Editor’s Note: The following essay was previously published in The Washington Post on May 11, 2026.

　　For years, I kept a favorite cartoon in my desk and pulled it out to open the annual business-plan meeting at the unit I led. It showed a frazzled executive standing in front of a screen displaying his multiyear sales projections. The line ran straight horizontally, close and parallel to the x-axis, almost to the right edge, where it leaped steeply upward, next to a label that said, “Miracle happens here!”

　　No impulse is more human than wishfulness, the tendency to grasp at any straw that enables us to avert our eyes from difficult realities and put off facing them. Members of America’s national political class personify this failing, in their continuing practice of fiscal denialism. Even as the inexorable arithmetic piles up, those responsible for the nation’s economic future and national security fasten on imaginary miracles to justify a gross default of their duty of stewardship.

　　A decade ago, as the national debt surged toward the once unthinkable level of $20 trillion (now nearing $40 trillion), denialists took brief refuge in an alchemist fantasy that called itself Modern Monetary Theory. The notion that a nation could borrow without limit, forever, in its own fiat currency was quickly demolished by full-spectrum critiques, in venues ranging from the Cato Institute to the Review of Keynesian Economics. The experts weren’t really necessary; you could have just consulted the Journal of Common Sense, or maybe your grandparents.

　　MMT has mercifully disappeared from serious discussion, but the wishful impulse has not. Its latest comfort station is the claim that the productivity boost that artificial intelligence will bring to the economy will bail us out of our sinking boatload of debt. Stop worrying; “Miracle happens here!”

　　In our post-truth world, facts aren’t as stubborn as they used to be, but the most obstinate of all are the mathematical ones. They tell us not to rely on even the powerfully positive impact of these new technologies to spare us the radical adjustments that a generation of procrastination has now made inevitable.

　　This is no time to be touting miracle cures to justify further procrastination.

　　That isn’t to say that no help is on the way. The evidence is persuasive that AI and related advances are already boosting the economy in the most important way possible, by raising productivity. That’s the biggest reason that GDP is surprising on the upside while job growth remains tepid. Moreover, forecasts that this favorable windage will accelerate seem highly credible.

　　What’s not credible is the idea that even an AI-led productivity surge can suffice to offset our decades of dereliction. The Congressional Budget Office, the Federal Reserve, and other forecasters peg average future economic growth at a little under 2 percent. Assume a 70 percent boost from the AI revolution, to 3 percent or so, and it becomes possible to imagine our current debt level stabilizing, not improving but merely getting no worse.

　　But even this daydream requires far too many improbable breaks. Simulations conclude that the chances of growth of even 2.6 percent are less than 1 in 20. That’s without factoring in the possibility of a military crisis, a recession, another pandemic, or any other macroeconomic setback. AI revenue increases could be partially offset by new spending requirements, for energy infrastructure, for example.

　　The Penn Wharton Budget Model credits AI with a healthy 1.5 percent productivity and GDP increase over the next decade. That would result in deficit reduction of some $400 billion over those 10 years. Not chump change, but only a fraction of what would be required, given the tsunami of entitlement spending, driving trillions of added debt, making landfall over that period.

　　AI enthusiasts assure us that the beneficial impact will be even bigger. Let’s hope they’re right, although that would mean a bigger productivity surge than those brought by electricity or the Internet. Even if it happens, it cannot conceivably get here before the trust fund insolvencies start in the early 2030s. Kent Smetters, a Penn Wharton Budget Model scholar, states flatly that AI, however positive, isn’t “a magic bullet” and that the call is “not even close.”

　　Let’s stipulate that AI will be the transformative wonder that its inventors foresee; that the CBO and other forecasters have often tended to underestimate US economic growth, especially in environments of lightened regulation and taxation; and that the United States somehow sails through an unprecedented streak without a single costly exogenous blow.

　　It still ain’t enough.

　　Otto von Bismarck supposedly proclaimed, “There is a Providence that protects idiots, drunkards, children and the United States of America.” After decades reelecting a Congress whose spending behavior qualifies for the first three categories, we can’t count on providential salvation.

　　This is no time to be touting miracle cures to justify further procrastination. Until America acts to make major changes in laws on the books, the right side of our national business-plan chart will continue to show a sharp downward line and the label, “Big trouble happens here.”