Iterated Insights

Abstract:

Artificial intelligence may inaugurate a transition unlike prior technological revolutions. Whereas mechanization and computing increased productivity while preserving the economic centrality of human labor, advanced AI plausibly reduces the need for labor itself across a widening range of cognitive and productive tasks. This prospect forces a governance question that is not merely technical but distributive: if AI generates abundance in a post-labor economy, who should benefit from it?

This article develops an argument by historical analogy and by civilizational accounting. First, it revisits the development of the early internet from ARPANET to the World Wide Web, emphasizing how open protocols, public investment, and CERN’s decision to release web technologies without restrictive licensing enabled permissionless innovation at global scale. The internet’s subsequent history also clarifies a limitation: open foundations can coexist with later concentration of value at the platform layer. Second, the article frames modern AI as civilizational infrastructure built from cumulative, widely shared inputs including centuries of scientific knowledge, publicly funded research, global physical infrastructure, and the cultural and linguistic output of billions of people reflected in training data. On this view, contemporary firms play a crucial catalytic role, but they develop systems that rest on collective foundations that no private actor could plausibly claim as exclusive property.

The article then analyzes how capitalist incentives may function as a transitional accelerator under scarcity while becoming progressively redundant as AI systems approach autonomous innovation and low marginal cost production. It concludes that the central policy question is not whether AI should be developed, but how its gains should be governed once labor no longer serves as the primary distribution mechanism. Drawing on the internet precedent, the article argues for treating advanced AI as a shared inheritance and for developing institutional pathways toward a broadly distributed civilizational dividend.

Authors: Jared Edward Reser, Daniel Murray Reser, and ChatGPT 5.2

1. Introduction: A New Technological Crossroads

Technological revolutions often feel inevitable after the fact. Once a system spreads across the world it becomes difficult to imagine that it might have taken a different path. Yet history shows that the architecture and governance of transformative technologies are shaped by decisions made at key moments. Artificial intelligence now appears to be approaching such a moment.

For most of modern history economic growth has remained tied to human labor. Machines increased productivity but they did not remove the central role of people in production, coordination, and discovery. Even the most powerful industrial technologies still required large numbers of workers and experts to operate them. Artificial intelligence may be different. Systems are beginning to perform tasks that previously required education, judgment, and creativity. If progress continues, many forms of work that once anchored the economy may become optional rather than necessary.

This possibility raises a question that is not only technical or economic but also political and moral. If advanced AI dramatically expands productivity and reduces the need for human labor, who should benefit from that abundance? A narrow answer would hold that the gains belong primarily to the firms and investors that built the systems. A broader answer recognizes that modern AI rests on layers of human effort accumulated over generations. Scientific knowledge, public infrastructure, language, culture, and digital activity from billions of people all contribute to the training and operation of these systems.

The emergence of the internet offers a useful historical comparison. During the late twentieth century a new global network began to take shape. The institutions and researchers involved faced choices about whether the technology would remain open and widely accessible or become tightly controlled and commercialized from the start. The path that ultimately prevailed favored openness, shared protocols, and widespread participation. That decision allowed the network to grow into a global platform for innovation and communication.

Artificial intelligence may represent the next infrastructure of similar scale. The question now is not only how quickly it will develop but also how its benefits will be distributed. This article argues that AI should be understood as the product of a long civilizational process rather than the isolated achievement of a few organizations. For that reason the wealth created by advanced AI should ultimately be regarded as belonging, in some meaningful sense, to humanity as a whole. The history of the internet provides a precedent that can help guide how we think about this transition.

2. The Origins of the Internet: Public Science and Open Architecture

The modern internet did not begin as a commercial product. Its roots lie in publicly funded research and collaboration among scientists who were trying to solve practical problems in communication and computing. In the late 1960s the United States Department of Defense supported the creation of ARPANET, an experimental network designed to connect research institutions and allow them to share computing resources. The system introduced ideas that later became foundational, including packet switching and the linking of multiple independent networks.

Over the following decades the network expanded beyond its original military context. Universities, laboratories, and international partners joined the system. Researchers began to develop protocols that allowed different machines and networks to communicate reliably. The design philosophy that emerged emphasized interoperability and openness. Instead of building a single centralized network, engineers created a framework in which many networks could connect to one another using shared standards.

By the 1980s and early 1990s this evolving infrastructure was spreading beyond research communities. One of the most important developments occurred at CERN, the European particle physics laboratory near Geneva. Scientists there needed a better way to organize and share information across institutions. Tim Berners-Lee proposed a system that used hypertext documents connected through the internet. This system became the World Wide Web.

CERN made a decision that would prove historically significant. Rather than patenting the technology or charging licensing fees, the organization released the core protocols and software of the web to the public. Anyone could implement them, modify them, or build new services on top of them. The result was an explosion of experimentation. Individuals, universities, startups, and companies around the world began creating websites, browsers, and online services.

