The Companies Changing Warfare Forever: Palantir & Anduril Execs on Drones, AI & the Future of War

U.S. Manufacturing Erosion due to Post-Cold War Consolidation and Globalization Leaves Nation Vulnerable To Supply Chain Failures During Military Conflicts

Trae Stephens highlights how, over the past 30 years of globalization, the U.S. offshored much of its manufacturing, leading to plant closures across the nation, gutting whole communities. Stephens shares personal stories of family members who worked at GM, Ford, Frigidaire, National Cash Register, and Armco Steel in Ohio—companies whose local factories no longer exist, their operations relocated globally. According to Shyam Sankar, the shift of key technology manufacturing such as integrated circuits and microelectronics to Southeast Asia, driven by U.S. trade policy, ultimately eroded American capacity and introduced vulnerabilities into U.S. supply chains, making them susceptible not just to cost competition but also adversarial sabotage.

The 1990s saw significant defense industry consolidation. After the Cold War, Pentagon budget cuts and explicit encouragement to consolidate led 51 major defense contractors to merge into just five or six primes. The Department of Defense told industry leaders that survival depended on merging or transitioning to commercial business, which ultimately didn’t succeed for most. By 1999, the Justice Department halted further consolidation, blocking a Lockheed-Northrop merger that would have further shrunk the base.

This consolidation marked a shift from a broad industrial base, where major civilian companies like Chrysler, General Mills, and Ford contributed directly to U.S. defense efforts, to a specialist defense-only model. In 1989, only 6% of weapons spending went to pure-play defense companies and the rest to dual-use firms, but today 86% goes to defense specialists. Sankar notes this structural shift undermines surge capacity, as mobilizing civilian industry is no longer a straightforward or rapid process.

A critical illustration of current vulnerabilities is the bottleneck in drone production: numerous U.S. drone companies all rely on brushless motors, a key component still lacking sufficient domestic capacity. A sudden surge in military demand could reveal fatal supply chain weaknesses unless new investment remedies these gaps.

Anduril's Arsenal one Factory Exemplifies a New Defense Manufacturing Model Emphasizing Modularity, Scalability, and Rapid Retooling Over Single-Product Focus

Determined to address modern military needs, Stephens describes Anduril’s establishment of the Arsenal One factory—a 5-million-square-foot campus in Columbus, Ohio. The factory strategically taps into the underemployed manufacturing workforce left after decades of deindustrialization. The facility is designed around a contract-manufacturer model, focusing on modularity and agility so production lines can pivot rapidly between weapon systems—such as Furies, Roadrunners, or Barracudas—based on shifting military demand signals, rather than being tied rigidly to a single product.

The Arsenal platform, with a modern software backbone, seeks to minimize overhead and automate processes efficiently. Stephens notes this is similar to how contract manufacturers build multiple brands’ electronics, leveraging volume, skill specialization, and optimized supply chains for flexibility and scale. This kind of modular approach enables the U.S. to be more responsive in conflict, avoiding bottlenecks when rapid production of certain systems becomes necessary.

The urgency for this model was revealed in the Ukraine conflict: U.S. inventory burn rates for Stinger and Javelin missiles rapidly eclipsed years of production. When the existing stockpiles ran out, there were no active assembly lines—retired workers had to be recalled just to restart production. Stephens sees Arsenal One’s adaptability as a solution, designed from the outset to avoid these reactivation lags and make the U.S. military less vulnerable to sudden shocks.

Sustainable Defense Production Requires Viewing Munitions and Military Systems As Consumables With Continuous Replenishment Cycles

Historically, the Pentagon has treated munitions more like durable goods—stockpiled en masse—creating periods of stagnant production. Over time, politicians take money from these programs, production lines drop to minimum rates, and industry loses incentive and knowledge to maintain or improve manufacturing.

Sankar and Stephens call for a reimagining of defense procurement: munitions and drones should be viewed as consumables—expended in training and combat, with replacement planned from the outset. This create ...

