The Next RMA: Autonomous Systems, Proliferated LEO, and Intelligentized War
Post 11 of 12 — From Clausewitz to Orbit: Strategy, Revolution, and the Future of War
Post 3 established that military revolutions require four components working in concert: new technology, new systems, new operational concepts, and organizational adaptation. Post 4 showed that the interwar period was the greatest laboratory for this process in modern history, and that the side which developed the concept while the other side accumulated the hardware consistently won. Posts 8, 9, and 10 examined three ongoing cases of revolutionary change in the character of warfare. This post examines what comes next — and whether the United States is organized to lead it.
The next military revolution is not a single phenomenon. It is the convergence of three concurrent developments that are individually significant and jointly transformative: autonomous systems enabled by artificial intelligence, proliferated commercial LEO constellations that are reshaping the scouting and communications architecture of warfare, and intelligentized warfare doctrines that treat military competition as fundamentally an algorithmic contest. All three are underway simultaneously. None is complete. And the organizational adaptation required to integrate all three into a coherent warfighting capability is, by any honest assessment, lagging behind the technology on the American side.
The Warfare-as-Software-Century Thesis
Karp and Zamiska, in The Technological Republic, state the central challenge directly: one of the most significant challenges facing the United States is ensuring that the DoD turns the corner from an institution designed to fight and win kinetic wars to an organization that can design, build, and acquire AI weaponry — the unmanned drone swarms and robots that will dominate the coming battlefield. The twenty-first century is the software century. And the generation best positioned to develop such weaponry is also the most hesitant, the most skeptical of dedicating its considerable talents to military purposes [1].
This is the organizational adaptation problem in its starkest form. The talent required to build the AI systems, the autonomous platforms, and the battle network software that will define the next generation of military competition exists primarily in the commercial technology sector. That sector has, for a generation, been systematically disinclined to work on defense applications. The cultural and institutional gap between Silicon Valley and the Pentagon is one of the most consequential asymmetries in the current military competition — and it is one that China does not have, because the PLA operates without the civil-military separation that structures American technology governance.
From Drone War to Swarm War
Post 8 established the Ukraine drone war as Phase 1 of the autonomous systems revolution: human-operated FPV drones, executed at industrial scale, iterating at machine speed. This phase has already reshaped the character of tactical warfare. But it is not the endpoint.
Defense firms will need to develop AI-powered systems that enable drones to communicate automatically — not just with one another but also with a host of sensors. These networks exist, but not at the required scale. And the task is getting harder each day: as the drone-versus-drone war escalates, the quantity of drones deployed in each operation will grow from hundreds to thousands, making their automated coordination increasingly difficult [2]. Defense companies are racing to create AI that can coordinate attacks by multiple drones in an automated drone swarm — the holy grail of drone operations [3].
China is already planning for this phase. China has ambitious plans for its swarms. One involves taking drone swarms into near space, as part of a “combined arms” strike force of stealth drones, hypersonic vehicles, and high-altitude airships [4]. This is not a theoretical concept. China’s air force has announced projects to develop fully autonomous swarms of intelligent combat drones — even while officially calling for international agreements to ban the use of autonomous lethal weapons [5].
Anduril’s Lattice platform represents the American conceptual response to this trajectory. Lattice for Mission Autonomy provides a fundamental paradigm shift in how we conduct military operations, enabling the military to regain affordable mass with teams of low-cost autonomous systems under the command of a single human operator, increasing speed and accuracy by processing and analyzing data far faster and accurately than humans, and most importantly, saving lives by reducing risk to human operators in dangerous, highly-contested environments [6]. The “affordable mass” concept is a direct response to the cost-exchange problem identified in Post 9: if cheap adversary systems can overwhelm expensive defensive systems, the answer is cheap friendly systems that can overwhelm at equivalent or greater scale.
Proliferated LEO: Commercial Space as Military Infrastructure
Starlink’s performance in Ukraine demonstrated the military value of proliferated commercial LEO constellations. The USSF has drawn the doctrinal conclusion. The U.S. Space Force Commercial Space Strategy explicitly directs leveraging the commercial sector’s innovative capabilities, scalable production, and rapid technology refresh rates to enhance the resilience of national security space architectures, strengthen deterrence, and support Combatant Commander objectives [7].
