An open reply to Peter Diamandis on what the 80s and 90s actually taught us

Peter,

You asked a good question on X, and it deserves a serious answer rather than a nostalgic one. You asked whether what made the 1980s and 1990s successful can still be applied today, with all the changes exponential technology is now bringing. It is the right question asked at the right moment, and I want to answer it in a way that takes your own commitments seriously, because I suspect I share more of them than my conclusion will suggest.

I am not going to argue against abundance. That is a losing argument, and worse, it is a wrong one. Technology genuinely does demonetize and democratize. The cost of computing, communication, and information has collapsed inside a single lifetime. On that, you are right, and the record is on your side.

But I want to draw a line that your question quietly steps over. Abundance is not access. An economy can manufacture staggering wealth and, at the same time, narrow the number of people who can actually reach it. The 80s and 90s were not the finished success we now remember. They were the opening phase of an experiment whose long-term results we can finally read. And what they teach, read honestly, is not a model to copy. It is a warning to correct.

The lesson was never GDP

The real measure of economic success is not GDP, corporate profits, or the level of the stock market. Those are numerators. They tell you how large the pie has grown. They tell you nothing about how many hands can reach it. The measure that matters is the denominator: how broadly the practical ability to earn, own, and benefit actually extends. Can ordinary work still buy a path to housing, healthcare, education, savings, retirement, and family formation? When the answer erodes, people are quietly reclassified. They move from participants into spectators, living inside a wealthy economy they increasingly cannot afford to fully join.

The 80s and 90s optimized the numerator with real skill. They also began, deliberately, to loosen the mechanisms that let ordinary people share in what they helped create: rising wages, bargaining power, affordable land, and a realistic path to ownership. This was not an accident of the business cycle. It was a reorientation of the firm itself around a single stakeholder, a turn I trace to Milton Friedman's 1970 essay and the Powell memorandum of 1971. Shareholder primacy did exactly what it was designed to do. It is uncharitable to pretend otherwise, and unnecessary.

The evidence is not rhetorical

Here is where the argument stops being a matter of taste. The clearest signature of the era is the gap that opened between what workers produced and what they were paid.

For most of the postwar period, productivity and pay rose together, roughly in lockstep. Then, around 1979, they split, and the distance kept widening through the very decades now remembered as the success story. The Economic Policy Institute's long-running analysis of federal data puts it starkly: from 1979 to 2025, net productivity of the American economy grew by roughly 90 percent, while the hourly compensation of a typical worker grew by only about a third of that. Output per worker climbed. The paycheck did not follow. That divergence is not a footnote to the era. It is the era's defining feature, and it tells you exactly where the gains went.

The RAND Corporation put a price on the same phenomenon from the other direction. In their study of income from 1975 to 2018, Carter Price and Kathryn Edwards asked a simple counterfactual question: what if incomes below the 90th percentile had merely kept pace with the growth of the economy, as they did in the first postwar decades? The answer was that the bottom 90 percent would have earned about 2.5 trillion dollars more in 2018 alone, and roughly 47 trillion dollars more, cumulatively, across the whole period. Price has since carried that measurement forward to 2023. In my own work I carried the same counterfactual through 2025, and the gap has only widened. This is not a leak in an otherwise sound system. It is a transfer, running steadily upward, for four decades and counting.

Now hold that against the abundance story, because both things are true at once, and the tension is the whole point. Many things did get radically cheaper: computing, communication, entertainment, most manufactured goods. But the goods that actually constitute a secure life moved the other way. Housing, healthcare, and education have risen far faster than income for a generation. So the economy demonetized the shelf of consumer goods while the shelf of middle-class security kept getting more expensive. Cheaper televisions do not build equity. A demonetized smartphone does not house a family. When the cost of participating rises faster than the ability to pay for it, abundance and exclusion grow together, in the same economy, at the same time.

Why AI closes the old exit

This is where your question sharpens into something more urgent than nostalgia, and where I think you and I actually agree about the stakes.

In every version of the market economy we have ever run, including the deeply unequal ones, wages were the transmission belt. Labor was the mechanism that carried productivity into ordinary households. Even when the belt slipped, as the figures above show it did, it still connected most people to a claim on what the economy produced. You worked, and through work you held a stake.

Artificial intelligence threatens to cut that belt. If human labor becomes less necessary while the productive assets stay concentrated, we can generate extraordinary wealth and, in the same motion, sever most people from the system that generates it. Concentration stops being a policy failure and becomes the default setting of the machine. The wealth still gets made. Fewer and fewer people have any structural reason to receive a share of it.