Deterrence Erosion Shows Adversaries Doubt American Military Superiority Deters Their Strategic Objectives Through Annexations and Provocations

Shyam Sankar and Trae Stephens observe a decline in U.S. strategic deterrence, citing clear empirical examples: Russia’s annexation of Crimea in 2014, China’s militarization of the Spratly Islands in 2015, Iran's attainment of breakout nuclear capability in 2017, the 2023 pogrom in Israel, and the Houthis holding Red Sea trade lanes hostage. Despite the U.S. outspending all nations on defense, these provocations and aggressive expansions have occurred, signaling adversaries are less deterred by American military superiority.

Purpose of Defense Capability: Deter Aggression By Making It Costly and Risky, Demonstrating Capacity and Will to Decisively Protect American Interests

Stephens asserts that the ultimate purpose of military capability is not to go to war, but to make it unthinkable for adversaries to challenge American interests. This is achievable by maintaining capacities so fearsome that any potential aggressor concludes it is not worth the risk to challenge the U.S. National security, Sankar notes, is a means to support economic prosperity for Americans. For decades post-World War II, the U.S. had a primary seat at the table for structuring global systems like supply chains, semiconductors, and trade lanes, which underpins both persistent economic advantage and geopolitical leverage.

Deterrence Depends On Credible Capability and Resolve; Uncertainty Invites Provocations to Test U.S. Thresholds

The strength of deterrence hinges on both perceived capability and displayed resolve. As Stephens stresses, credible deterrence prevents adversaries from contemplating aggression. When the U.S. retreats from leadership, it loses the ability to set terms; instead, rivals dictate terms in their interests, and global order shifts quickly away from U.S. norms. If American deterrence is doubted or appears uncertain, adversaries are tempted to test U.S. thresholds, resulting in more frequent provocations.

"2027 Taiwan Window" Marks Critical U.s.-china Military Parity Decision Point

David Friedberg and Stephens highlight the "2027 Taiwan window" as a point of acute risk, with Taiwan’s vulnerability increasing as China rapidly closes the military capability gap. Stephens explains that U.S. dominance in semiconductor manufacturing eroded once TSMC in Taiwan gained momentum, and no comparable domestic industry grew to match it. As a result, Taiwan now hosts the majority of advanced semiconductor fabrication capacity, which constitutes a major vulnerability—both to global technology supply chains and to U.S. defense systems reliant on those chips.

Taiwan's Vulnerability Grows as China's Military Capabilities Advance

The accelerated advancement in China's military and strategic posture by 2027 exposes Taiwan to both military and economic coercion. The U.S. faces the danger that, should conflict arise, industrial and manufacturing gaps cannot be closed on wartime timelines; urgent defense readiness and industrial capacity expansion must be achieved before a crisis.

Taiwan's Semiconductor Manufacturing Concentration Poses Strategic Vulnerability to Global Supply Chains and U.S. Defense Systems

Taiwan’s concentration of advanced chip fabrication becomes a critical risk factor: the loss or compromise of this capability would deeply impact not just the U.S. military’s edge, but also the global commercial and consumer electronics markets.

Addressing Manufacturing Capacity Gaps For 2027: Urgent Need For Defense Readiness as Industrial Base Projects Can't Be Accelerated Retroactively Once Conflict Begins

Stephens warns that preparing the industrial base—especially for semiconductors and advanced defense technologies—is not something that can be accomplished in response to a crisis. Unless those gaps are closed proactively, they will remain insurmountable once a conflict emerges.

Defense Offset Theories: Ai and the Third Offset in Warfare

Sankar outlines the history of U.S. "offset strategies": The first offset leveraged nuclear weapon ...

The defense technology sector is undergoing a significant transformation as venture-backed firms challenge traditional models and attract heavy capital investment. Multiple voices, including Trae Stephens, Shyam Sankar, and David Friedberg, discuss these shifts, outlining new approaches to innovation, capital concentration, and sustainable growth in this rapidly evolving industry.

Venture-Backed Defense Firms Prioritize Innovation and Competition Over Relationship-Driven Cost-Plus Contracts

Traditional defense contractors, commonly known as the Primes, have long operated by responding passively to government requirements. Stephens and Sankar argue that these firms typically wait for government specifications and then build to those requirements, leading to incremental rather than breakthrough innovation. Because of the monopsony structure of defense procurement—where the government is the single buyer—contractors lack incentive to push forward with independent research and development, innovating only when expressly required.

