The Claim

Every functioning society operates on a theory of who has a claim on what it produces, and why.

This is not a comfortable sentence. It sounds like the opening of a political argument, and political arguments about distribution have a way of generating more heat than light. But the claim theory is not primarily a political question. It is a structural one. Productive systems generate output. Output must go somewhere. The rules, formal and informal, that determine where it goes constitute a claim theory whether or not anyone has chosen to articulate one.

For most of the industrial era, the dominant claim theory was legible and, within its own logic, coherent: you have a claim on the surplus of the productive system to the extent that you contribute to producing it. Labor contribution generated a wage. Wage was the claim. The system was not fair in its distribution of labor’s rewards, and a great deal of political history consists of arguments about that unfairness. But the underlying logic was understood by all parties: contribution earns a claim. The argument was about how much the claim was worth, not about whether contribution-based claiming made sense as a framework.

Automation does not merely challenge this framework. At scale, it dissolves it.

When the productive system stops needing the contribution, the contribution-based claim expires. Not because the person became less worthy of a decent life. Not because their need for income, structure, identity, and belonging diminished. Because the theory of the claim was always conditional on a labor market that valued what they had to offer, and that labor market is being restructured.

What theory of the claim survives when the contribution mechanism fails?

This is the question that the next century of political economy will attempt to answer, with varying degrees of honesty and varying degrees of success. What follows is an attempt to map the terrain.

Assets: What You Actually Have

#

The first move in any credible framework is an honest inventory.

Countries, communities, and populations that will navigate this transition with any agency will need to know what assets they actually hold, not what assets the development textbooks assumed they would hold, not what assets they wish they had. What they actually control that the rest of the productive system values.

Labor is one asset. It has historically been the primary one for most of the world’s population. Its value is declining at the bottom of the skill distribution and shifting in nature at higher levels. This is the asset whose erosion the previous two essays described.

But labor is not the only asset, and for many populations it will not be the decisive one in the coming decades.

Land remains valuable in ways that are being reconfigured. Agricultural land, under pressure from climate disruption and under transformation from precision agriculture, is not uniformly valuable, but the countries with arable land in a world where food security is becoming geopolitically significant hold something real. More interestingly, the geographic features that the digital infrastructure economy needs, locations with the right climate for data centers, routes through which submarine cables run, territorial waters through which global shipping passes, proximity to major consumer markets, these are forms of geographic capital that have not historically been treated as negotiating assets. They are becoming negotiating assets.

Critical minerals are the most discussed asset in this category, and the discussion has a way of collapsing into resource curse pessimism. The record of countries with concentrated mineral wealth achieving broad-based development is not encouraging. Extractive industries generate revenue that often concentrates in state institutions and elite networks rather than distributing through the broader economy. The employment multiplier is low. The infrastructure requirements are capital-intensive and frequently met by foreign investment that repatriates returns. All of this is true. But the structural importance of lithium, cobalt, rare earth elements, and other inputs to the energy transition and technology manufacturing is not diminishing. Countries that hold these resources and have the institutional capacity to negotiate their extraction terms rather than simply accepting them have leverage. Developing that institutional capacity is a project, but it is a tractable one.

The most underexamined asset is data.

Every large population has been generating training material for global AI systems for roughly two decades. Behavioral data, linguistic data, medical patterns, agricultural practices, urban mobility patterns, consumer preferences, social network structures: all of this has been collected, transmitted, processed, and incorporated into systems that are now generating enormous economic value. The populations whose lives generated this data have received nothing for it. The data was collected under terms of service that no one read, governed by legal frameworks that no one designed with this purpose in mind, and the value it generated flowed to the companies that built the systems.

This is not inevitable. It is a choice, made by default, in the absence of a collective framework for treating population-scale data as an asset subject to negotiation rather than a resource available for extraction.

A credible data sovereignty framework, one that enabled populations to collectively negotiate the terms on which their behavioral and biological data trains the systems that then compete with their labor, would represent a genuinely new development mechanism. The populations with the most valuable data are, in many cases, the populations with the largest and youngest demographics: billions of people whose consumption patterns, health outcomes, agricultural practices, and urban mobility have never been captured by global training datasets at scale. They have something the global AI economy needs. They have not yet organized to extract value from it.

The asset inventory matters because assets create negotiating position, and negotiating position is the precondition for extracting value rather than simply receiving what the system offers. Countries and communities that approach AI civilization without legible assets receive it as a product designed elsewhere for someone else’s priorities. Countries and communities that have organized their assets can negotiate terms.

Distribution: The Instruments Available

#

The surplus that automated productive systems generate is real and it is large. Productivity per unit of human labor is increasing at rates without historical precedent. The question of who captures that surplus, which the contribution-based claim theory resolved automatically through labor markets, is now a question that requires deliberate design.