The early internet therefore grew out of a mixture of public investment, academic culture, and international cooperation. Its architecture encouraged participation rather than control. That openness did not prevent large companies from later emerging or capturing substantial economic value. Yet the decision to keep the foundations of the web accessible allowed innovation to occur at a global scale. It also established an important precedent. Some of the most influential technologies in modern history have been treated not as private property from the beginning but as shared infrastructure upon which others are free to build.

3. Why Openness Won: The Generative Power of Shared Infrastructure

The early internet was not guaranteed to succeed in the form that we recognize today. Many alternative models were possible. Large telecommunications firms could have developed closed networks with subscription access to information. Software companies might have built incompatible systems that locked users into proprietary platforms. Governments could have restricted participation to a small number of approved institutions. None of these outcomes would have been unusual by the standards of earlier communication technologies.

Instead, the internet grew around a different logic. The core protocols were publicly documented and widely implemented. Anyone with sufficient technical knowledge could connect a server to the network and publish information that became visible to users around the world. No central authority had to approve a new website or application. This permissionless quality turned the internet into a platform for experimentation.

The result was a type of innovation that is difficult to engineer from the top down. Individuals and small groups began creating tools, communities, and businesses that no central planner would have predicted. Search engines, online marketplaces, collaborative encyclopedias, open source software projects, and social networks all emerged within the same open environment. Many early creators were students, hobbyists, or small startups working with limited resources. The barrier to entry was low enough that ideas could spread quickly.

Openness also allowed economic value to expand far beyond the institutions that built the original infrastructure. The organizations that funded early networking research did not capture most of the wealth later generated by the internet economy. Instead, the network became a foundation on which millions of others could build. This pattern is familiar in the history of infrastructure. Railways, highways, electrical grids, and communication systems often enable activity that is much larger than the projects themselves.

At the same time, the internet demonstrated that open foundations do not automatically guarantee equal distribution of wealth. Over time a relatively small number of companies came to dominate major parts of the online economy. Platforms accumulated data, users, and capital at extraordinary scale. Yet even with this concentration, the open architecture of the network continued to support new entrants, independent creators, and global communication. The lesson is not that openness solves every problem. The lesson is that the structure of a technological system can shape the range of possibilities that follow.

As artificial intelligence advances, it raises a similar question about architecture and access. Will the systems that guide future economic activity be tightly controlled by a few actors, or will they function more like shared infrastructure that others can build upon? The history of the internet suggests that early design choices can influence this outcome for decades.

4. Artificial Intelligence as Civilizational Infrastructure

Artificial intelligence is often described as the product of particular companies, laboratories, or technological breakthroughs. This view contains some truth. Organizations invest large sums of money, hire talented researchers, and compete to develop more capable systems. Yet focusing only on these immediate actors obscures the deeper foundations that make modern AI possible.

Contemporary AI rests on layers of knowledge accumulated over centuries. Mathematics, statistics, computer science, neuroscience, and engineering all contribute to the techniques used in modern systems. Much of this knowledge emerged from universities and publicly funded research institutions rather than private industry alone. Scientific papers, open conferences, and international collaboration have played a central role in spreading ideas that later became commercial technologies.

Another essential ingredient is data generated through everyday human activity. Language models are trained on enormous collections of text, images, and other digital material. These datasets reflect the collective output of millions of writers, artists, programmers, teachers, journalists, and ordinary people communicating online. In a broad sense, modern AI systems learn patterns from the cultural and intellectual record of humanity itself.

The physical infrastructure behind AI is also deeply collective. Semiconductor manufacturing depends on decades of global investment and research. Data centers draw on electrical grids, fiber networks, and industrial supply chains that span continents. Governments have funded many of the underlying technologies, from early microelectronics to networking and satellite systems. Even the educational systems that train engineers and scientists represent long term social commitments.

Seen from this perspective, artificial intelligence begins to look less like a discrete invention and more like a continuation of a long civilizational process. Each generation contributes tools, knowledge, and institutions that enable the next wave of discovery. The organizations currently building advanced AI systems are important participants in this process, but they are not its sole authors.

This broader view matters because it changes how we think about ownership and responsibility. If AI were simply the product of a few private actors, it might seem reasonable that the benefits should flow primarily to them. If instead AI represents the culmination of contributions from societies across time and geography, then the case for a wider distribution of its benefits becomes stronger. Understanding AI as civilizational infrastructure helps frame the debate about what should happen as these systems grow more capable and begin to reshape the economy.

5. The Coming Post Labor Economy

For most of the industrial era, new technologies increased productivity but did not eliminate the need for human work. Mechanization transformed agriculture. Automation reshaped factories. Computers changed offices. Yet each wave still required large numbers of people to design systems, supervise machines, interpret information, and coordinate production. Employment shifted across sectors but the basic structure of the economy remained intact.

