By Lauren Weber
Inside a glass-encased building in downtown Washington, D.C., this past week, a group of experts hunkered down to strategize how to avoid the AI apocalypse that many doomsayers predict.
They were some 40 leading thinkers in economics, technology and public policy, and they had come to the Peterson Institute for International Economics to contemplate a vision of U.S. society in 2030. Provided with workbooks, easel pads and Sharpies, they arrived ready to game out responses to some of AI's biggest risks -- to jobs, the economy, political stability.
One conclusion they reached: Political polarization will make ambitious changes difficult. And yet inaction will be far more costly.
So "we start to put down smart bets on some scenarios," said Rep. George Whitesides (D., Calif.), who stopped by for a brief Q&A during lunch.
The event was arranged by Windfall Trust, a nonprofit devoted to heading off AI's economic disruption by developing potential policy responses. It has been running scenario-planning exercises like this in cities around the world to catalyze proactive strategies for confronting an AI-driven future.
The group chose to focus on AI's economic implications rather than safety risks such as cyberattacks and bioweapons. "Those are likely to have political consequences sooner as people's anxiety grows about how AI will disrupt the economy and their jobs and their children's jobs," said Adrian Brown, Windfall Trust's chief executive.
It doesn't help that even some of the most powerful people in AI can't decide what it will unleash on humanity. OpenAI's Sam Altman and Anthropic's Dario Amodei have softened earlier predictions of a white-collar job apocalypse, albeit ahead of recently announced plans for initial public offerings.
The public isn't so reassured. Only 17% of Americans say AI will have an overall positive effect on the U.S. over the next 20 years, according to a March survey by the Pew Research Center.
The scenario the Washington participants were given, called "Paper Prosperity," presented an economy on fire by some metrics. AI's speed and efficiency had nearly doubled the gross domestic product and labor productivity growth rates. The S&P 500 was soaring, too.
But underneath was a simmering social and economic crisis: Unemployment had risen only slightl, but underemployment had leapt from 8% to 14%. It indicated that millions more people were in part-time jobs, doing gig work or in roles they were overqualified for because AI had triggered layoffs of college-educated workers.
"Underemployment, gig work and precarity -- long features of working-class American life -- were finally reaching the upper tiers of the middle class," read the scenario.
During the morning session, each table discussed what this world might look like beyond the numbers. The tone was collegial and open-minded, even as experts sometimes questioned one another's assumptions and predictions.
They reached consensus on a few things: Political unrest would certainly rise. Generational divides would grow wider, and fertility would decline because of young people's pessimism. The social contract -- go to school, work hard, prosper -- would fray.
Wages for some jobs would increase, at least temporarily, especially for physical roles like welding, nursing and plumbing, some suggested. But as white-collar workers and others pour into job training for those roles -- a trend already happening -- a surge in supply for those now-scarce workers could drive wages down.
There might be some upsides in this future. Healthcare and education costs might fall because much of that expertise can be delivered cheaper and easier with AI-augmented tools, said Neil Thompson, director of MIT FutureTech, a research group that studies technology trends. Americans might be healthier and live longer lives thanks to AI-driven advancements in medicine. And underemployed workers might wind up with more hours of free time for creative pursuits, or what one facilitator jokingly called "the crochet economy."
What policies can be imagined and enacted now, the facilitators asked, to avoid the toxic outcomes described in the scenario?
Worker reskilling came up over and over, with several caveats. The U.S. doesn't have a great track record with government-administered reskilling programs, like those designed to help workers displaced by industrial offshoring.
Also, we don't yet know what new tasks and occupations will be created because of AI, or which jobs will simply be augmented, said Harry Holzer, a labor economist at Georgetown University.
There was much discussion about redistributing the windfall that AI will create, through mechanisms such as a universal basic income, taxes on AI companies and their shareholders, or a version of what Sen. Bernie Sanders (I., Vt.) proposed this month: putting ownership of half of AI companies' stock into a sovereign-wealth fund.
Other suggestions included boosting the social safety net with universal health insurance, job guarantees, wage subsidies for laid-off workers who fell into lower-paying jobs and huge investment in jobs in re-envisioned, high-quality child care and eldercare, fields that rely on human skills like care and empathy.
Over the course of the day, it became clear that one of the biggest challenges to proactively addressing AI is government dysfunction and the resulting public distrust of elected leaders.
"Congress is slow-moving and technophobic," said Whitesides. The congressman, who is a former chief of staff at NASA and the first CEO of Virgin Galactic, said he is particularly concerned about risks like biological threats and digital safety, especially for kids and teenagers.
"How do we act as policymakers in a world where we don't know what the future holds?" he said.
He mentioned a bipartisan proposal called the Great American AI Act, released this month by Reps. Jay Obernolte (R., Calif.) and Lori Trahan (D., Mass.). The draft legislation calls for everything from AI-specific workforce data to requiring AI labs to disclose their assessments of catastrophic risks.
The U.S. can also look to other countries, said Windfall Trust's Brown. Last week, the U.K. launched the AI Economics Institute, a government group whose research will inform public policy.
"It's heartening that some governments are taking steps to put serious attention on these issues," he said.
Write to Lauren Weber at Lauren.Weber@wsj.com
(END) Dow Jones Newswires
June 13, 2026 14:00 ET (18:00 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments