Trump Accounts opened on the Fourth of July. Every American child born between 2025 and 2028 is now entitled to a thousand dollars from the government. The money goes into an index fund, a simple investment that owns a small piece of the five hundred biggest American companies.

My own reaction surprised me. I have little sympathy for this administration. Most of what it does offends my sense of how institutions should work.

But a good idea does not become a bad one because of who signs it. This is a good idea. It may be the most important ownership idea to come out of Washington in a generation. The principle underneath it is simple. Every citizen a shareholder.

Start with what the program actually does. Every eligible child gets the same thousand dollars. It goes into the same fund, on the same terms. The money stays locked until the child turns eighteen; and even then it does not turn into spending money. At eighteen the account converts into a retirement account in her name, and the usual retirement rules apply. This is a stake, not a check.

Families, relatives, and employers can together add up to five thousand dollars a year on top of the seed. Up to twenty-five hundred of that can come from an employer, tax free to the employee.

The idea did not come from the president. It came from Brad Gerstner, an investor who runs a fund called Altimeter Capital. He spent five years pushing it through both parties under the name Invest America.

Business is already lining up behind it. More than fifty companies have pledged to put money into the accounts of their employees' children, among them Dell, Uber, Goldman Sachs, Nvidia, BlackRock, JPMorgan Chase, Visa, Mastercard, IBM, Comcast, Charles Schwab, Robinhood, and Wells Fargo.

I should pause here, because skipping this would be dishonest. I have been hard on the technology industry these past two years. Many of its leaders moved too close to this administration. They lined up at the inauguration. They kissed the hand of the king. They rewrote company policies to match the political weather and stayed silent about things they knew were wrong, because access mattered more than principle.

I said so at the time and I stand by it. But this episode shows what that same access can do when it points somewhere useful. Gerstner and the executives behind these pledges spent their influence and their creativity on children who have no power at all. Proximity to power is a tool. This time it built something.

Your finance close runs in #finance. Stripe and QuickBooks reconciled, runway updated, posted Sunday night without you asking.

Engineering review lands in #eng. Viktor pulled the open PRs, left comments on auth-refactor, flagged a dependency blocking api-pagination.

Campaign brief lands in #growth: Meta CPA up 18%, recommendation to pause broad match, a draft landing page already deployed for the variant test.

You hired him on day zero. He lives in Slack and Microsoft Teams alongside your contractors and investors, connects to 3,000+ tools, pushes back when you ship something dumb.

"Viktor is now an integral team member, and after weeks of use we still feel we haven't uncovered the full potential." Patrick, Director, Yarra Web.

Get Started for Free

America has done this kind of thing before, in its best moments. In the 1860s the government gave ordinary families free land through the Homestead Act, because land was how families built wealth back then. After World War Two, the GI Bill paid for college, because by then education was how families moved up. The vehicle of mobility keeps changing. First it was land. Then it was the degree. Now it is equity, and soon, I will argue, it will be the machines themselves. The great American programs did not hand out cash. They handed out the vehicle.

Then the vehicle changed again, and this time nobody handed it out. For the past forty years, the thing that builds wealth in America has been the stock market, and most Americans are simply not in it.

The richest ten percent of households own about ninety-three percent of all stock market wealth. The bottom half of the country owns about one percent.

Two families can earn the same salary for thirty years and end up in different worlds. The difference is whether their money was invested or sitting still.

Our main answer to inequality has been taxes and transfers. The government taxes people who earn a lot and sends money to people who earn little. I want to explain, as plainly as I can, why that approach is incomplete and why this one adds what it lacks.

The first reason is mathematical. The French economist Thomas Piketty became famous for a simple observation: money invested in assets grows faster than wages do. His own answer was to tax wealth itself, and nearly every country that tried has retreated, because capital moves and enforcement is hard. Taxing wages to fund monthly checks loses the race by arithmetic. Taxing wealth loses it by politics. The route that remains is to own the thing that compounds.

The rich family's stock portfolio earns returns on top of returns, year after year. The poor family's monthly check does not compound. It gets spent, because it must.

The second reason is what transfers actually deliver. Let me be precise here, because the case against them is usually overstated. Transfers work. Children who grew up with food stamps and Medicaid earn more and live healthier as adults; the research on this is solid. What transfers cannot do is create an asset. A family can receive support for thirty years and own nothing at the end. Nothing to pass to their children. Nothing that grows while they sleep. The help changes lives, and it still never changes what the family holds.

The third reason is friction. Means-tested programs require an industry of eligibility rules, caseworkers, audits, and appeals. The money itself mostly gets through; the process taxes the time and the dignity of the people it serves, and benefits vanish at cliffs that punish a raise. Compare that to this program. The law caps the fund's fees at a dollar a year on every thousand, and enrollment for newborns is being folded into birth registration itself. One system employs a bureaucracy to means-test. The other opens the same account for everyone, almost for free.

The fourth reason is political. Means-tested programs split the country into two groups: the people who pay and the people who receive. Both sides end up resenting it. The payers fight the taxes. The receivers carry the stigma. So the programs get cut and restored and cut again with every election. Universal programs work differently. Social Security survived every attack not because it was generous but because everyone was in it. Nobody will campaign against a whole generation of shareholders. This program builds its own protection, one child at a time.

Let me be fair to the other side of the argument, because it is easy to overstate mine. A locked account does nothing for a family that cannot pay rent this month. Ownership does not replace transfers. It adds what transfers could never provide. One keeps people fed today. The other changes what their children hold tomorrow.

The critics also say the money is too small, and on the numbers they are right. A thousand dollars growing for eighteen years becomes roughly six thousand dollars. Nobody retires on that; though left untouched, the government's own projections say the same seed reaches nearly a quarter million dollars by age fifty-five. But the balance is not the point. The position is the point.

