The Motley Swarm: How a Dyson Structure Could Outgrow Every Plan That Built It

One familiar way of drawing a Dyson swarm is clean: a star inside an evenly spaced halo of identical collectors. Technical models often make similar choices because uniform components at a common radius make the physics easier to calculate.[1] The illustration compresses not only the machinery but its history. A multitude of independent objects begins to look like a single project.

Freeman Dyson’s original paper proposed looking for an artificial stellar source through the starlight it reradiated as infrared.[2] When readers objected to the implied shell and the authority required to build it, Dyson replied that he had meant a loose collection of objects traveling on independent orbits. He then imagined an open society expanding through proliferating city-states, each following its own orbit, and wrote that “such an expansion need not be planned or dictatorially imposed.”[3] He acknowledged that the political analogy was anthropomorphic. It was nevertheless part of the idea in 1960.

That part is easy to lose because we speak of an advanced civilization deciding to build a Dyson swarm, as though a civilization were a very large person who could approve the project, fund it, and wait for the contractors to finish. A central authority might do exactly that, but the physics does not require it. Organic growth is a description of the result’s relation to intention, not a construction method. Jason T. Wright later compared the growth of a swarm to a forest gradually closing its canopy, while a recent workshop report separated emergent construction from coherent operation: a system could grow without a grand design and still require shared tracking and station keeping.[4][5] The route has been named several times. What remains to be supplied is much of the middle—the local plans, interfaces, jurisdictions, repairs, failures, and inherited investments between the first orbiting settlement and the astronomical object.

Coordination at Several Scales

China’s export-industrial expansion offers a smaller case in which such a middle is visible. During the 1980s and 1990s, Hong Kong firms developed the large-scale “front shop, back factory” system across the border with the Pearl River Delta. Orders, sales networks, and finance remained in Hong Kong while labor-intensive processing moved into mainland factories. Taiwanese electronics companies later extended production networks into the Pearl River Delta, Greater Suzhou, and other mainland clusters, often bringing established suppliers with them. National reform, special economic zones, public infrastructure, local-government competition, global customers, and firm-level decisions all contributed, but no program specified the complete industrial system that resulted.[6][7]

Some Taiwanese computer networks in the Pearl River Delta maintained strong links to their existing Taiwanese suppliers and weak ones to domestic firms.[8] Their presence expanded production in China without producing substantial local upgrading. That complication belongs in the comparison. A system can become more capable as a whole while knowledge remains unevenly distributed and organizational boundaries remain surprisingly hard.

The operative distinction is therefore not between coordination and its absence. Coordination occurred at nested scales: firms planned factories, suppliers moved with customers, local governments built infrastructure and competed for investment, and the state created legal and physical conditions under which production could multiply. Each successful plant then changed the choice facing the next investor by adding suppliers, transport, skilled labor, information, or a market for specialized services. What no authority designed was the full network produced by those decisions. From a sufficient distance, the accumulated result resembled a coordinated program because many acts of coordination had produced order at a scale none of them controlled.[9]

What Changes in Orbit

The Chinese case does not supply a map of future orbital industry. It supplies mechanisms: complementary investments alter one another’s value; interfaces and sunk costs extend early choices; and plans made inside firms and jurisdictions do not add up to a plan for the whole. Cheap labor, territorial competition among local governments, and the division between a commercial front shop and a mainland factory do not travel into orbit in any obvious form. Orbital physics must decide the geography.

A mining operation might choose an asteroid for composition and accessibility; a processor might locate near the resource or at a transport node where removing unwanted mass saved more than the processing equipment cost. Depots, repair yards, fabrication plants, and some forms of computation might concentrate where material, energy, cooling, communication delay, and transport favored them.[10] Solar collectors would face a different pressure: as their density increased, they would shadow one another and require more elaborate collision management, giving them reasons to occupy different orbits rather than reproduce a terrestrial factory district.[4] An orbital industrial system could therefore thicken without becoming a thicker ring.

The result would be a differentiated network. Mines and processors could create demand for power; local power could make fabrication and maintenance cheaper; fabrication could reduce dependence on imported parts; more reliable transport and repair could make the next mine less speculative. Habitats, storage, communications, navigation, and traffic management could follow the equipment and the people responsible for it. None of these additions would need a Dyson swarm as its objective. Each would solve a problem created or made tractable by what was already there.

