Assetlessness Is Driving America Toward State Dependence

David Friedberg sees the United States moving toward socialism because too many people have no credible route to economic security outside the state. His central distinction is between people whose lives depend on the next paycheck and people who own assets that generate income and widen their choices.

He puts the immediate problem in terms of affordability. Friedberg cites a figure of 63% of Americans living paycheck to paycheck, without enough savings for a household emergency. Costs have risen faster than wages, he says, and the resulting strain has become a politics of class conflict: people who cannot cover bills watch wealth accumulate elsewhere and conclude that public support is their only available answer.

That is why he calls socialism inevitable under current conditions. His claim is not that Americans are necessarily drawn to a political theory. It is that people who feel locked out of housing, healthcare, childcare, and retirement will support the candidate offering a check, service, or guarantee.

Friedberg connects that condition to American decline. Asked whether the United States is an empire in decline, he answers yes in terms of the dollar’s power and the prosperity of the average American. But he does not treat decline as fixed. America still has, in his telling, an unusual record of allowing people to arrive with debt or little inherited advantage and accumulate independence. The failure is that millions of people no longer believe that route is open to them.

His preferred measure of success is therefore neither homeownership nor employment alone. It is the rate at which people move “from labor to capital.” Labor means relying on wages to meet monthly obligations. Capital means owning enough assets that the assets themselves produce income and give a person more freedom over how to spend time.

Friedberg uses his own history as an illustration. He graduated with $19,000 in debt, paid it down through work, and later gained assets through Google’s IPO. The important change, he says, was not simply a higher salary but the possession of savings and ownership that could produce returns. At a simplified 10% annual return, $100,000 produces $10,000 and $1 million produces $100,000. The figures are illustrative; his point is that assets can reduce dependence on the next paycheck.

He proposes that the country should aim to move 2% of Americans across that threshold each year. The relevant question in his framework is how many people could retire if they chose to—not merely how many have jobs, receive benefits, or own a home with debt attached.

Friedberg estimates that government employees, contractors, pensioners, and Social Security recipients together account for roughly half the U.S. population. For him, that is already approaching de facto socialism: once a large share of voters receives a government check, reducing that check becomes politically difficult. Politicians promise more, recipients have reason to support the promise, and government grows into a larger part of the economy.

He traces the pressure behind that outcome to debt, inflation, and the rising cost of essentials. A chart displayed during the discussion placed U.S. national debt at $0.37 trillion in 1970, $27.75 trillion in 2020, and $40.04 trillion in 2028. Friedberg describes the mechanism as government borrowing, Federal Reserve purchases of debt, and new dollars entering the banking system. He considers the trajectory unsustainable.

But the political conclusion matters more to him than the monetary mechanics. If people cannot build assets of their own, their demand for public support will continue. In his telling, a government that responds to unaffordability with larger transfers and public provision does not solve the underlying problem; it entrenches it.

Friedberg’s critique is not simply anti-tax or anti-government. He describes the problem as capitalism combined with policies that allow existing capital to compound more easily than workers can accumulate capital in the first place.

A worker is taxed through wages as those wages are earned. Owners of appreciating assets can hold them for years without realizing a taxable gain. If they already have substantial capital, Friedberg argues, compounding accelerates the gap between them and people who have not yet crossed from labor into ownership.

His proposed remedy is to tax capital gains at a rate equivalent to labor income, potentially lowering tax on wages while raising taxes on gains. He also wants borrowing against appreciated assets to be a taxable event.

Steven Bartlett offers a concrete example from his own experience. After his company went public, a European bank offered to lend against his shares without requiring a sale. Against a million dollars of Facebook stock, Bartlett says, the bank offered hundreds of thousands of dollars in cash without a taxable sale. The borrower could spend the money or buy other assets while continuing to hold the original shares.

Friedberg treats that as a basic defect in the current system. Borrowing against an asset to fund consumption should be treated as a transaction, he says, because it gives someone access to the economic value of property without triggering the tax that would accompany an outright sale. In his analysis, that arrangement lets people with substantial assets continue acquiring capital while wage earners are taxed at the point they receive income.

What he rejects is a recurring tax on wealth simply because an asset exists. He calls that an asset seizure rather than tax reform. His objection is both philosophical and constitutional: private property, in his account, is a foundational American right, and an annual claim on property before sale or exchange gives government a power it should not possess.

