The Architecture of Power
Palantir, The Technological Republic, and the manifesto as commodity-form
There is a particular vertigo that attends the moment when a corporation publishes a philosophical manifesto. The genre is not new: IBM had its THINK posters, Apple its 1984, Google its mantra against evil; but Palantir’s twenty-two-point thread, drawn from Alexander Karp and Nicholas Zamiska’s The Technological Republic, belongs to a different register. It is neither slogan nor advertisement. It is a programme. And the question is not whether one finds it persuasive but what kind of object it is, and what work it performs in the political economy of the present.
What follows is a reading in three movements: the book as it stands, the thread as a distinct ideological artefact, and the contradiction that emerges when the two are placed in the same frame. The interesting argument is not that Palantir is sinister (the company has done the work of saying so itself) but that its self-presentation makes legible something about how capital now conducts its quarrels with itself.
I. The thesis of exhaustion
The Technological Republic is, in its better moments, a diagnosis rather than a prescription. The diagnosis is civilisational fatigue. Karp and Zamiska argue that the West has lived for nearly a century inside what John Lewis Gaddis named the “long peace,” a stability secured not by virtue but by the cold arithmetic of mutually assured destruction. The peace was anomalous, fragile, and (this is the authors’ central claim) corrosive. Three generations have grown up without a great-power war, and the engineering class that once served the state has migrated into the shallow waters of the consumer internet.
The book’s scaffolding rests on three structural arguments, each of which deserves to be taken seriously before being taken apart.
The hollowing of the technologist
The first argument is that the postmodern dismantling of Western grand narratives has left a void where civic purpose once stood. Stripped of a coherent national project, the technological elite has defaulted to a thin, utilitarian morality: optimising click-through rates, compressing video for delivery, mediating loneliness through application layers. The authors are not nostalgic for empire; they are nostalgic for ambition. The atom was split, the moon was reached, and the inheritors of those projects, they argue, now build photo filters.
The diagnosis has a structural force that resists easy dismissal. There is something genuinely true about the strange asymmetry between the technical sophistication of the present and the triviality of its dominant applications. To dismiss this observation as reactionary is to miss what it identifies.
The pathology of bureaucracy
The second argument is more surprising, and is consistently underread by hostile reviewers. The book is not, in fact, a paean to state power. It is a scathing autopsy of the procurement state. The illustrative anecdote is precise: in the early 1990s, with American forces deploying into Kuwait, the Air Force discovered it lacked enough handheld two-way radios, the kind sold for twenty dollars at any electronics store. Motorola had stock available. The deal collapsed because Motorola’s accounting systems could not produce manufacturing-cost data in the format the Federal Acquisition Regulation required. The most powerful military in history could not legally buy commodity hardware off the shelf in wartime.
This anecdote is doing real argumentative work. It establishes that the authors’ quarrel is not between tech and the state but between two configurations of state-tech intimacy: the post-Reagan procurement bureaucracy on one side, and an idealised mid-century fusion (the Manhattan Project, ARPA, the early Stanford Research Institute) on the other. The book is, in this register, a critique of administrative ossification masquerading as a critique of pacifism.
The Eck Swarm and the foxes
The third argument is the book’s most distinctive and is, predictably, the part the X thread excises most thoroughly. Karp and Zamiska reach for two unlikely sources to articulate what an unalienated engineering culture might look like. The first is the Eck Swarm: Martin Lindauer’s 1951 observation, in a Munich park, of honeybees collectively deliberating among eight potential nesting sites through decentralised dance. The second is Keith Johnstone’s improvisational theatre, in which the cardinal rule is that one must say yes to what is offered before complicating it. Behind both stands Isaiah Berlin’s distinction between the hedgehog who knows one big thing and the fox who knows many. Silicon Valley, the authors insist, is the consummate fox, and the fox should not pretend to be a hedgehog.
This is the most theoretically generative passage in the book. It articulates a vision of organisational form (flat, decentralised, anti-authoritarian, governed by emergent consensus rather than hierarchy) that is in pointed tension with the manifesto’s own demand for unified national purpose. Whether the authors register this contradiction is unclear. They proceed as though the swarm and the state could be married without incident.