The instruments available are not equivalent in their applicability. Each has conditions under which it works and conditions under which it does not.

Universal basic income is the instrument that has received the most philosophical attention and the least successful implementation at scale. The logic is sound: decouple income from labor contribution, provide a floor below which no one falls, allow people to participate in economic life as consumers and agents regardless of their position in the labor market. The conditions for this to work are demanding. It requires a productive base large enough to be taxed at the level necessary to fund it. It requires political institutions capable of building and maintaining consensus around redistribution at that scale. It requires a monetary system stable enough that the income floor maintains real purchasing power. These conditions exist in wealthy, institutionally stable countries. They do not exist in most of the world. The instrument is real. Its geographic applicability is limited.

Sovereign wealth funds represent an approach that has worked where it has been tried: countries that control productive assets accumulate capital collectively, invest it, and distribute returns broadly. Norway’s Government Pension Fund is the clearest demonstration that this can work at national scale. The Gulf state models are variations with different governance structures and different distribution mechanisms. The conditions required are: a productive asset whose returns can be captured by the state rather than by private extraction, institutional capacity to manage the fund without corruption or capture, and political will to maintain the commitment to collective accumulation across political cycles. These conditions are not widely met. Where they are met, or could be built, the instrument is powerful.

Cooperative and community ownership of productive AI infrastructure is the instrument least developed in current policy discourse and potentially the most significant over the longer term. The argument is simple in structure: if the surplus generated by automated production is captured by the owners of the productive infrastructure, then distributing ownership of that infrastructure distributes the surplus. Worker cooperatives, municipal ownership of utility-scale automation, community investment structures that allow local populations to hold equity in the systems operating in their contexts: these are not novel concepts. They have existed in various forms for centuries. What is novel is their potential application to the automated productive infrastructure that is displacing labor.

The barriers are substantial. Capitalized productive AI infrastructure is expensive. Access to the capital required to acquire it is unevenly distributed in precisely the ways that mirror the existing inequality this framework is trying to address. But infrastructure that is built with public resources, whether municipal, national, or international development finance, can be structured for collective ownership from inception rather than requiring purchase after private accumulation.

Managed trade with employment or social contribution floors is the instrument that is politically most difficult and economically most coherent. The logic: access to large consumer markets is a privilege that can be conditioned. Countries or companies that wish to access the consumer markets of the European Union, the United States, or other large economies can be required, as a condition of access, to demonstrate minimum employment levels in specified communities, minimum contribution to community development funds, or minimum compliance with labor and environmental standards in their production. This is a departure from the free-trade consensus that has governed global economic policy since the 1990s. That consensus was built on a world where comparative advantage in production generated employment somewhere. In a world where comparative advantage in automated production generates employment almost nowhere except at the ownership and engineering layer, the intellectual foundation of the free-trade consensus requires reconsideration. The political conditions for that reconsideration are arriving, somewhat chaotically, in the form of industrial policy resurgence across major economies. The intellectual apparatus has not caught up with the political moment.

Data dividends as a development mechanism occupy an unusual position: they are logically compelling, potentially significant in scale, and entirely dependent on collective bargaining capacity that does not currently exist and would have to be built. If a country with a large population negotiated, collectively, the terms on which its population’s data trains the models of the major AI companies, the potential revenue stream is not trivial. The companies that built foundation models on the behavioral data of billions of people without compensation are sitting on claims that have not been asserted. Asserting them requires governance architecture: data sovereignty legislation, international negotiating capacity, technical infrastructure to audit and meter data use. These are buildable things. They require political will and institutional investment.

The Participation Question

#

The income problem and the participation problem are related but not identical, and solving one does not solve the other. Part 60 of this series mapped the interior experience of what happens when cognitive indifference and connected loneliness become the normal condition rather than the exceptional one. That mapping is relevant here, at the civilizational scale.

A society that solves income distribution without solving participation has bought time. It has not resolved the problem.

Employment provided participation automatically, as a byproduct. You went to work, and in going to work you had a reason to leave the house at a specific time, a set of relationships organized around shared purpose, a role in a social structure that assigned you a defined position, and the daily experience of your effort mattering to something beyond yourself. None of this was designed as participation provision. It happened as a consequence of the economic transaction. When the economic transaction disappears, the participation it bundled disappears with it, and nothing replaces it automatically.

Human societies have organized meaningful participation through forms that are not primarily economic: religious practice, civic institutions, military service, athletic competition, craft traditions, artistic production, community governance, educational participation, care networks. These forms exist. They are not new inventions. The question is whether they can carry, at scale and under conditions of material adequacy, the weight that employment carried for a century.