And the position is stricter than most coverage admits. At eighteen the account does not pay out; it becomes a retirement account, which the young adult cannot raid without penalty. She cannot cash it for a car. What she owns at eighteen is not a windfall. It is a stake she will carry for decades, growing through every job she takes and every crash she survives. A child who owns a piece of the American economy relates to it differently than a child who owns nothing. She is a shareholder, not a spectator.

The French sociologist Pierre Bourdieu spent his career showing that wealth is never just money. It is confidence, belonging, a sense of where you stand. The deepest divide in America today is between people who feel they own a piece of the economy and people who feel the economy is something that happens to them.

Now the honest criticism, because praise without it is worthless. The design has real flaws. Rich families can fund the full five thousand dollars every year. Poor families can add nothing. Run the math and the gap is brutal: the seed alone grows to about six thousand dollars by eighteen, while a fully funded account reaches more than a quarter million. So the accounts will grow furthest for the children who need them least. A program sold as an equalizer that doubles, in practice, as an engine of divergence.

Senator Cory Booker and the economist Darrick Hamilton proposed a better version years ago, called baby bonds, in which the poorest children receive the most: a thousand dollars at birth, up to two thousand more each year for low-income families, close to fifty thousand dollars by eighteen for a child born into poverty. Their design was fairer. It also never passed.

Britain offers a warning too. It seeded an account for every child born between 2002 and 2011, then abandoned the program and stopped reminding anyone it existed. By 2023, more than four in ten of the young adults whose accounts had matured had never touched the money.

Today more than 750,000 accounts, worth over 1.6 billion pounds, still sit unclaimed, and they belong disproportionately to the poorest families. And Britain's accounts paid out in cash at eighteen. If free money can be forgotten, a locked retirement stake can be forgotten more easily.

America's version is already showing the same symptom, at the front end. The thousand dollars is not automatic; a parent has to claim it. Six million accounts were opened in the program's first days, 1.4 million of them in line for the seed, out of tens of millions of eligible children.

Folding enrollment into birth registration will close the gap for newborns. Until it does, the children least likely to be enrolled are exactly the children the program exists for. The lesson from Britain is simple. Opening the account is easy. Keeping families connected to it for eighteen years is the real work.

There is, however, a repair kit inside the statute itself, and almost nobody has noticed it. States, cities, and charities can contribute to these accounts without limit, outside the five-thousand-dollar cap, as long as they give equally to a defined class of children. The Michael and Susan Dell Foundation has already committed 6.25 billion dollars to put 250 dollars into the account of every American child under eleven.

Flat gifts are a start. The real opportunity is progressive. A state, a city, or a foundation could top up the accounts of its poorest children every year, building Booker's baby bonds inside the architecture that actually passed. The equalizing version of this program does not need new legislation. It needs funders who understand what the law already allows.

Most importantly, I worry about how the money will be managed. The law has guardrails: the funds must track a diversified index of American companies, with fees capped at a tenth of a percent. The default today is a State Street fund tracking the S&P 500, administered by Bank of New York Mellon, with a family app built in partnership with Robinhood. So far, so boring, which is exactly what a child's account should be.

But the Treasury decides what counts as a qualified index, and definitions can stretch. Given the track record of the president and his family, a slow drift toward affiliated ventures is not a far-fetched scenario. If it happens, we will lose a rare chance at genuinely good government policy. The guardrail worth watching is not the fee cap. It is the definition.

The reason all of this matters right now is AI. Here is where I part ways with the tech bros: I do not believe this wave nets out to more jobs. I believe it will eliminate more work than it creates, deepen inequality, and consolidate the power of the companies that own the systems. I may be wrong about the timing. I do not think I am wrong about the direction. This program is one of the few tools that scales against that risk.

Over the past year I have noticed the same thing in every conversation with investors. Everyone wants exposure to AI. Almost nobody asks who will end up owning the gains. The economist Daron Acemoglu has spent years on that question, and his answer is uncomfortable: the profits from automation flow to whoever owns and directs the machines. Wages get some of it. Ownership gets most of it.

I have watched this up close, building with AI for a decade. It is making routine knowledge work cheaper by the month, and the value is flowing to the firms and people who direct the machines.

Left alone, AI will produce enormous wealth and deliver it to the people who already own everything. The stock market is where those gains will land.

So the only equalizer that scales for the AI age is ownership itself, given to everyone, starting at birth. Every citizen a shareholder, from the first day of life. A stake in the index is a stake in the machines.

Even Sam Altman, who funded the largest basic income experiment in history, a thousand dollars a month to three thousand Americans for three years, argued in his essay "Moore's Law for Everything" for giving every citizen equity in the wealth the machines create, rather than monthly checks. He is right to point in that direction.

There is a selfish case for this too, and the new AI rich should hear it. A system that delivers all the gains to a few thousand owners will not survive politically. People who own nothing eventually turn against the machine, and against the people who run it.

There is a business point underneath as well. Companies need customers, and machines do not buy anything. If AI takes income from workers while its profits sit with a few, the market for everything shrinks.

Sharing the ownership keeps money flowing through the people who spend it. Broad ownership is good for the AI rich themselves. Their wealth is only durable in a system that most people still want to defend.

I still believe that skills, judgment, and agency remain the core of any individual answer to this era, because they are built, not given. But every capacity needs a base to grow from. For the first time, that base is being issued at birth, equally, to every child who is signed up for it.

That is the American dream, restated for the age of AI. Not a promise that hard work will make you rich. A guarantee that when the machines produce wealth, some of it is already yours, whoever your parents are. The design needs fixing. The fix is already legal. The direction is right.