The numbers explain the urgency. An anticipated 50,000 spacecraft in orbit by 2030 [8] will represent a total transformation of the orbital environment. Commercial operators — most of them not subject to direct military authority — will constitute the majority of the space infrastructure on which military operations depend. This creates both an opportunity and a vulnerability. Proliferation increases resilience: an adversary cannot attrite a constellation of thousands of small satellites the way it can attrite a constellation of dozens of large exquisite ones. But proliferation also expands the attack surface, multiplies the number of ground terminals requiring cybersecurity, and creates attribution problems when any of those thousands of satellites is interfered with.
The architectural response is disaggregation: distribute capability across many platforms rather than concentrating it in a few. This is the proliferated LEO strategy applied to military space architecture. It is also, not coincidentally, the strategy that best leverages the commercial sector’s cost curve and production rate advantages.
Algorithm Confrontation: The Decisive Variable
The sensor-to-shooter compression is the central competitive dynamic of the next era. Warfare has become a contest of networks as much as of weapons. Precision fires, autonomous systems and AI-driven analytics all rely on fast, secure data exchange. The side that can connect sensors to shooters faster wins [9].
The bottleneck is not sensors. Between proliferated LEO constellations, ground-based radar and optical networks, and tactical ISR drones, the modern battlefield is drowning in data. Traditional workflows can’t scale to the data volumes streaming in from space, air, sea, land, and cyberspace, meaning critical clues might be missed, or arrive too late, if processed manually [10]. The competitive variable is the AI that processes that data, fuses it into a coherent operational picture, and generates targeting solutions faster than a human analyst can. The next phase of space domain awareness will depend on how effectively companies can integrate diverse data streams — radar, optical, RF, and on-orbit — into unified, AI-powered systems [11].
This is the algorithmic warfare that China’s 2013 Science of Military Strategy anticipated and that the PLA has been building toward through three doctrinal revolutions. It is also the domain in which the American commercial AI sector has the deepest technical capability — if it can be mobilized for defense purposes at the required scale and speed.
The Organizational Adaptation Problem
Applying the four-component framework to the next RMA:
Technology is present in abundance. Commercial AI development is advancing at a pace that dwarfs any government program. Autonomous platforms are being deployed in Ukraine at industrial scale. Commercial launch is reducing the cost of orbital access dramatically.
New military systems are being built. Lattice, Starlink, proliferated small satellite constellations, AI-enabled targeting systems — these constitute the building blocks of a genuinely new combined arms architecture.
Operational concepts are being developed, but unevenly. The Ukraine frontline has generated more conceptual innovation in three years than the Pentagon has produced in a decade. Commercial defense firms like Anduril and Palantir are building operational concepts into their products. But the institutional processes for evaluating, validating, and scaling those concepts — doctrine, training, wargaming, acquisition — have not kept pace.
Organizational adaptation is the critical gap. The Army’s NGC2 program, building a modular open-architecture command-and-control backbone to integrate AI and commercial space connectivity, represents a serious attempt at the organizational transformation required [12]. It is also a single program in a department that has launched and killed similar programs repeatedly. For the Army, success won’t mean fielding a flawless system. It will mean building one that can evolve — fast enough to survive the next war, not the last [9].
Krepinevich’s verdict from Post 3 bears repeating: from the post-Cold War transformation effort through the rise and fall of Joint Forces Command through repeated attempts to develop operational concepts, the United States’ armed forces exhibit few, if any, of the characteristics of military organizations that succeed at disruptive innovation. That verdict was delivered before the current moment. It has not become less accurate.
Referenced Highlights
[1] “One of the most significant challenges that we face in this country is ensuring that the U.S. Department of Defense turns the corner from an institution designed to fight and win kinetic wars to an organization that can design, build, and acquire AI weaponry — the unmanned drone swarms and robots that will dominate the coming battlefield. The twenty-first century is the software century. And the generation best positioned to develop such weaponry is also the most hesitant, the most skeptical of dedicating its considerable talents to military purposes.”