I do not know the answer to this. I suspect the answer varies. Societies with strong civic institutions and high baseline social trust, societies where non-economic participation has been consistently valued and resourced alongside economic participation, may be better positioned to maintain cohesion through a transition that removes employment as the primary participation mechanism. Societies that organized social value almost entirely through labor market position, where your worth was your wage and your identity was your occupation, face a more severe version of the problem. The stripping away of the economic mechanism strips away most of what organized social meaning.

This is not a problem that markets solve. It is not a problem that governments typically design for. It is a problem that requires being named as a problem before anything can be done about it. We are still in the naming phase.

The Honest Cartography

#

The framework has three components: assets, distribution mechanisms, and participation. Their combination differs across contexts. And honesty requires mapping that variation, including the parts that are uncomfortable.

Some countries and communities have strong positions across all three dimensions. Wealthy, institutionally stable, small-to-medium population nations with geographic assets, existing infrastructure, and strong civic traditions have real room to navigate this transition. Their challenge is primarily political: building the will to deploy available instruments at the necessary scale, overcoming the incumbent interests that benefit from the current distribution, designing the participation systems that employment provided automatically. These are not trivial challenges. But they are challenges from a position of genuine choice.

Some countries have strong asset positions but weak distribution and participation infrastructure. Large young-population countries with significant data assets, critical mineral endowments, or geographic leverage have negotiating position they have not yet organized into an effective claim. Their challenge is collective agency: recognizing that their assets have value, building the institutional capacity to negotiate rather than simply receive, resisting the extraction relationships that have historically characterized resource-rich development contexts. The path is not obvious. It exists.

Some countries have sophisticated institutional traditions and strong participation culture but limited productive asset bases and aging populations. These face the hardest version of the distribution problem: the productive base is shrinking as the population requiring support from it ages, and the instruments available require a surplus that a contracting productive economy may not generate. Automation may be the only way to maintain output, but the politics of automation in a context of labor scarcity and aging populations is different from the politics of automation in a context of labor surplus. Both are difficult. They are differently difficult.

And some populations face a combination of conditions in which no currently available instrument provides a credible pathway. Young, poor, resource-light, institutionally fragile, geopolitically peripheral: the combination of all five characteristics simultaneously places populations outside the reach of most of the mechanisms described above. Universal basic income requires a productive base and institutional stability. Sovereign wealth funds require assets to capitalize them. Cooperative ownership requires capital. Data dividends require governance capacity. Managed trade requires negotiating leverage.

This sentence must be in this essay: for some populations, within the current global system, there is no clearly available path.

That is not a comfortable sentence. It is an accurate one. A framework that papers over this reality is not a framework. It is a map with blank spaces filled in by wishful geography.

What Agency Actually Means

#

The word agency has appeared throughout this trilogy of essays, and it deserves a precise definition before this inquiry closes.

Agency is not the capacity to choose freely among unlimited options. That is not available to any individual, community, or nation. Circumstances are inherited. Resources are unequally distributed. History constrains the present. No country chooses its geography, its colonial history, its demographic structure, or the timing of its encounter with a given technology.

Agency is the capacity to act on an accurate understanding of the situation you are actually in, with the assets you actually have, toward goals you have actually chosen rather than inherited without examination.

The employment frame denied agency in a specific way: it offered a single answer to the question of how people and societies achieve wellbeing, presented that answer as natural law rather than historical contingency, and then attributed failure to achieve it to individual or institutional deficiency rather than to the failure of the mechanism. Countries that did not develop were told they had not followed the recipe. Workers who could not find stable employment were told they had not acquired the right skills. The mechanism’s failure was invisible because the mechanism was invisible.

Dissolving the employment frame makes the mechanism visible. And making the mechanism visible makes it possible to ask, honestly, what mechanisms are available, for whom, under what conditions.

The framework is: know your assets. Understand which distribution instruments your conditions actually permit. Ask the participation question deliberately, because no structural mechanism will address it automatically. And be honest about what combination you face.

Agency, at the scale of nations and communities, is the capacity to act on an accurate map. The development economics of the late twentieth century gave countries a map that worked often enough to become canonical. The terrain has changed. The map has not been updated.

The updated map is harder to read. It does not show one road. It shows a landscape of assets, mechanisms, conditions, and possibilities, some combinations of which open paths and some of which, at present, do not. The cartographer’s obligation is to show the terrain as it is. False comfort is the most dangerous thing the map could offer.

Reading the map honestly, acting on what it shows rather than what we wish it showed, building the institutional capacity to use the instruments that are available while being clear about those that are not: this is what agency looks like at civilizational scale.

It is the most demanding thing any of this asks of us.

It is also the only place from which something real can be built.