He also argues that a wealth tax fails the arithmetic test. The figures displayed in the discussion put total U.S. household net worth at roughly $178 trillion to $183 trillion, while the bottom half of Americans collectively held $6 trillion.

Even confiscating all billionaire wealth, Friedberg says, would fund about one year of federal spending. The problem, as he sees it, is not a shortage of one-time assets available to seize. It is persistent spending and a tax code that favors unrealized ownership over work.

Bartlett raises the possibility that billionaires’ political views may reflect their interest in remaining wealthy. Friedberg’s answer is to distinguish ownership from use: tax people when they sell, spend, or borrow against assets, he says, but do not give government an annual claim on assets they continue to own.

That distinction drives his political fear. Friedberg believes that a wealth tax creates a widening constituency for taking assets: first billionaires, then people worth tens of millions, then millionaires, and eventually anyone with property. He says he has seen billionaires move or plan to move out of California because they anticipate policies of that kind. In his view, their departure would mean not only lost tax revenue but fewer company builders, investors, employers, and spenders in the local economy.

He cites France as an example of a wealth tax driving capital away and reducing total tax revenue. His preferred reforms are less dramatic but structural: equalize the treatment of labor income and capital gains, tax asset-backed borrowing, and reduce federal spending. Those changes, he says, would make it easier for people to retain income and accumulate assets rather than depend on public guarantees.

Homeownership and student debt belong to the same diagnosis. Friedberg treats the conventional American route to stability—college, mortgage, retirement guarantee—as a set of promises that no longer works reliably for people without existing wealth.

Owning a house with substantial debt, insurance, property taxes, and maintenance obligations does not necessarily make a person financially free, he argues. It may simply make monthly dependence on income more expensive. He does not deny that homeowners have benefited from rising property prices. He regards those gains as evidence of the deeper problem: when a house bought 20 years ago appreciates dramatically, the same house becomes harder for the next buyer to afford.

The chart shown during the exchange marked the S&P/Case-Shiller home-price index at 25 in 1976, 184 at the 2006 bubble peak, 134 at the 2012 trough, and 335 at a 2026 historic high. Friedberg’s interpretation is that housing has become a protected vehicle for middle-class wealth accumulation whose price must keep rising for current owners to feel secure.

He recommends broader ownership of liquid financial assets instead. Depending on the market, he says, many people would have been better off over the past 30 years buying the S&P 500 than treating a primary residence as their main investment. A stock-market holding does not require property taxes, insurance, repairs, or maintenance, and it avoids placing most household capital in one asset.

This is not an argument that every individual should sell a house or avoid buying one. It is an objection to the claim that purchasing a home is itself the American dream. The dream, as Friedberg defines it, is having enough capital to choose how to spend one’s time.

He applies the same critique to federal student loans. Government-backed loans, he says, do not discriminate sufficiently among institutions, fields of study, and borrowers’ prospects. That gives colleges access to a large stream of money without forcing them to demonstrate that tuition produces useful educational or employment outcomes.

Friedberg cites a sharp expansion in university administration over the past decades. The on-screen note attributed to the National Center for Education Statistics says administration has grown 156% since the 1970s, faster than faculty and enrollment, and now accounts for roughly one-third of total staff. Friedberg’s broader contention is that federal loan funding allows institutions to expand costs without a normal market constraint.

He wants to end or phase out the federal student-loan program and let private lenders underwrite loans. A private lender would have an incentive to assess the quality of a school, the likely value of a degree, and a borrower’s ability to repay. Programs that did not lead to viable work might no longer support $200,000 in debt; stronger programs would still obtain financing. Scholarships, he says, could continue to serve students who need them.

Steven Bartlett resists a version of Friedberg’s argument that credits individual agency while minimizing the role of public foundations. He describes his own move from southern Africa to the United Kingdom and the support structure he had not fully recognized at the time: free schooling, healthcare, laws, infrastructure, internet access, and the possibility of Jobseeker’s Allowance if his early ventures failed.

The existence of that floor, Bartlett suggests, may have made risk-taking psychologically possible. He could leave university, start a business, and tolerate the prospect of failure partly because the consequences were not limitless.

David Friedberg agrees. Basic stability, infrastructure, and a social foundation are a great role for government to play. He also says no one should be without healthcare in a civil society, though he does not offer a complete model for providing it without allowing costs to rise sharply. He thinks any workable healthcare arrangement would have to be multi-tiered.