Companies like Palantir and Anduril break from this passive model by investing heavily in private R&D to create products proactively, then pulling the market toward their innovations instead of waiting for the government to push them. Stephens states that Anduril, for example, uses private capital to fund extensive R&D and then sells the resulting products, operating with a fundamentally different business model oriented around aggressive product-led strategies.

This product-led approach, as Friedberg and Stephens note, requires substantially higher private capital to shoulder the risk inherent in R&D and allows for faster commercialization once government recognition follows. The ability to raise and efficiently utilize large amounts of capital is critical, as the demands of hardware and software development for defense tech are much greater than those of a typical tech startup.

Capital Concentration in Defense Tech Follows Power Law: A Few Companies Capture Most Value, Biggest Winners Return Entire Funds' Worth

As defense tech becomes a booming venture category, Friedberg notes the influx of venture capital and a sharp focus on finding the sector’s dominant players. Stephens draws parallels to other tech sectors, pointing out that in space, crypto, and social media, investors who missed out on category leaders like SpaceX, Coinbase, or Facebook failed to realize significant returns. Defense tech is no exception; history shows that concentrating capital into high-potential winners yields stronger results than spreading investments thinly across many contenders.

Sankar criticizes the tendency to "peanut butter spread" innovation capital, saying that this approach underfunds all companies and prevents any single firm from scaling up to true breakthrough innovation. Acknowledging the “authentic power law curve” in venture returns, both commentators explain that a few major winners—like Palantir and Anduril—can return an entire fund’s worth of capital.

Palantir's valuation journey illustrates this dynamic, rising from skepticism at a $20 billion projection in early employee equity offers to today's environment where even seed rounds sometimes claim such lofty valuations. Anduril’s rapid achievement of significant revenue milestones, reaching targets in 22 months that took Palantir five years, further demonstrates cumulative learning and capital efficiency prevalent in the sector’s major players.

Defense Startups Must Resist High Revenue Multiples to Avoid Unsustainable Growth

With surging capital and rising company valuations, there are risks of unsustainable growth. Stephens warns against the pressure to raise increasingly larger rounds at ever-higher revenue multiples, which can force companies into “hypergrowth” and lead to short-term failures or lacklus ...

Origin of Silicon Valley and Defense Divide: Vietnam War Schism and Generational Military Disconnect

Shyam Sankar explains that Silicon Valley’s original foundation was defense: in the 1950s, Lockheed was the region's largest employer and the Corona spy satellites were built there. The region’s posture was shaped by the Cold War and the threat of the Soviet Union, leading to active collaboration between technologists and defense institutions. After the Cold War, a new era of globalism led to growing cynicism toward defense, with threats seeming less tangible.

The Vietnam War significantly misrepresented the defense mission and broke this trust. Sankar argues that the war created a fundamental schism between academia and defense that has never healed, as people felt lied to about U.S. motives and actions. Since then, distrust has brewed and escalated through society.

Trae Stephens describes that at Stanford, undergraduates almost never have immediate family in the military, revealing how distant military service has become for elite tech circles. He observes that only among veterans at the business school do such connections exist. This disconnect means younger technologists lack direct contact with the military and rely on abstract cultural narratives and media depictions rather than firsthand observation to form beliefs about defense institutions. Shyam Sankar adds that, with few people having prior service or meaningful community connection, a "cartoon version" of defense work—shaped more by fears and Hollywood depictions than reality—dominates young people's perceptions. This erodes an informed and balanced understanding of both defense challenges and military professionalism.

Stephens further notes that society has lost connection with “salt of the earth, middle of the country, veteran community”—present in larger numbers during WWII and the Cold War—causing even deeper generational disconnect.

Adversaries Exploit Divisions By Funding Activism and Discouraging Defense Innovation

Sankar highlights that foreign adversaries have exploited these American divisions. During the Vietnam War, the Soviet Union spent $7 billion funding the American peace movement, adding "gasoline on the fire" by injecting money and messaging into activism—weaponizing skepticism to create division and discord. The Soviet role built upon an organic antiwar movement but magnified its impact.

Today, Sankar claims the Chinese Communist Party (CCP) funnels money to organizations protesting against American defense tech companies on domestic grounds, escalating the divide and discouraging innovation. David Friedberg and Stephens note that few Americans are willing to admit they could be influenced by outside actors, but this reluctance plays to adversaries’ advantage by weakening psychological defenses and leaving the country’s science, industry, and military vulnerable to destabilization through information operations and activism.