Standards could give those local decisions a longer reach. Separately built spacecraft already depend on common interfaces and communication protocols because compatibility reduces integration cost and permits equipment to be supported or replaced by organizations other than its maker.[11] Once a particular interface had accumulated compatible tools, suppliers, and trained maintainers, choosing something else could become expensive even if something else was better. Successful designs would also be copied because learning from another organization’s costly mistakes is among civilization’s more durable technologies. The swarm could acquire a recognizable order without anyone having selected the order as a whole.

Such a process would not be ungoverned. Present space activity already combines national authorization, international spectrum coordination, private operation, technical standards, debris rules, and jurisdiction that remains attached to registered spacecraft. These arrangements do not create ownership of orbital territory, and no world authority currently polices collisions. They do show how many bounded authorities can govern parts of one physical environment.[12] Future governments might subsidize trunk infrastructure, license operations, allocate access to spectrum and transport routes, require disposal or safe storage, and establish procedures for tracking and deconfliction. Firms, states, cooperatives, and their successors would plan large pieces of the system, and some might explicitly intend to expand it. Their plans would enter a structure already shaped by earlier equipment, claims, standards, and sunk costs.

The need for coordination would eventually be enforced by the orbits themselves. Brian Lacki finds that in a fully randomized swarm with a large covering fraction, the time between collisions could approach the orbital period, while the fragments from early impacts could drive a much faster cascade. Even more orderly swarms would be vulnerable over longer periods to gravitational and radiative perturbations unless something kept their trajectories compatible.[13] That something need not be one sovereign or even constant intervention. Carefully structured families of displaced, parallel orbital planes might reduce the burden of collision avoidance through architecture.[14] A durable swarm would require active upkeep, disciplined orbital design, or some combination of the two. Emergence would not release its builders from coordination; it would make compatibility across separately built elements a condition of survival.

Given enough time and enough builders, the installed equipment would probably arrive in vintages. New systems would be grafted onto old ones; useful components would be repaired, repurposed, and operated beyond their intended lives. Spacecraft have already displayed an unreasonable reluctance to die on schedule, and in-space servicing is being developed precisely because replacement is not always the best answer.[15] Some abandoned installations might become sources of parts. Others would become dangerous debris, drift beyond economical recovery, or survive as obstacles whose removal remained slightly more expensive than everyone’s present willingness to pay for it.

The resulting heterogeneity would not necessarily be visible as picturesque disorder. Builders facing the same gravity, radiation pressure, thermal limits, and collision risks might converge on similar designs; a common standard or traffic regime might impose regular orbital patterns on equipment belonging to many actors. A centrally directed system could travel in the opposite direction as successive designs, failures, and repairs accumulated. Geometry could supply clues about function and physical constraint, but it would not by itself disclose the system’s political history.

Institutionally, the swarm could still be motley. Ownership might be fragmented, and control would not necessarily follow ownership. Commercial nodes could depend on public navigation systems; nominally independent settlements might rely on one supplier’s maintenance protocols; strategic facilities might be privately operated under political constraint. The system could contain tightly governed regions, open regions, obsolete regions, dangerous regions, and pieces whose status had become clear to no one except the lawyers billing by the hour. Its order would consist partly of planning and partly of accommodations forced by everything that could no longer be moved.

Histories the Halo Cannot Distinguish

Automation creates the strongest rival history because it changes governance as well as speed. Models of a bootstrapped space industry describe a small seed installation making an increasing share of its own equipment and eventually expanding exponentially; partially self-replicating probe studies identify the microelectronics and other components that such systems would still need supplied from elsewhere.[16] If one actor controlled the first successful industrial lineage, its interfaces and designs might spread through the expanding industry before rival institutions had time to develop. Later ownership could fragment without undoing that technical ancestry. The first lineage would not need to own the later collectors if they depended on its standards, replacement parts, or descendants. The swarm could be physically modular yet historically centralized. If inexpensive replication became available to many actors instead, it could multiply incompatible lineages and accelerate institutional fragmentation.

A Dyson structure is not the inevitable endpoint of either history. Wright’s thermodynamic analysis finds task-dependent optima and steeply diminishing returns from capturing ever larger fractions of a star’s light.[4] A civilization could develop mines, settlements, factories, transport networks, and a great deal of energy collection without finding it worthwhile to enclose most of its star. There need be no internal milestone corresponding to the point at which a distant astronomer decides that the covering fraction has become large enough to deserve the name. “Dyson swarm” may be an observer’s category rather than a builder’s objective.