That is why returning to the 80s and 90s model is not merely undesirable. It is becoming mechanically impossible. That model assumed a labor market that distributed purchasing power widely enough to keep the whole thing running. Remove the assumption, and the model does not just become unfair. It becomes unstable. An economy of extraordinary supply and collapsing demand is not a triumph. It is a contradiction waiting to resolve itself, usually badly.

The answer is not redistribution. It is Ordered Liberty.

So what replaces the old model without throwing away what was genuinely good in it? Competition, innovation, open markets, and entrepreneurship are not the problem, and I have no interest in trading them for managed stagnation. The task is narrower and harder: to rebuild the link between productivity, ownership, and participation without breaking the engine that creates the wealth in the first place.

I call the framework that does this Ordered Liberty, and its organizing instinct is a single reversal. The 20th-century reflex was to let wealth concentrate and then argue, endlessly, about how much of it to tax back. That is redistribution, and it is a permanent rearguard action fought after the fact. Ordered Liberty prefers predistribution: widening the base of who owns the productive assets before the returns pool at the top, rather than only taxing them once they have. You do not need to claw back a fortune you never allowed to become dangerously concentrated in the first place.

Adam Smith drew the distinction this rests on, long before anyone needed reminding of it: the difference between the maker and the taker, between income earned by producing and income extracted by owning a chokepoint. An economy is healthy to the degree that it rewards the first and disciplines the second. Ours has spent forty years doing the reverse. Ordered Liberty is, in that sense, less a revolution than a reset: a restoration of the older and frankly more American understanding that markets are supposed to reward creation, not capture.

That principle becomes a concrete policy stack. A few of its load-bearing pieces, each aimed directly at a failure named above:

A land value tax, in the tradition of Henry George, aimed at the single largest driver of the cost-of-living crisis. Land does not respond to demonetizing technology, because they are not making more of it, and its rising value is overwhelmingly the community's creation, not the owner's. Taxing that unearned rent, rather than labor and enterprise, attacks the housing problem at its root instead of subsidizing around it.

A Homestead Ladder and a Housing and Down-payment Savings Account, to rebuild the ownership on-ramp that wages used to provide. If labor is going to be a weaker path to a stake in the economy, then ownership has to become a wider one, and it has to start early and low on the ladder rather than being reserved for those who already own.

Postal banking and a universal basic income, understood not as charity but as a floor: the demand-side counterweight to an economy where labor income can no longer be assumed. A dividend on shared productivity, closer to a citizen's rightful share than to a handout.

And at the top of the structure, the piece that speaks most directly to your world: a graduated ladder of corporate duty. The privileges a firm enjoys are granted by the public through its charter, and under the older concession theory of the corporation, duties should scale with those privileges. A sole proprietor owes little. A billion-dollar firm owes more. And a firm large enough to be a systemic entity, measured by its share of GDP, its market concentration, and, critically, its command of AI training compute relative to the moving frontier, owes the strongest duties of all, with the nation itself recognized as a stakeholder.

I want to be plain about why that last tier exists, because it is your field that makes it necessary. Under current law, an AI-directed enterprise can incorporate in the most permissive jurisdiction available and amass nation-exceeding power with no federal restraint. "Too big to fail" becomes "too big to answer to anyone." The graduated ladder is designed so that greater scale only ever adds obligation, never escapes it. This is not hostility to scale. It is the refusal to let scale become sovereignty.

Back to your question

So, can what made the 80s and 90s successful still be applied today? The spirit of it, yes, and it must: the competition, the invention, the entrepreneurial nerve. Those are the parts worth carrying forward at almost any cost. But the model, the specific machine that let productivity and wages come apart and called the result a success, cannot be copied into an age of exponential technology. It was already leaking when it was new. Under AI, it does not leak. It breaks.

The 80s and 90s were not a finished success to reproduce. They were a first draft. The technology now arriving will decide which of two things we build on top of that draft: an economy that widens the circle of people who create, own, and benefit, or a far richer theater built for a steadily growing audience of spectators.

An economy is not successful merely because it creates more wealth. It is successful when more people have the practical ability to participate in creating, owning, exchanging, and benefiting from that wealth. That is the whole of the disagreement, and, I suspect, the whole of the work ahead of us both.

With respect,

Darin Lawson Hosking The Moral Algorithm

This letter draws on the argument developed at length in From Moral Algorithm to Ordered Liberty: A Citizen's Pathway. Figures on the productivity-pay divergence are from the Economic Policy Institute; the income counterfactual is from the RAND Corporation's Price and Edwards study, extended through 2025 in the author's own analysis.