Cultural View of Defense-Industrial Complex as Corrupt Clouds Policy Debate and Skepticism Regarding Defense Professionals' Competence and Integrity

The prevailing cultural narrative depicts the defense-industrial complex as fundamentally corrupt. Trae Stephens references the long-running fallout from the Snowden disclosures, where criticism and distrust over surveillance and intelligence collection persist despite investigations finding almost no abuse—less than a dozen cases, largely due to technology errors.

Sankar underscores that every government agency has an independent inspector general empowered to investigate complaints—anonymous or named—ranging from defense to housing and urban development. These mechanisms provide transparency and accountability, often rendering illegal conduct in the defense sector more visible and prosecutable than in private industry. He describes how these mechanisms are taken extremely seriously—sometimes ...

Existing Autonomous Weapons Systems: Focus On Accountability and Human Oversight Over Autonomy Debates

Autonomous weapons systems are not new to modern warfare. Trae Stephens points to established systems like the SeaWiz (Phalanx CIWS), which is deployed on naval vessels and automatically engages aerial threats due to the need for split-second responses. In these cases, automation is essential—human operators do not have the time to intervene before a threat strikes. Yet, even with fully automated engagement, the vessel’s captain retains legal and moral responsibility for the system’s actions. This principle of accountability mirrors that of a soldier with a weapon or any military operator: command and responsibility are inseparable, regardless of a system’s autonomy.

Weapon development has evolved from brute-force approaches to increasingly precise and discriminate methods. Modern warfare now prioritizes precision-guided munitions, which can, for instance, neutralize threats by firing non-explosive missiles into specific windows rather than causing widespread destruction. The goal is to minimize civilian casualties and unintended harm.

Artificial intelligence is rapidly enhancing targeting precision, reducing collateral damage, and accelerating decision-making in combat environments. Stephens contends that developing and deploying superior, more precise technology is a moral imperative—abstaining from advancing such technology is itself a moral decision, potentially less just than active participation in improving warfare’s ethical landscape. He asserts that entrusting these advances to democratic institutions ensures their use aligns with the public interest.

Instances like Anthropic’s refusal to deploy the Claude AI model for the Department of Defense’s Maven project without explicit human oversight highlight the ongoing debate over the appropriate degree of human control in autonomous systems. Friedberg underscores the ethical necessity of maintaining clear human authority, especially when it comes to life-and-death decisions.

Vendors Must Enforce Policy Compliance Via Architecture, Detecting Activity Rather Than Relying On User Self-Compliance to Prevent Illegal Platform Use

Stephens stresses that technology vendors have a responsibility not just to establish ethical boundaries, but to enforce compliance through the technical architecture of their platforms. Rather than depending on end users to follow rules, systems should be designed to log, detect, and surface potential misuse for transparent oversight.

Palantir serves as a model in this regard. Their platform is built to highlight potentially illegal activity, aiding oversight and enabling the detection and prosecution of illegal government actions instead of concealing them from scrutiny. If any government agency illegally uses vendor technology, there are existing mechanisms—such as Inspector General (IG) offices—where such activity can be reported, investigated, and addressed. Every agency has an independent IG empowered to conduct thorough investigations when concerns arise.

Stephens argues that vendors’ normative role is to enforce legal limits and maintain transparent records of activity—not to independently determine what counts as ethical or permissible government action. The systems must ensure lawful use and aid accountability, rather than making subjective decisions about appropriate use.

Democratic Legitimacy Requires Elected Representatives to Decide On Surveillance, Data Collection, and Military Operations, While Tech Companies Imposing Ideological Constraints Exercise Unaccountable Veto Power Over Democratically Authorized Activities

Democratic legitimacy is foundational in governing the use of advanced technologies in defense and national security. Stephens underscores that U.S. intelligence community policies on data collection and retention have endured through administrations of both parties and repeated congressional reviews. These programs, continually debated and renewed, carry the imprimatur of democratic consensus and national interest.

When technology company executives decline to provide tools for government use based on their own ethical standards—as with Anthropic refusing ...