What if the rise of artificial intelligence doesn’t herald a jobs apocalypse but instead sparks a blue-collar renaissance? That’s the provocative question raised by a new economic analysis from Anthropic, the company behind the AI model Claude.
According to a paper released this week by Anthropic’s economics team, the future of AI and work could take several very different paths by 2030, depending on how capable the technology becomes, how quickly businesses adopt it, and how easily workers adapt. The authors stress they attach no probabilities to the scenarios and caution that the real-world outcome may fall somewhere in between.
Three Futures for AI and the Economy
In the most modest scenario, the economy looks much like today: GDP is just 1.6 percent higher than its no-AI path by 2030, annual growth ticks up to 2.4 percent, and the job market barely registers a change. At the opposite extreme, AI transforms everything β the economy balloons to 32.4 percent larger than the no-AI baseline, growing at a staggering 15.4 percent annually, but overall unemployment jumps to 11.9 percent as many displaced workers struggle to find new roles.
It’s the middle scenario, labeled “substantial,” that deserves a closer look. In that world, AI drives a powerful productivity and investment boom while still leaving plenty of work for human beings. By 2030, GDP is 8.3 percent above its no-AI trajectory, annual growth reaches 5.4 percent (versus 2 percent in the baseline), and the capital stock is 13.8 percent larger β meaning businesses have more productive equipment and assets at their disposal.
Anthropic’s model treats jobs as collections of tasks. As AI is adopted, it takes over some tasks and assists with others, while also creating entirely new tasks β some of which we may never have imagined before. In the substantial scenario, AI could handle half of all knowledge work by 2030, but most tasks overall would still be performed without it.
The Factory Floor Effect
To understand why this might be good news for blue-collar workers, consider a company weighing whether to expand a factory. AI could make engineering, scheduling, and administrative work cheaper. That cost saving can transform what was once a marginal project into a profitable one, prompting the company to greenlight the investment. The result? Increased demand for labor.
Someone has to pour the concrete, install the equipment, and keep the machinery running. Savings in the back office become opportunities on the shop floor. Higher returns encourage additional investment, which makes workers more productive, which in turn creates more work for the building trades and the people who operate factory machines. Demand for skilled and unskilled manual labor expands in this scenario β even if those workers never touch AI directly.
Winners and Losers in Wages and Employment
The numbers in the substantial scenario bear this out. Real wages for occupations outside knowledge work β a broad category that includes service and blue-collar jobs β are projected to be 5.9 percent above their no-AI path by 2030. Knowledge workers, however, face a different fate: their wages are 0.3 percent below the projected path, and employment in knowledge work falls 3.9 percent from mid-2026 as AI takes over many of their tasks.
Overall unemployment in this scenario reaches 4.6 percent, compared with a 3.8 percent baseline. While that’s a meaningful increase β and if it happened suddenly, it would trigger recession signals under the Sahm rule β it would still be historically low.
The model assumes wages adjust slowly, which researchers say is realistic. Employers are hesitant to cut pay for fear of damaging morale or losing valued employees, and workers naturally resist earning less for the same work. That’s why layoffs, rather than pay cuts, tend to be how the labor market adjusts to falling demand for a particular type of work.
The result is a rise in unemployment. A displaced accountant can’t simply change their LinkedIn profile to become an electrician. Workers must reset their expectations about what field they’ll enter and often bear retraining costs. Unlike job losses during a pandemic lockdown or a monetary-policy-induced recession, many of the jobs AI displaces won’t come back β they’re likely gone for good.
The Investor Side of the Ledger
Before white-collar professionals despair, there’s an upside. Total capital income in the substantial scenario is 18.9 percent above the no-AI baseline by 2030. As machines perform more tasks, the share of income going to capital rises. But that’s not simply a story of triumphant capitalists and impoverished workers, the analysis notes, because the two groups overlap heavily.
Workers β especially knowledge workers β are also capital owners, typically through retirement accounts and stock portfolios. The Federal Reserve’s 2022 Survey of Consumer Finances found that 78 percent of households in the 50th to 90th income percentiles owned stocks, either directly or indirectly. Among the top 10 percent, ownership reached 95 percent. That means many of the professionals whose jobs are exposed to AI also hold a financial interest in the companies likely to benefit from the technology.
Stronger profits can boost investment income and share values, cushioning the blow of weaker earnings for professional households. A larger retirement account reduces how much a family needs to save from each paycheck. For workers whose wages dip only slightly below their previous trajectory, that offset could prove decisive.
Anthropic’s paper doesn’t forecast stock prices or calculate these household offsets β that’s beyond its scope. But the analysis suggests that investment gains would likely cushion a significant portion of the pain from job losses and transition costs for established professionals. Younger workers with little invested would remain more exposed.
The Fed and the Fog of Transition
Of course, the Federal Reserve wouldn’t sit idly by during this transition. If productive capacity expands faster than spending, unemployment rises, and inflation weakens, the Fed could ease monetary policy to support demand. Standard central bank models suggest the Fed should recognize that faster productivity growth permits faster economic growth without necessarily reigniting inflation.
Over the long run, that doesn’t guarantee lower interest rates. A vigorous investment boom can increase demand for financing and push up the rate consistent with stable inflation. The Fed would have to judge which forces dominate. It can’t retrain an accountant, but it could help prevent weak spending from adding another layer of unemployment.
Blue-Collar Gains and Immigration Policy
The economic changes in Anthropic’s scenario carry significant implications for immigration policy. Better technology and more capital allow a slowly growing workforce to produce substantially more β meaning the U.S. may not need foreign workers to supplement a slowly growing, aging workforce to maintain growth.
More provocatively, mass immigration could undercut the wage gains projected for blue-collar workers. Those gains partly reflect their growing scarcity amid rising demand. Large inflows of competing workers could dilute that scarcity and blunt wage increases. Protecting those gains, the analysis argues, gives reason to restrain immigration even where hiring is strong.
The familiar push for more “skilled” immigration also looks different in this scenario. Most so-called skilled immigrants are cognitive workers β computer-related occupations accounted for 64 percent of approved H-1B petition beneficiaries in fiscal 2024. With AI handling many of the tasks now performed by cognitive workers, the demand for such imported labor may weaken.
Anthropic’s model may also underestimate the economy’s resourcefulness. As blue-collar wages rise, those workers will want financial advice, legal counsel, real estate agents, psychologists, and other white-collar services. As those services become cheaper, more households and businesses can afford them, expanding the market even as AI automates parts of the work.
History suggests economies abhor unused potential. That resilience, combined with the offsetting effects of capital gains and new job creation, may mean the disruption is more manageable than the headline numbers suggest. What’s clear is that the neat divide between AI’s winners and losers is blurrier than it first appears.
Source: www.breitbart.com β https://www.breitbart.com/politics/2026/09/09/breitbart-business-digest-the-blue-collar-boom-inside-anthropics-ai-model/