AI's Job Disruption Is Coming. We're Nowhere Near Ready.

A new opinion piece in The Hill raises a question that should have been urgent years ago: What do we do when AI eliminates millions of jobs?

The article proposes "fiscal insurance", a combination of worker adjustment assistance, wage insurance, and expanded capital ownership. Those aren't radical ideas. They're what other wealthy democracies already do. What's radical is that we don't.

Why This Matters Right Now

Automation isn't new. But the speed and scale of AI disruption is different. A cashier, a truck driver, a claims processor, a junior attorney, all of them could see their work fundamentally changed in the next five to ten years. Not in the distant future. Soon.

In the past, when factories closed or industries shifted, workers who lost jobs faced a brutal choice: retrain yourself (on your own dime, in your own time), move to find work, or fall behind. Many did fall behind. Their wages never recovered. Their kids had fewer options. Entire towns hollowed out.

We know this happens. We've watched it happen. And we've done almost nothing to stop it from happening again.

The Real Problem: We Spend Almost Nothing on Retraining

Here's the gap that matters: Denmark spends 2% of its GDP on active labor market policies, programs that help displaced workers retrain, find new jobs, and transition with dignity. The United States spends 0.1%. That's not a typo. One tenth of one percent.

The result? Only 24% of workers displaced by automation receive any retraining at all. The rest are left to figure it out alone.

This isn't just unfair. It's economically stupid. A worker who retrains stays in the economy, keeps spending, keeps paying taxes. A worker who gives up stops doing all three. Communities that lose too many workers lose tax revenue, which means worse schools, worse infrastructure, worse futures for kids born there.

Why "Adjustment Assistance" Is a Fancy Way of Saying "Not Enough"

The U.S. has Trade Adjustment Assistance programs that are supposed to help workers displaced by trade deals. They're underfunded, hard to access, and don't cover most displaced workers anyway. Extending them to cover AI displacement would be a start, but only a start.

Real retraining takes time, money, and support. Community colleges that actually work. Wages while you're training. Help with childcare, transportation, housing. Job placement after you're done. Most of our current programs offer fragments of that, and only to workers who happen to know the programs exist and can jump through the hoops.

A serious plan wouldn't be insurance. It would be a guarantee: if your job is automated, here's what you get. Retraining that actually leads to work. Income support while you train. Healthcare that doesn't disappear when your job does. A path, not a prayer.