The refrain is by now well-known: China steels itself for war while America’s defense industrial base is, by its government’s own admission, failing.¹

The key weakness: a constrained capacity to keep pace with Chinese weapons production. As the 2024 report of the U.S.-China Security and Economic Review Commission made clear, “Whichever country secures a lead in key technologies – particularly those with first mover advantages – will tip the balance of power in its favor and reap economic benefits far into the 21st century.”²

There is a shrinking window of time to deploy these solutions and prevent a major war. Fortunately, recent breakthroughs in AI provide new opportunities to rethink manufacturing, including in ways that can expand the productivity of American workers, giving them what can only be described as superpowers.

It is with such superintelligent workers that the nation’s future hope rests, and policymakers must do all they can to precipitate and enable these Americans.

China Gathers Its Strength

Consider the vast scale of China’s shipbuilding industry, which not only serves its commercial interests, but also drives a rapid naval build-up. In 2020, China launched more naval ships than the combined total of all other nations; the People’s Liberation Army Navy has surpassed the United States in total ship count, with over 350 warships compared to the U.S. Navy’s 296.³ China has also grown its arsenal of precision-guided ballistic and cruise missiles. China’s DF-21D “carrier killer” ballistic missile poses a significant threat to American naval operations, while its DF-17 hypersonic glide vehicle is designed to evade missile defenses.⁴

Chinese stockpiles of these advanced weapons have ballooned, supported by an industrial base that can rapidly produce and deploy these systems. In addition to conventional missile development, China is investing heavily in emerging technologies, such as AI, quantum computing, and cyber warfare, with the aim to assert a decisive technological edge over adversaries. These efforts all contribute to a broader strategy to ensure China is able not only to launch, but also to sustain, prolonged military operations – something the American industrial base is increasingly unprepared to match.

A Brittle Deterrence

In stark contrast, America’s defense industrial base has struggled to keep pace with the demands of modern warfare. Decades of offshoring, budget cuts, and deindustrialization have severely weakened the manufacturing sectors critical to defense production. This is not wholly recent: during the wars in Iraq and Afghanistan, the United States faced shortages of basic equipment such as body armor, armored vehicles, and munitions.⁵ These deficiencies worsened, as the Pentagon’s munitions stockpile was depleted further through its support to Ukraine.⁶

While recent reports indicate that the United States has begun to replenish stockpiles of key precision-guided munitions at rates fast enough to meet current and future demand – including Javelin anti-tank missiles and HIMARS rocket systems – the production capacity for these weapons is constrained by supply chain bottlenecks. It also lacks surge capacity in key industries. For example, Raytheon, a major defense contractor, has faced delays in producing advanced air defense systems due to shortages in microchips and other critical components. Likewise, U.S. Navy shipyards also face challenges, with construction times extended far beyond acceptable timelines and in stark contrast to Chinese pace of production.⁷

In any protracted conflict, the United States would be hard-pressed to replace military assets at the rate necessary to maintain its global military commitments. Without substantial changes, the United States risks being outmatched in a future fight, not due to any shortcomings in military expertise, but rather those in industrial capacity.

The United States cannot compete with China on labor costs or scale; where America employs 100,000 workers, China can deploy one million.⁸ To overcome this disadvantage, the average American worker must become exponentially more productive.

Two critical developments – AI action models and digital twin platforms – are poised to bring about a revolution in industrial production. These innovations enable machines to learn from data, adapt to new conditions, and perform complex tasks with unprecedented precision and speed. By integrating these systems into the defense industrial base, the United States can significantly enhance its manufacturing capacity, allowing American workers to compete with China’s massive labor force. In short, AI could be the catalyst by which America closes the productivity gap and reclaims its position as the global industrial leader.

Dawn of The Dexterous Robots

For much of modern history, a manufacturer’s output was roughly tied to the number of workers on the production line. While automation has made significant strides, industrial output still depends heavily on human labor. Automating complex production processes at scale, with the necessary accuracy and throughput, remains extraordinarily difficult. Despite decades of advances in robotics and automation, human labor continues to fill critical gaps in production where machines lack the precision, flexibility, or intelligence.

However, recent breakthroughs in AI Action Models are rapidly expanding the possibilities of industrial automation. AI Action Models enable machines to go beyond traditional programming and learn from data, adapt to changing conditions, and even develop dexterity for intricate tasks.

Unlike previous generations of automation, which relied on predefined routines, AI Action Models empower robots and machines to perform a wider array of tasks and improve their performance continuously over time. Companies like Tesla already leverage AI-driven robotics for flexible manufacturing, and ABB uses intelligent robots in environments that demand precision, such as electronics assembly.⁹ These AI-driven machines are not only performing repetitive tasks, but they also are beginning to solve problems and carry out delicate operations that once required human skill.

The introduction of intelligent, dexterous robots into modern manufacturing has the potential to revolutionize production by allowing factories to scale output with fewer human workers. AI Action Models, with their ability to refine motor skills and adapt to new tasks, will enable machines to perform tasks that previously required direct human intervention, reducing the need for labor-intensive processes. As more of these dexterous robots become central to modern production lines, the role of human workers will likely shift from performing tasks that are repetitive and dangerous, to overseeing those performed by fleets of robots.