The disagreement is about the point at which a floor becomes a ceiling. Friedberg fears a system in which the state becomes the default employer and provider of permanent income. Bartlett’s account implies that public institutions can expand agency by protecting people from ruin. Friedberg’s concern is that public dependence can instead narrow agency, especially if people are taught that their only secure path is government employment or government support.

He does not believe everyone should become a founder, creator, or influencer. Bartlett calls that opposite extreme a mistake, and Friedberg agrees. The objective is not entrepreneurship for its own sake. It is the capacity to create something another person values—within a company, through a trade, in public service, or by starting a business.

Friedberg points to occupational licensing as an example of avoidable friction. Someone opening a barbershop may need to pay for schooling, obtain cosmetology credentials, navigate health inspections, and meet other requirements before beginning to work. He sees such layers as barriers to ordinary forms of agency.

His account of China follows the same logic. Friedberg does not describe China as a pure free market or deny the state’s large role. But he believes a major source of its growth came from allowing individuals to trade, produce, and identify what other people would pay for—not simply from central planning.

That faith in individual initiative shapes his view of inequality. Progress does not diffuse evenly at first, he says. His metaphor is a goose that lays golden eggs: the first owner becomes wealthy, breeds more geese, and pulls ahead while others still lack access. Over time, the geese spread and more people receive the benefit. The initial period feels unequal because it is unequal, but eventually everyone may possess something that once seemed extraordinary.

He applies that argument to medicine as well as technology. CAR T therapies can be extremely expensive and scarce, he notes, but their existence is still better than having no therapy at all. AI may similarly benefit early users before becoming broadly available. Friedberg sees unequal diffusion as a cost of progress, not proof that progress should be suppressed.

Bartlett is less persuaded that the trade-off is acceptable. People who receive the golden goose late may find that housing, healthcare, and other necessities have become less affordable in the meantime. Public support then becomes less an ideological preference than a response to diminished purchasing power.

Their contrast between Denmark and the United States makes the disagreement explicit. Bartlett describes the Danish proposition as a higher floor: more public services, social trust, shorter work weeks, paid leave, and less exposure to financial ruin. Friedberg accepts that many people might choose it.

But he thinks the choice is not merely between a harsh system and a humane one. The American system offers fewer guarantees but a chance at greater progress, wealth, and individual upside. The Danish model, he says, offers more security but a more limited path to extraordinary outcomes.

Bartlett’s strongest point is that not everyone shares the temperament of a founder. Ambition, risk appetite, childhood experience, and personal drive are unevenly distributed. A system designed by people who constantly want to build may not serve those who do not.

Friedberg’s answer is educational rather than redistributive. Schools should teach that people have multiple ways to create value and choose a path. They should not imply that everyone must become an entrepreneur, but neither should they direct students automatically toward government work as their defining future.

The unresolved problem is how much security is necessary to make risk-taking possible, and how much security changes the incentive to take risks at all.

David Friedberg grounds his optimism about AI in a historical claim: previous technological revolutions did not reduce employment in aggregate. Mainframes generated fears about accountants and clerical staff; desktop computers generated another wave of anxiety; industrialization moved labor out of agriculture and into factories. Friedberg says that in these transitions, productivity rose, the economy expanded, and new forms of work emerged.

His mechanism is that technology increases the output of each worker and creates new revenue opportunities, not simply lower costs. A painter who supervises robots can complete more work. A materials-science company can use AI to screen vast numbers of potential materials, automate experiments, and develop more products. The relevant corporate question, he says, is not only which support roles can be eliminated, but what new things a business can now make and sell.

That is also what Friedberg reports from his own businesses. Teams using AI do more work and then ask to hire more people, he says. In life sciences, his organizations use internal models trained on their own data, specialized models, and general-purpose systems. The result, according to Friedberg, is not reduced ambition but a much wider scientific frontier.

He rejects the claim that AI will leave humans with nothing to do. In his view, it discounts human adaptability by treating today’s job categories as the limit of useful work. Fifty years ago, he says, people could not have named podcasters, dog walkers, personal trainers, or social-media creators as large occupational categories. The inability to specify future work now does not, for Friedberg, establish that future work will not exist.

Steven Bartlett accepts that AI could create new work eventually, but focuses on the transition. AI affects cognitive work, while robotics may also affect physical work. A junior employee, delivery driver, customer-service worker, or factory worker may face a much larger leap into whatever work comes next than did workers in earlier technological shifts.