It is at this point, between the bee and the bomb, that the book begins to fall apart.
II. The thread as a different object
The twenty-two-point summary that Palantir’s corporate account published is not a précis. It is a different object than the book it claims to compress. To read the two side by side is to watch a contemplative diagnosis transmuted into something closer to a recruitment poster, and the transmutation is itself the symptom worth examining.
Consider what survives the compression and what does not. What survives: the demand for hard power, the call for AI weapons, the mockery of pacifism, the assertion that some cultures are “middling, regressive, harmful,” the insistence that Silicon Valley owes a moral debt, the call to reconsider the all-volunteer force. What does not survive: the entire psychological architecture (Asch’s conformity experiments, Milgram’s obedience studies) that Karp and Zamiska use precisely to warn against the formation of monolithic ideology. What does not survive: the Eck Swarm, the improvisational startup, the fox. What does not survive: the critique of bureaucratic ossification, which constitutes nearly a third of the book.
The compression is not random. It systematically excises every passage that complicates the manifesto’s call to arms. A book that argues for decentralised, anti-authoritarian engineering culture becomes, on the timeline, an argument for unified national militarisation. A book that warns against the psychological pressure to conform becomes a warning that one must conform to the new programme. The thread is not a summary of the book; it is the book purged of its self-doubt.
This is not a complaint about reading comprehension. It is a structural feature of the medium. The thread cannot accommodate the book’s contradictions because contradictions do not propagate on platforms optimised for affective uptake. What propagates is the assertion stripped of its qualification. And so the thread does what the book refuses to do: it speaks with one voice, in declarative cadence, with the rhetorical posture of a state department that has been disintermediated.
There is something instructive in the observation that the very platform Palantir’s authors disparage as a symptom of civilisational decay is the one chosen to propagate the cure. The contradiction is not lost on the company; it is the operating logic. The thread is performative rather than constative. It does not describe a politics; it enacts one. And what it enacts is the spectacle of seriousness: the aestheticisation of strategic resolve, packaged for an audience of investors, recruits, and rival capitals.
The interesting feature of the Palantir thread is not its content but its publication. Other frontier-AI firms hold positions on military and surveillance applications that range from active participation to formal restriction; none has felt the need to publish a twenty-two-point credo. The manifesto-form is itself the symptom. It appears where capital needs to coordinate not just labour and contracts but also the legitimacy of what it is about to do, and where the appeal to internal corporate policy or quiet contractual practice is judged insufficient to that task. To publish a manifesto is to confess that something must now be said aloud.
III. The materialist undertow
Read at this register, the book and the thread cease to look like text and apparatus and begin to look like one continuous gesture. The gesture is the ideological work of a particular faction of capital, addressing a particular constituency, in a particular conjuncture. The interesting question is not whether the authors are sincere (they almost certainly are) but what their sincerity is for.
Factional warfare in the bourgeoisie
The book’s indictment of consumer technology is not, as it presents itself, a critique of capitalism. It is a quarrel within capital. Defence and surveillance capital, long the junior partner of the consumer-internet behemoths, is making a bid to redirect engineering labour, valuation multiples, and political attention toward its own balance sheet. The “moral debt” owed by Silicon Valley turns out, on inspection, to be denominated in contracts. The authors are not asking the technological class to renounce capital. They are asking it to pick a different patron.
This factional reading also explains the otherwise curious structure of the book’s anger. Why is Karp angrier with Google’s engineers than with, say, the offshoring of fabrication to Taiwan, or the financialisation of the engineering profession itself? Because the former is a competitor for talent and capital; the latter are not. The polemic does not, of course, reduce without remainder to the cash flows; the authors’ civilisational alarm is real, and the reduction would itself be too clean. But the polemic tracks the cash flows closely enough that a reader watching only the money would predict its grammar correctly.