This will be a sea change, especially for countries like the United States, where labor costs and availability place manufacturers at a significant disadvantage compared to China. By closing the productivity gap with automation powered by AI, American manufacturers could dramatically increase efficiency, reclaim competitiveness, and reduce reliance on human labor on the production floor. While such AI-enabled advantages are also available to China, the CCP has proven less adept at diffusing newer technologies across its society (according to the U.S.-China Commission’s 2023 report).¹⁰ Herein lies the opportunity for the West.

Western economic history is punctuated by a series of industrial process revolutions, which created step-changes in production output. In the early 20^th century, Henry Ford’s assembly lines were a step change in manufacturing. Previously artisanal processes were mechanized, replaced by the structured division of labor – with man and machine operating more in tandem than ever before. This fueled the largest creation of wealth and power in human history. Robots enabled by AI action models may soon unlock the greatest revolution in production since the assembly line.

Digital Twins and the End of Human Knowledge Bottlenecks

Another critical innovation that AI brings to the table is the digital twin platform. Since the Industrial Revolution, the scale and efficiency of industries have been limited, not only by the number of workers, but also by the amount of “process knowledge” embedded in their expertise. Process knowledge — gained through years of experience on the factory floor — determines everything from optimizing machine operations to managing supplier performance. Process knowledge can only be acquired by constant human “learning by doing,” making it inherently hard to transmit beyond the factory floor. With process knowledge historically siloed, this constraint creates a bottleneck in industrial productivity.

AI-enabled digital twin platforms have the potential to break this bottleneck. Digital twins are virtual replicas of physical systems and processes, such as machines or entire factories, that can be continuously updated with real-time data. By optimizing the entire industrial system, both the individual factories and the supplier networks, for the generation, analysis, and AI-driven actionability of data, process knowledge can spread across an organization and improve over time. For workers, AI-enabled digital twin platforms have two radical implications: first, the most experienced workers become several orders of magnitude more productive; second, the least experienced workers can perform at the level of the most knowledgeable experts.

A first principle of manufacturing, as Elon Musk explains, is building assembly lines with “a linear adjacent flow [and] a tempo so that every station has roughly the same amount of work time, things move from one station to the next station, and the specialization of labor at each station.”¹¹ Creating a “linear tempo” to build highly complex systems like rocket ships or fighter jets means perfectly synchronizing suppliers, workers, machines, components and parts, schedules and deliveries, and more. AI not only makes this synchronization possible, but also learns to improve production process on the fly, and exponentially over time.

Today’s defense industrial base, however, is not capturing data on their fragmented processes to the degree required for AI-powered optimization. A fighter jet or battleship has hundreds of thousands of parts from a vast, diverse supply base. Each part can be supplied at different frequencies subject to its complexity and the capacity of the manufacturer, yet the defense industry does not track the lead time of these components.

The U.S. defense-industrial base could benefit enormously from adopting similar AI-driven practices. By using AI to model and optimize production processes, defense manufacturers could create a “GPS for production” that continuously improves speed, cost, and efficiency. Just as SpaceX has pioneered serialized rocket production, the United States must aim to accelerate the production of critical military assets, from warships to fighter jets, by integrating AI into its manufacturing systems. Without this technological leap, the country risks falling further behind China in terms of industrial defense output and capabilities.

AI: America’s Competitive Edge – And Strategic Imperative

The central question for American grand strategy in the 21st century is whether the United States can leverage its technological strengths, particularly in AI, to rebuild its industrial base and restore its global military dominance. Much of the current debate in Washington has focused on the risks associated with AI, such as job displacement and ethical dilemmas. However, these concerns should not overshadow the enormous potential of AI to revolutionize manufacturing. To remain competitive with peer adversaries like China, the United States must deploy AI aggressively in the manufacturing sector, or risk being outpaced by its rival.

1. USCC, “2024 Report to Congress of the US-China Economic and Security Review Commission,” US Government Publishing Office, November 2024. ↩︎
2. Ibid. ↩︎
3. Carla Babb, “China far outpacing US in military, commercial ship numbers,” VOA News, March 9, 2025.
   ↩︎
4. Maj. Christopher J. Mihal, PMP, “Understanding the People’s Liberation Army Rocket Force,” Military Review, Army University Press, September 2021. ↩︎
5. “US troops in Iraq, Afghan wars lack equipment-poll,” Reuters, August 9, 2007. ↩︎
6. Tara Copp, “Pentagon pause on arms shipments to Ukraine part of a global review of pressures on stockpiles,” AP, July 2, 2025. ↩︎
7. Evan Beebe, “Can New Legislation Reverse Navy Shipbuilding Failures?,” Institute for Defense and Government Advancement, July 16, 2025. ↩︎
8. Torsten Støk, “US Wages vs Wages in Chin and India,” Apollo Academy, September 16, 2024. ↩︎
9. “AI & Robotics,” Tesla; “Robots and humans take intelligent manufacturing to the next level,” ABB, May 15, 2024. ↩︎
10. USCC, “2023 Report to Congress of the US-China Economic and Security Review Commission,” U.S. Government Publishing Office, November 2023. ↩︎
11. Everyday Astronaut, “First Look Inside SpaceX’s Starfactory w/ Elon Musk,” YouTube, June 22, 2024. ↩︎