Bartlett offers examples from his own operations. AI agents can organize email, scheduling, and administrative workflows that might once have supported assistant roles. He is hiring aggressively in some areas but more cautiously in others, especially entry-level work. An on-screen note citing Stanford and ADP data said early-career tech roles had dropped 19% as AI replaced entry-level tasks, while junior developer roles had declined by up to 10% within six quarters and senior engineers remained valuable.

Friedberg agrees that this pattern is emerging in software. Experienced engineers can use AI to write far more code, while junior engineers lack the judgment and background needed to direct those systems well. Companies therefore compete for senior staff while becoming less eager to hire inexperienced programmers to write their first lines of code.

But he disputes the conclusion that an enterprise will simply pocket savings and shrink. If AI lets a company make five products rather than one, he says, management has reason to invest more: more product lines, more revenue, and ultimately more hiring. In his view, money saved through automation does not disappear. It is spent by the business on new initiatives, or invested elsewhere and eventually directed into other productive activity.

Bartlett’s counterexample is a personal assistant. If software manages email and scheduling, he may simply decide not to hire the additional assistant he would otherwise have hired. Friedberg asks what happens to the money Bartlett no longer spends. If it is invested in a productive asset, he argues, it supports another business and other employment.

The disagreement turns on timing, retraining, and distribution. Friedberg assumes that productivity gains and investment will create more valuable work. Bartlett’s concern is that people whose tasks disappear may not be able to reach that work quickly, or at all, while their income is already gone.

The exchange over Klarna captures the ambiguity. Bartlett says Klarna’s chief executive told him that the company had moved from around 6,000 employees to fewer than 3,000 after it stopped recruiting, while revenue doubled. The company had adopted AI customer service while increasing the value it placed on human VIP service. Friedberg accepts that AI was part of the story, but describes the case as a startup restructuring as well. He does not know how fully occupied the displaced customer-service workers had been, nor where they went afterward.

He expects many jobs to become more human rather than disappear. As digital production becomes cheap, he expects in-person services and experiences to gain value: coffee made by a person, exercise classes, concerts, dinners, and work built around gathering with other people. He imagines people spending more time in this “life-space economy,” supported by greater automation elsewhere.

Bartlett’s own test of an AI-generated audio show complicates that confidence. An AI wrote the script and synthesized Bartlett’s voice for founder stories; viewers were told it was AI, yet 40% to 50% watched an hour-long program through to the end. Informational media, at least, may be highly substitutable. What may retain value is the human component: presence, community, trust, and shared experience.

Friedberg does not claim a smooth transition. The condition that would prove him wrong is mass unemployment. Rising unemployment alongside falling wages, he says, would be the point for policy intervention, potentially including AI taxes or other measures.

Bartlett’s concern is that waiting for broad labor-market deterioration may mean acting after a rapid shift has already taken hold. Friedberg replies that early evidence of job loss is precisely what policymakers should track. He does not support stopping AI development preemptively, because he believes the technology will continue to advance elsewhere even if the United States restricts it.

Friedberg’s answer to AI concentration is open source—or, more precisely, open-weight models that can be downloaded and used without paying a proprietary provider for every output.

He compares the opportunity to the early internet. Proprietary browser and server software initially created tolls and gatekeepers. Open-source projects such as Firefox and Apache made it easier for people to browse the web and build sites without those fees. Friedberg sees open AI models as a similar release of productive capability: a company or individual can run a capable model themselves, paying principally for the computing power needed to operate it.

The distributional claim is central to his optimism. Friedberg does not believe the value of AI will end up with a handful of frontier-model companies or their founders. In the next decade, he predicts, someone with no current net worth will download open-source AI, start a company, and become a billionaire. He expects many versions of that story.

He points to Chinese open-weight models as an immediate challenge to the assumption that American proprietary systems will retain all the value. During the discussion, he contrasts an estimated 50 cents per million output tokens for such models with $50 for Anthropic’s private model. The policy question, as he frames it, is whether American companies should be allowed to use cheaper Chinese models, what that does to domestic AI competitiveness, and whether their use introduces security risks.

Friedberg rejects the notion that open weights are inherently backdoors from China. In his description, an open-weight model is fundamentally a large set of numerical parameters, not a covert program operating inside a computer. But he accepts the broader dual-use problem. Widely distributed models could potentially aid cyberattacks, biological threats, or weapon design.