The Technological Republic — Alexander C. Karp and Nicholas W. Zamiska. Open in Readwise
[2] “Defense firms will need to develop AI-powered systems that enable drones to communicate automatically — not just with one another but also with a host of sensors. These networks exist, but not at the required scale. And the task is getting harder each day: as the drone-versus-drone war escalates, the quantity of drones deployed in each operation will grow from hundreds to thousands, making their automated coordination increasingly difficult.”
The Dawn of Automated Warfare — Eric Schmidt. Open in Readwise
[3] “Defense companies are also racing to create AI that can coordinate attacks by multiple drones in an automated drone swarm — the holy grail of drone operations.”
The Dawn of Automated Warfare — Eric Schmidt. Open in Readwise
[4] “China has ambitious plans for its swarms. One involves taking drone swarms into near space, as part of a ‘combined arms’ strike force of stealth drones, hypersonic vehicles, and high-altitude airships.”
The Origins of Victory — Andrew F. Krepinevich. Open in Readwise
[5] “In April 2018, the Chinese government expressed support for an international agreement ‘to ban the use of fully autonomous lethal weapons systems.’ China explicitly did not call for a ban on developing these advanced weapons... On the same day that China called for banning the use of autonomous weapons, its air force announced a project to develop fully autonomous swarms of intelligent combat drones.”
The Kill Chain — Christian Brose. Open in Readwise
[6] “Lattice for Mission Autonomy provides a fundamental paradigm shift in how we conduct military operations, enabling the military to regain affordable mass with teams of low-cost autonomous systems under the command of a single human operator, increasing speed and accuracy by processing and analyzing data far faster and accurately than humans, and most importantly, saving lives by reducing risk to human operators in dangerous, highly-contested environments.”
Anduril Unveils Lattice for Mission Autonomy — Anduril Industries. Open in Readwise
[7] “The USSF will leverage the commercial sector’s innovative capabilities, scalable production, and rapid technology refresh rates to enhance the resilience of national security space architectures, strengthen deterrence, and support Combatant Commander objectives.”
U.S. Space Force Commercial Space Strategy — Rudolph Bowen et al. Open in Readwise
[8] “Adoption of artificial intelligence could transform an industry with an anticipated 50,000 spacecraft in orbit by 2030.”
Military Spending and Direct-to-Device Competition Are Reshaping the Space Economy — Debra Werner. Open in Readwise
[9] “Warfare has become a contest of networks as much as of weapons. Precision fires, autonomous systems and AI-driven analytics all rely on fast, secure data exchange. The side that can connect sensors to shooters faster wins... For the Army, success won’t mean fielding a flawless system. It will mean building one that can evolve — fast enough to survive the next war, not the last.”
Space Is Key to the Army’s Long March to a Connected Force — Sandra Erwin. Open in Readwise
[10] “Traditional workflows can’t scale to the data volumes streaming in from space, air, sea, land, and cyberspace, meaning critical clues might be missed, or arrive too late, if processed manually. That’s where AI agents excel — they accelerate triage, automate first-pass analysis, and surface anomalies and patterns in near real-time.”
From Analyst to AI Orchestrator: Evolving Roles in the Age of Autonomy — NV5. Open in Readwise
[11] “Many space trackers say the next phase of SDA will depend on how effectively companies can integrate diverse data streams — radar, optical, RF and on-orbit — into unified, AI-powered systems.”
Managing Space Domain Awareness Data Has Become a Greater Challenge Than Collecting It — Jason Rainbow. Open in Readwise
[12] “Now the Army is trying again, under a new banner: Next Generation Command and Control (NGC2). Instead of the closed, bespoke systems that have defined Pentagon programs for decades, NGC2 is being built around an open architecture — a modular tech stack of software, data infrastructure and resilient communications. The Army wants to plug in innovations from the commercial world: cloud computing, artificial intelligence and, increasingly, space-based connectivity.”
Space Is Key to the Army’s Long March to a Connected Force — Sandra Erwin. Open in Readwise