation of investigations under Section 301 to address unfair practices of trading partners.
Ambassador Greer also delivered an excellent speech at his alma mater, the University of Virginia School of Law, on international law. Of particular note: “The United States raised the alarm again and again, but it became clear that this problem could not be resolved using old, slow moving, and outdated international tools.”
This is the point where even the most ardent enthusiasts of the “rules-based trading system” ultimately get stumped. The conclusion of Oren’s interview with Ernesto Zedillo, former Mexican president and professor at the Yale Jackson School of Global Affairs, was a particularly good example.
Prior to Trump, was the WTO working as intended? No.
Was there any plausible path to WTO reform, given its failure to conclude a successful negotiating round after its establishment in 1995? No.
If the WTO was not working and could not be reformed, should the United States simply have lived with a dysfunctional status quo indefinitely? Hmmm, well things get harder there…
So then what was the alternative to abandoning the broken system and starting anew?
BROKEN SYSTEM, CHINA EDITION
China, of course, is exhibit A in this broken system. The rest of the world largely ignored the problem as the United States absorbed its imbalances, but with Trump saying no mas, the problem is everyone’s now.
What Say You, Germany? As Thorsten Benner, Jakob Hensing, and Florian Klumpp write in an especially comprehensive look at “Germany’s China Shock” in the new issue of Internationale Politik Quarterly, “for a long time, Germany’s political and business elites have chosen to ignore the economic threat emanating from Xi Jinping’s China. It’s now time to confront it.”
Will they confront it? The Economist gives reasons for hope, and reasons for doubt:
Exports of German cars and other goods to China have fallen off a cliff, and surging imports from Chinese companies facing price pressure at home have driven the trade deficit to €90bn ($105bn), a head-spinning 2% of German gdp. With domestic demand subdued, Chinese exporters of cars and other goods are seizing market share from German rivals in third countries (see chart 1). Parts of German industry are crying foul over Chinese subsidies and a grossly undervalued yuan. “Our companies are competing not only with Chinese rivals, but with the Chinese state budget,” says Oliver Richtberg, head of foreign trade at the VDMA, an association of mainly German machinery firms….
Still, industry does not speak with one voice. Some multinationals, like basf, a chemicals giant, have doubled down on their investments in China. A growing number of German firms are “localising”: using Chinese supply chains, developing products with local workers and reinvesting profits in China. Volkswagen, among others, is accelerating its plans to use China as an export hub to the rest of the world. Meanwhile it is slashing jobs in Germany.”
What Say You, Japan? According to the Financial Times, “China has restricted exports of rare earth magnets and other critical materials to dozens of leading Japanese companies in an escalation of a dispute with Tokyo. China’s commerce ministry on Tuesday said it would freeze the flow of ‘dual-use’ materials, which have applications to both civilian and military industries, to 20 Japanese companies. Another 20 groups are being added to a new ‘watch list.’”
And What Will We Say? In the New York Times, Jonas Nahm explains that the decline in U.S. competitiveness is a function of not only unfair trade practices and economic distortions, but also a genuine productivity shortfall:
The barriers are practical rather than technological, especially for small and midsize companies. In many factories, production data is incomplete or still recorded manually, making it impossible to use digital tools that rely on continuous information. Three-quarters of manufacturers surveyed struggle to connect older machines to systems that could help them run more efficiently. And eight in 10 report shortages of workers who can use A.I.-powered manufacturing tools. More than half say the upfront cost of A.I. projects is prohibitive.
…
America has emphasized invention and breakthroughs over deployment. But sometimes, technology needs to be treated as factory work — unsexy and perhaps boring, but essential for being competitive.
REINDUSTRIALIZATION
The New York Times goes into the details of the U.S. dependence on Taiwanese semiconductors, landing on this stumper of an “intractable problem”: “New plants won’t be built in the United States unless companies agree to buy the chips produced in them, which would be more expensive and cut into profits. It has been a Catch-22 that federal intervention has struggled to solve.” What a Catch-22! How could federal intervention possibly make the purchase of domestic semiconductors more attractive than the purchase of foreign semiconductors, when the foreign ones are cheaper? Think, people, think.
And be sure to read on for this anecdote:
In an Oval Office meeting, the president told Mr. Huang that he planned to put tariffs on semiconductors because making them in Taiwan was risky, two people familiar with the meeting said. Mr. Trump told Mr. Huang that when he spoke with Mr. Xi about the island, China’s leader would breathe heavily, said one of these people who was briefed on the conversation. The president didn’t like it. He urged Mr. Huang to make chips in America.
What has us breathing heavily here at Understanding America? “Apple Plans to Manufacture Mac Mini in Houston” (Wall Street Journal). Read also the inside story of the reshoring:
The scale of construction, at TSMC and other suppliers, shows Apple’s effort to reshore its chip supply chain is bearing fruit. The Wall Street Journal toured the desert Southwest with Apple executives to see facilities that its purchasing heft and investments are helping to build.
But what has us huffing angrily? SemiAnalysis reports: “Micron spent 612 days on the environmental impact study, including a 45 day public input period. Yet hours before groundbreaking, they were hit with a lawsuit calling the process ‘unnecessarily rushed.’”
This sort of garbage has to stop. Maybe we could start a “Concerned Citizens for Reindustrialization” non-profit that goes around filing onerous litigation against whatever clowns file this sort of litigation? As SemiAnalysis notes:
The lawsuit itself didn’t come from a ground roots uprising. A CA-based workers rights group went door to door in NY seeking plaintiffs. They eventually found just 6 people willing to sign on. But that’s enough to potentially halt the project. … One member of the suit, who as a former lawyer you might expect to be smart, says that “Syracuse has the highest child poverty rate in the country. What is Micron doing about that?”
“Permitting reform” isn’t going to fix this, we need to repeal the procedural environmental statutes. Their costs wildly exceed their benefits.
FINANCIALIZATION
And finally, on Wall Street, hyper-rational financial actors whose value creating innovations are questioned only by unsophisticated rubes are up to their old value creation again.
Rubes like Jamie Dimon: “Dimon Sees Pre-Crisis Parallels as Rivals Do ‘Dumb Things’” (Bloomberg).
Rubes like this “senior partner at a leading international law firm and an adjunct professor at a leading United States law school,” who just published anonymously an article in the University of Chicago Business Law Review, which “argues that the core tools of PE value creation—high leverage, cash extraction, and short-term exit incentives—externalize predictable risks to third parties including workers, healthcare patients, consumers, unsecured creditors, communities and the environment.” Huh. Could the fact that such insider-experts feel that they cannot say these things publicly have anything to do with the fact that not a lot of insider-experts say these things publicly?
At least private equity continues to deliver tremendous value for its investors, like pension funds and universities, by… nope, not true either. “Private Equity’s Dry Spell Worse Than 2008 Crisis, Bain Says” (Bloomberg).
At least we found one good use for financialization. If you’ve made it this far, you’ll definitely want to take a few minutes to read about hero policy wonk Alan Cole betting his life savings against DOGE’s ability to reduce government spending (Wall Street Journal). Alan, we salute you.
Enjoy the weekend!