The imperial monopoly on violence
The call for a “new Manhattan Project” is best read not as historical analogy but as a demand for state-guaranteed demand. The Manhattan Project worked, from the perspective of the firms that participated, because it offered cost-plus contracts, regulatory cover, and indefinite procurement horizons. What Karp and Zamiska are advocating is the reconstitution of those conditions for the AI sector. The “West” in this account is the imperial core, and the dispensation it requires is one in which the volatile dynamics of consumer markets are replaced by the predictable rhythms of military Keynesianism.
The book’s description of German and Japanese pacifism as an “overcorrection,” a phrase that landed with notable force in European policy circles, should be read in this light. The genuine geopolitical anxiety it taps into is not manufactured; the elite panic is real, and rearmament debates were already advancing in Berlin and Tokyo without Palantir’s prompting. But real panic and market opportunity coexist without difficulty. A rearmed Europe and a remilitarised Japan are, for a defence-software firm, addressable markets that have been politically off-limits for seventy years. The polemic against pacifism does not invent the anxiety it mobilises; it monetises it.
The fantasy of the missing centre
Where the materialist reading needs to be supplemented is in accounting for the manifesto’s persistent gesture toward a missing object: the “shared culture,” the “national purpose,” the “civic rituals” the authors mourn. This is the moment at which a Lacanian register becomes useful. The thread does not, strictly speaking, propose a politics. It proposes a fantasy that organises desire around a void. The repeated insistence that the West must rediscover what it has lost performs the constitutive function: the loss is not contingent but structural, and the manifesto’s political utility lies precisely in its refusal to specify what would count as recovery.
Žižek’s observation that ideology is most effective when it cynically acknowledges its own ideological character applies here with unusual exactness. Karp and Zamiska know the West they invoke is a constructed fiction. They say so. The point is not to mistake the fiction for reality but to use the fiction as a coordinating mechanism for capital and labour that would otherwise have no shared horizon. The thread is, in this sense, a quilting point: a master signifier that retroactively organises the field of meaning so that drone swarms, cultural cohesion, and procurement reform appear to belong to the same project.
Synthesis: the manifesto as commodity-form
The deepest contradiction in The Technological Republic is the one its authors cannot resolve and the thread cannot afford to expose. The book’s most original passages (the Eck Swarm, the improvisational startup, the warning against Aschian conformity) describe an organisational form that is constitutively incompatible with the project the same authors propose. One cannot have a flat, decentralised, dissent-tolerant engineering culture and a unified national defence apparatus animated by shared cultural purpose. These are not complementary visions. They are antagonists. The book’s rhetorical sleight of hand is to present them as continuous, and the thread’s function is to ensure no one notices the seam.
This is what makes the manifesto useful as a diagnostic instrument, even (perhaps especially) for those who reject its conclusions. It exposes, with unusual clarity, the structure of the present conjuncture: a fraction of the technological bourgeoisie attempting to reconstitute the conditions of mid-century state-capital fusion under the sign of a manufactured cultural emergency, while propagating the appeal through the very platforms whose hollowness it diagnoses. The diagnosis is not wrong about the hollowness. It is wrong, or perhaps strategically silent, about who emptied the room.
To diagnose the manifesto’s fantasy is not to stand outside it. The materialist reading shares the longing for a coordinating frame; what it refuses is the form Karp and Zamiska propose. The refusal is not yet an alternative, and to pretend otherwise would be its own ideological move. The conjuncture’s difficulty is precisely that the absence of a frame is intolerable and that no available frame is innocent.
Karp and Zamiska identify a real symptom: the alienation of the technologist, the sclerosis of the procurement state, the disappearance of any frame larger than the next quarterly earnings call. They are correct that something will rush in to fill the void; they are wrong to imagine that what they are offering is recovery rather than capture. What they describe as a return to civic purpose is, in operational terms, the subsumption of the engineering class under the disciplinary regime of military capital; a regime which, like all such regimes, will require its dissenters and its swarms and its foxes to fall in line. The hedgehog wins in the end.
The question the manifesto refuses to ask, but which the conjuncture poses regardless, is whose hands the means of coordination (computational, military, cultural) will fall into, and on what terms. “The West” is not an answer. It is the form the unanswered question takes when capital needs to keep moving.