Construction history might leave traces in temperature distribution, gaps, covering factor, or variation in the intercepted light. None of those features uniquely identifies decentralized growth. A central program could build equipment for many functions at many temperatures, while independent builders facing the same constraints could converge on one design. Even simplified Dyson models can contain more physical parameters than the observations can recover.[1][4] The search would still look for missing starlight, waste heat, unusual variability, or other evidence of artificial structure. What would change is the inference drawn from a candidate: an energetic architecture is not a recovered polity.

An emergent Dyson structure would have several plausible founding dates and no single technically privileged one. Its history might begin with the first mine, the first self-sustaining off-world factory, the first collector, a famous statute, or the point at which somebody noticed that a collection of useful things had become something else. Whatever date later histories selected, the structure could have arrived one locally intelligible addition at a time. A distant observer might see mainly starlight intercepted and returned as infrared rather than an inspectable arrangement of machines. Thermodynamics would preserve the energy balance and discard the politics. The observer could draw a clean halo around the accumulated system and give it a name, but the name would say almost nothing about who built it, under what arrangements, or whether anyone ever intended to build the thing it described.

Notes

[1] Jason T. Wright, “Dyson Spheres”, Serbian Astronomical Journal 200 (2020): 1–18. [2] Freeman J. Dyson, “Search for Artificial Stellar Sources of Infrared Radiation”, Science 131 (1960): 1667–1668. [3] Freeman J. Dyson, “Response: Artificial Biosphere”, Science 132 (1960): 252–253. [4] Jason T. Wright, “Application of the Thermodynamics of Radiation to Dyson Spheres as Work Extractors and Computational Engines, and Their Observational Consequences”, The Astrophysical Journal 956, no. 1 (2023): 34. [5] Olivia Curtis et al., “The Dyson Minds 2025 Workshop: SETI around Black Holes”, arXiv workshop report (2026). [6] Hong Kong Education Bureau, “Regional Cooperation in the Pearl River Delta”; Douglas Zhihua Zeng, “How Do Special Economic Zones and Industrial Clusters Drive China’s Rapid Development?”, World Bank Policy Research Working Paper 5583 (2011). [7] You-Ren Yang and Chu-Joe Hsia, “Spatial Clustering and Organizational Dynamics of Transborder Production Networks”, Environment and Planning A 39, no. 6 (2007): 1346–1363; Siyue Chen et al., “Value Chain Organization Models of Taiwanese Electronic Information Enterprises in Mainland China”, PLOS ONE 16, no. 7 (2021): e0254402. [8] Chun Yang and Haifeng Liao, “Backward Linkages of Cross-Border Production Networks of Taiwanese PC Investment in the Pearl River Delta, China”, Tijdschrift voor Economische en Sociale Geografie 101 (2010): 199–217. [9] Gilles Duranton and Diego Puga, “Micro-Foundations of Urban Agglomeration Economies”, Handbook of Regional and Urban Economics 4 (2004): 2063–2117; W. Brian Arthur, “Competing Technologies, Increasing Returns, and Lock-In by Historical Events”, The Economic Journal 99 (1989): 116–131. [10] Marc M. Cohen et al., Robotic Asteroid Prospector, NASA technical report (2018); Gerald B. Sanders, Space Resource Utilization, NASA technical presentation (2015). [11] Consultative Committee for Space Data Systems, “What Is CCSDS?”; Daniel J. Sadey et al., “International Space Power System Interoperability Standards”, NASA technical paper (2022). [12] United Nations, Treaty on Principles Governing the Activities of States in the Exploration and Use of Outer Space (1967), especially Articles II, VI, and VIII; International Telecommunication Union, “Regulation of Satellite Systems”; Federal Communications Commission, Mitigation of Orbital Debris in the New Space Age, FCC 22-74 (2022). [13] Brian C. Lacki, “Ground to Dust: Collisional Cascades and the Fate of Kardashev II Megaswarms”, The Astrophysical Journal 985, no. 2 (2025): 191. [14] Colin R. McInnes, “Stellar Engines and Dyson Bubbles Can Be Stable”, Monthly Notices of the Royal Astronomical Society 546, no. 4 (2026): stag100. [15] NASA, Voyager Fact Sheet; NASA Goddard Space Flight Center, “In-Space Servicing, Assembly, and Manufacturing”. [16] Philip T. Metzger et al., “Affordable, Rapid Bootstrapping of the Space Industry and Solar System Civilization”, Journal of Aerospace Engineering 26, no. 2 (2013): 18–29; Olivia Borgue and Andreas M. Hein, “Near-Term Self-Replicating Probes: A Concept Design”, arXiv preprint (2020).