His answer is that society repeatedly faces a choice with powerful technologies. It can attempt to ban or centrally control anything that could cause harm, or it can allow beneficial technologies to spread while investing in defenses. He compares AI to DNA sequencing and DNA printing, which could theoretically be misused but have not been prohibited wholesale. In an AI era, he says, cyberdefense and biodefense become more important.

He also doubts the United States can stop AI development unilaterally. If it blocks data centers, models, or domestic deployment, he says, other countries can build the infrastructure and connect it to the internet. The policy question is whether the United States wants the jobs, capital, and capabilities associated with AI development at home or elsewhere.

His ownership argument extends to data centers. Friedberg agrees that data centers can raise local electricity costs, though he dismisses concerns about water use. His proposal is that every new data center operate on its own grid and generate sufficient power to send energy back to the wider system. In that model, data centers could lower electricity prices rather than strain households, while publicly traded companies and real-estate investment trusts would let retirement accounts own part of the buildout.

The practical vehicle he names repeatedly is the 401(k). Friedberg wants people anxious about AI to own shares in the businesses and infrastructure benefiting from it, so that the technology’s gains do not accrue only to people who already hold productive assets.

Friedberg’s life-sciences work is the clearest example of why he sees AI as expansionary rather than merely labor-saving. He describes aging as a process of epigenetic dysfunction: cells share the same DNA, but molecular markers regulate which genes are switched on or off. Over time, he says, DNA damage and repair can move those markers into the wrong positions, causing cells to make the wrong proteins or fail to make necessary ones.

That process, in his account, explains why the body seems to age broadly at once. Cells throughout the body experience the same statistical accumulation of damage and repair. As gene-regulation markers drift, tissues begin to malfunction: skin wrinkles, vision fades, hearts become less efficient, and other systems lose performance.

He identifies Shinya Yamanaka’s discovery of four proteins as a foundation for reversing aspects of that process. In Friedberg’s account, high doses can push a mature cell toward a stem-cell state, while smaller doses can move a cell toward a younger version of its existing identity. The danger is resetting too far and creating an embryonic state that can become cancerous.

Friedberg calls the biological proof of concept established and describes the remaining challenge as engineering: getting the right dose of the right factors into the right cells. He cites work in mice, monkeys, and retinal treatment as evidence that epigenetic reprogramming is becoming a new lane of medicine. He expects further human trials in the coming years, initially in discrete tissues such as the eye, skin, or heart.

AI matters because biology is too complex for ordinary trial and error. Friedberg estimates that a cell contains roughly 10 billion proteins interacting with one another and that the body contains around 10 trillion cells. AI can generate and simulate large numbers of candidate proteins in silicoA term for computer-based simulation or experimentation., narrow the list for lab testing, and automate portions of the experimental process.

The result, he says, is the ability to assess far more plausible interventions before committing years to a program. That is why, from his vantage point, AI creates more scientific work and more reasons to hire: it makes previously impractical questions tractable.

He applies the same view to materials science, engineering, agriculture, and industrial development. The value of AI, in his account, is not that it makes existing work cheaper. It opens domains where the experimental cycle was previously too slow, expensive, or difficult to pursue.

Friedberg expects economic pressure to make democratic socialism increasingly attractive. He predicts that Alexandria Ocasio-Cortez could run for president in 2028, defeat more conventional Democratic contenders, and potentially win. He qualifies the forecast—“what do I know?”—but sees the political terrain clearly: people want help with bills, healthcare, childcare, and food more than they want an argument about property rights.

If AI produces substantial displacement, he expects the political answer to be more government checks, government jobs, UBI, or another form of direct support. He does not think those measures solve the underlying issue. Publicly created jobs, he says, may cost more than the value they generate; the system becomes politically appealing because it offers immediate relief, not because it creates productive independence.

Steven Bartlett presses the practical point: for people in genuine hardship, there may be no other credible offer. Friedberg agrees that this is how it feels. His concern is that the guarantee comes at the cost of liberty and agency, especially if the state’s role becomes permanent rather than transitional.

The contrast between their positions is less about whether people need security than about how security should be delivered. Bartlett emphasizes the foundation that permits people to survive and take risks. Friedberg wants that foundation too, but believes people must also own the upside of economic growth. Without ownership, in his view, public guarantees become politically irresistible.