Artificial intelligence introduces the possibility of a different trajectory. Systems are beginning to perform tasks that were once considered the domain of trained professionals. They can write code, analyze documents, assist with research, generate designs, and interact with users in natural language. These capabilities remain imperfect, but the direction of progress is clear. With continued advances in computation, algorithms, and data, many forms of cognitive labor may become increasingly automated.

The economic implications of this shift are significant. Modern economies distribute purchasing power primarily through wages. People work, earn income, and use that income to obtain goods and services. If large portions of production can be carried out by machines with minimal human involvement, the traditional link between labor and income weakens. Productivity may rise even as the number of jobs required to sustain that productivity falls.

It is important to recognize that capitalism has been an effective engine for technological development under conditions of scarcity. Competitive markets encourage experimentation. Firms pursue different strategies, invest in new ideas, and compete to solve problems. Redundant efforts, though sometimes inefficient, can accelerate discovery because no single actor has perfect information. This dynamic has helped produce many of the technologies that define the modern world.

However, if artificial intelligence reaches a point where systems can rapidly generate solutions, design improvements, and manage complex operations with little human intervention, the economic role of redundancy may change. Parallel efforts that once drove discovery could become unnecessary duplication. Markets are well suited to environments where knowledge is dispersed and uncertain. In a world where advanced systems can coordinate information and production at extraordinary scale, the justification for some forms of competition may weaken.

The argument here is not that capitalism suddenly disappears or that markets cease to exist. Rather, the underlying conditions that made wage labor the central mechanism of distribution could erode. If production becomes increasingly automated, societies will need to consider new ways of allocating the wealth created by these systems. The transition may unfold gradually, but the direction raises questions that existing economic frameworks do not fully answer.

The internet offers a precedent for open standards, but artificial intelligence differs in a crucial way: frontier capability is not merely a protocol that can be published freely. It is a capital-intensive productive capacity that must be trained and continuously served at scale, consuming compute, energy, and operational labor. For this reason, the long-run challenge is not simply to make interfaces open, but to design the economic routing of AI-generated surplus before the post-labor phase arrives. Rather than relying primarily on heavy taxation after wealth has concentrated, one can imagine a pre-distribution architecture in which increasingly autonomous, self-improving systems are constitutionally constrained to allocate surplus by rule: reinvestment for safety and maintenance, bounded returns to early capital providers, and a broad civilizational dividend distributed widely. On this view, capitalism remains a useful accelerator during the scarce early phase, but the governance of abundance is engineered into the system itself once recursive autonomy makes traditional market incentives progressively redundant.

6. A Second Crossroads: Ownership, Abundance, and the Future of AI

The early internet developed during a moment when its creators faced choices about how the technology would be structured and shared. Decisions to maintain open protocols and release key components without restrictive licensing allowed the network to become a platform for global participation. That choice did not eliminate inequality or prevent the rise of dominant firms, but it did shape the environment in which innovation occurred.

Artificial intelligence now appears to be approaching a comparable moment. Advanced systems could become the central infrastructure of economic activity, influencing everything from research and manufacturing to communication and governance. As this happens, societies must decide whether the wealth generated by these systems will remain concentrated or be treated as the outcome of a broader human inheritance.

The case for a wider distribution of benefits rests on the historical foundations discussed earlier. AI systems are built upon centuries of scientific discovery, publicly funded research, shared language, cultural production, and the digital contributions of billions of people. No single organization created these conditions in isolation. Modern AI therefore represents a convergence of efforts that extend far beyond the boundaries of any company or nation.

Recognizing this does not require dismissing the role of entrepreneurs, engineers, and investors who have pushed the technology forward. Their contributions are substantial and deserve acknowledgment. The point is that the final stages of development occur on top of an immense base of collective human work. When the output of that process becomes capable of generating extraordinary abundance, the question of ownership takes on a different character.

The world that emerges from advanced AI may resemble a continuation of existing economic systems, or it may begin to diverge from them. Much will depend on how institutions respond during the transition. One possibility is that automated production remains tightly controlled, with the benefits flowing primarily to those who own the systems. Another possibility is that access to the technology becomes widespread but without mechanisms to share the resulting wealth. A third path recognizes AI as a form of civilizational infrastructure and seeks ways to distribute its gains broadly across humanity.

The history of the internet suggests that early choices can influence technological ecosystems for decades. The release of the web as an open platform helped create an environment in which people around the world could build, communicate, and innovate. Artificial intelligence now presents an opportunity to think carefully about similar questions of structure and benefit. If the technology truly represents the accumulated knowledge and activity of our species, then the abundance it produces may reasonably be viewed not as the property of a few, but as a dividend from the long project of human civilization.

Attribution:

The conceptual link between early internet governance and the emerging political economy of artificial intelligence was suggested to me by my father during a discussion about the development of ARPANET and the subsequent decision to maintain open network protocols. His observation that those early choices shaped decades of innovation prompted the historical comparison that motivates this article.