In her new book Valley of Death: How Big Tech is Fueling the Future of War, journalist and US security expert Sharon Weinberger provides a compelling account of how a new cohort of military-oriented tech firms is reshaping the US security apparatus and how it conducts business with the private sector. Referring to a Pentagon study from 2013 on the potential danger of a privately-owned nuclear bomb, the book opens with a simple yet important question: “What happens when Silicon Valley does finally bridge the proverbial Valley of Death and gets its products adopted by the US military – unleashing a new generation of weapons that belong to private industry?”. Spanning ten chapters and over 250 pages, Weinberger tackles this question in great depth by tracing the chronology and mechanics of what she terms the “Pentagon-Silicon Complex”. After Imagineers of War: The Untold Story of DARPA, the Pentagon Agency that Changed the World (2018), the book offers another insightful glimpse into the public-private dimension of military technology development, this time with a focus on the shifting political economy of defense acquisition as a key arena of contemporary military-industrial reconfiguration. The following review aims to provide a brief (and certainly incomplete) overview of the book’s central insights and my own key takeaways.
In the first three chapters, Weinberger examines the historical links between the US security apparatus and Silicon Valley. Chapter one recalls the widely recognized role of military funding in the core technologies of the 20th century, whilst also shedding light on lesser-known anecdotes, e.g. about how Steve Jobs’ firm NeXT bridged difficult years with CIA contracts before eventually merging into Apple. Chapter two and three then trace major changes in US warfare after 9/11 and explore the associated rise of a new breed of defense firms that are central to the book: Palantir and SpaceX. Here, Weinberger already provides some insights into the companies’ varying strategies and business models, for example when describing how SpaceX sought to establish new procurement logics (see below) or how Palantir created “internal demand and a network of trained users” (ch. 3) for its software products by offering free access and support to certain units.
Chapters four to seven trace the political and cultural shifts in Silicon Valley from the mid-2010s to the present day. Through the backstory of Anduril (chapter 4), the development of Project Maven (chapter 5) and the struggles over the JEDI cloud contract (chapter 6), Weinberger systematically explores how Silicon Valley has evolved from a supposedly liberal tech mecca into the stronghold of a new private-military complex, outlining the ideological battles fought along the way – between staff and management, but also between firms in their struggle for Pentagon billions. The seventh chapter builds on this, revealing the sector’s full militarization along the lines of a growing ‘China threat’ and the war in Ukraine. Here, Weinberger not only captures the importance of Ukraine as a testing ground and marketing opportunity for this new generation of US defense firms, but also how the embrace of great-power rivalry fueled a true gold rush among them and their investors. These developments, the book suggests, carry the risk of prioritizing vibes and profitability over functionality and diligence, as evident in an interview with an anonymous venture capitalist: “I honestly don’t care about the product […] It’s like, I want it to be cool and I want the thing that they work on to be something that the DOD is going to spend money on, right? But like doing technical diligence? It’s not how the DOD buys things” (ch. 7).
The book’s core argument and, in my view, greatest contribution lies in chapters eight to ten. These are particularly valuable from an analytical perspective, as they reveal a central contradiction in the ongoing transformation of the defense business: the so-called ‘neoprimes’ around Palantir, Anduril and SpaceX are widely regarded (and never tire of presenting themselves) as disruptors of an oligopolistic market. Their primary objective, so the narrative goes, would be to break up the gatekeeping role of the traditional prime contractors in order to increase competition and efficiency in military procurement. While the problem statement might be valid, the book repeatedly illustrates that the proposed solution put forward by these firms targets precisely this gatekeeping position for themselves. This paradox becomes particularly evident in Weinberger’s recap of a press meeting with Palantir’s Chief Technology Officer Shyam Sankar, one of the most vocal advocates of procurement reform:
"Palantir’s role in this new American mobilization would be to lead those defense startups out of the Valley of Death. It would do this, Sankar said, by allowing them to build on the company’s government-certified software platform, avoiding the months or even years it might take them otherwise to get accredited for federal work. What Palantir had created, Sankar said, was a 'wormhole' that would allow hundreds of software companies to sell their products to the Pentagon. Palantir would become the gateway for dozens if not hundreds of new companies seeking government work. […] In the world of military products, the government is a monopsonist, meaning the sole buyer, which allows it to dictate terms to the sellers. Sankar’s criticism of major defense companies, the primes, was that they act as gatekeepers for the monopsonist, because they were put in charge of major weapons, controlling which firms would supply the parts and subsystems. Yet Palantir’s plan for startups sounded similar. 'Doesn’t this make you the monopsonist gatekeeper then, if you’re the one that’s handling all of the separate entry points?', one of the reporters of my team asked. 'Well, for better – not for better or worse – those decisions aren’t mine', Sankar replied" (ch. 8).
As my colleagues Marijn Hoijtink, Jasper van der Kist and I have similarly argued in a recent forum article for Science as Culture, the new potential gatekeepers of the US defense market follow a playbook that centers around the platform business model and the exploitation of software lock-ins, which will arguably intensify rather than dissolve the perils and dependencies of the traditional military-industrial complex.
Another key aspect of the unfolding acquisition transformation is highlighted in Weinberger’s assessment of Anduril’s central procurement pledge. In her words, “Anduril appeared set to do for hardware what Palantir had done for software: conquer the Pentagon market” (ch. 8). In contrast to traditional defense contractors, the company likes to present itself as a ‘defense product company’, which implies a move away from classic waterfall procurement, i.e. the long process from problem definition and requirement specification to competitive tendering and cost-plus contracting to prototyping and production. Instead, Anduril and its peers propose an ‘off-the-shelf’ model, in which companies invest private (venture) capital in the development and production of new weapon systems and subsequently sell these to the Pentagon. Such a model, so the logic goes, would open doors for smaller businesses that cannot compete with the administrative and financial resources of the ‘primes’, thereby creating a competitive market and redistributing risk from the government to the private sector. Whilst advantageous at first glance, Weinberger points to the potential pitfalls of this model by bringing in the question of ownership. Who owns a software-based weapon system if it is developed and built with private capital? Who controls it when it is sold to the military ‘as a service’? And what motives then drive the development (and deployment) of these systems under capitalist profit imperatives? The fact that these questions are more than just hypothetical thought experiments is repeatedly demonstrated in the book, for example, with reference to Elon Musk’s decision to cut-off the Ukrainian military from the Starlink network.
By skillfully weaving in interviews and anecdotes with in-depth background knowledge, Weinberger presents some compelling answers to the important questions she raises throughout the book, highlighting the potential implications of private tech firms controlling the fate of what are, in some cases, revolutionary weapons systems. For what ultimately “differentiates the Andurils, Palantirs and SpaceXs from traditional defence companies”, she argues, “is that they own and control the technology they have developed, placing an immense amount of power in private hands” (ch. 10). One of the most important contributions of this book then lies in linking this very question of power and ownership to the shifting political economy of defense acquisition. As the title ‘Valley of Death’ – a common metaphor for the Pentagon’s cumbersome acquisition bureaucracy – already reveals, Weinberger captures the struggle over how the US military buys stuff from the private sector as a central arena of contemporary military-industrial restructuring around which new dependencies realize. The book convincingly traces the chronology of these organizational shifts and, more importantly, exposes its central contradictions and consequences. This makes it not only a refreshing and captivating read, but also an important one given the seemingly universal (and in some cases myopic) consensus on the necessity of a rapid acquisition overhaul.
