Imagine one in five Americans out of work, unemployment jumping from 3% to 18% in just two years as machine-replaced Americans went looking for work. Soup kitchens lining the streets of every major city. Whole states even harder hit: over a third of New York workers unemployed, nearly half in Michigan.
That was America in 1893 — not the Great Depression, but the one that came before it. The proximate triggers were financial: a railroad bubble that burst, a run on gold reserves, bank failures cascading across the country. But the kindling had been laid over decades.
The violent lurch from an agricultural economy to an industrial one had uprooted millions: Workers flooded cities unprepared to house and care for them, and the migration left rural America in a prolonged slump long before the panic hit. Frustrated, unemployed Americans marched on Washington for the first time ever, demanding a jobs bill. Technological disruption didn’t cause the Depression directly, but established the conditions that made catastrophe, when it came, so devastating and so hard to escape.
With so many now predicting yet another Depression-level event, the social and economic devastation of the “forgotten depression” is eerily instructive. We are again at the doorsteps of a transition we do not fully understand, but this time, a technological transformation is throwing gas on a fire that is already alight.
Because at the dawn of the AI era, the economic ladder is already broken. If you were born in 1940, you had a better than 90 percent chance of out-earning your parents. If you were born in 1985, the odds dropped to about 50-50. Today, nearly half of American households can’t make ends meet.
People aren’t just imagining that they’re falling behind, they actually are.
The good news is we know far more about what works than we did the last time around. From Benjamin Bloom’s two-sigma tutoring research to Raj Chetty’s work mapping both the drivers of economic mobility and the stark disparities in who actually achieves it, we have an increasingly clear picture of not just the causes, but the environments, experiences, and relationships that can enable Americans to beat the economic odds.
We can now trace roughly two-thirds of the decline in American mobility to public and private policy choices that decoupled the country’s growth and prosperity from that of workers. Understanding the implications of those decisions affords us the opportunity to act differently this time.
What’s even more encouraging is that the technologies that risk leaving millions of Americans behind may hold the potential to transform the institutions where opportunity will be won or lost.
That is the idea behind what Anthropic co-founder Daniela Amodei has dubbed “Claude Corps,” which aims to embed a thousand AI-fluent fellows within hundreds of organizations that hold civil society together: food banks, legal aid offices, veterans’ networks, refugee resettlement agencies.
To prepare corps members, Anthropic is teaming up with CodePath, a nonprofit founded by tech entrepreneurs that, according to their last annual report, has quietly helped to prepare more than forty thousand alumni (forty percent of whom are from households earning under $50,000) to compete for and win coveted roles at the most competitive engineering organizations in the country. Modeled after the pathway from low-income to high that CodePath’s cofounders each took themselves, Codepath is now teaming up with Anthropic engineers to create entirely new approaches to training that can equip young Americans with the skills they need to not just thrive, but make an impact.
A motivated 22-year-old with the right training can now develop solutions inside an organization that would have required a team of consultants five years ago. Multiply that by a thousand fellows inside institutions touching millions of Americans, and the question becomes not just how the fellows’ lives will improve or even how much more the institutions they serve can deliver to their communities, but how the social sector will be transformed.
For decades, institutions that benefit the public have been last in line for new technology, not because the people inside them are slow, but because they’ve never had the staff, the budget, or the in-house expertise to move first. We can now put AI in the hands of the people and organizations that enable us to mitigate or recover from literal and figurative tsunamis. From education and job training to community lending to housing to mental health supports to making public benefits easier to get, new tools in the right hands will make a vital difference. Claude Corps is a down payment on that possibility.
Staving off a devastating economic shock will require a sweeping effort on multiple fronts to prepare people for the jobs that will exist in a fast-changing economy, to lower the cost of a decent existence, and to reinvent safety nets built for a different era. It will require investment in new approaches to education and training.
That should grab funders and policymakers. We have built moonshots before. The WPA in the Depression. The Marshall Plan. The space program. The polio vaccine. Each required the discipline to face brutal facts honestly and the faith that we could prevail in the end. We are in that kind of moment now.
Today, that work has not just a new imperative, but new instruments. The question is whether we can race fast enough to prepare more Americans to harness their potential, and embed them within the institutions and the places where opportunity is hardest to build.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
