Karp has performed a necessary act of clarity. He has described, without euphemism, the redirection of the most concentrated technical talent in history toward the trivial. Silicon Valley’s dominant output, advertising systems, status platforms, logistical conveniences for the already comfortable, is not an accidental market outcome. It is the preferred activity of a class that inherited the security of a powerful state and then treated that security as a permanent background condition rather than a condition that must be continually maintained. The authors are right: the early American technical achievements that made the modern software industry possible were not pure market creations. They were the product of a deliberate partnership dissolved into consumerist self-enclosure, the horizon of ambition shrank. The book’s diagnosis of this shrinkage is its strongest contribution, and it should be stated without softening.

Here is the concrete evidence: engineers who will optimize the placement of advertisements refuse contracts that would improve targeting systems for the military that still underwrites the order in which they live. Large language models and the coming generations of autonomous systems are treated primarily as consumer interfaces or ethical puzzles rather than as the domain that will decide relative power among states. Karp correctly identifies the asymmetry: authoritarian competitors do not share the same cultural inhibitions. They are building the systems that will coordinate sensors, effectors, and decision loops at machine speed. To treat this primarily as a question of consumer product design or campus discourse is to misread the decisive arena. Schelling’s insight remains relevant here: the power to hurt, when anticipated, is bargaining power. In the software century that power will be exercised through code and autonomous platforms. The side that refuses to build them does not become more virtuous. It becomes more vulnerable.

The historical recovery is equally necessary. The myth of Silicon Valley as a spontaneous market phenomenon free of the state is a form of self-serving forgetting. Early semiconductor work, reconnaissance systems, and the research lineage that produced the internet itself were embedded in national technical requirements. When the technical elite turned away from those requirements after the Cold War, they did not become more independent. They became smaller in purpose. The market, left to itself, allocates talent to whatever can be monetized most rapidly among the secure. That allocation is not destiny. It is the failure of direction.

The organizational chapters supply a further strength. The lessons drawn from collective decision processes in natural systems, from the mechanics of improvisation, and from the experimental literature on conformity are practical. They describe how effective technical work actually occurs: through friction rather than consensus theater, through status earned by results rather than credentials, through the willingness to absorb external disapproval when the work demands it. These habits are portable. They explain why certain organizations deliver under pressure and others dissolve into process. The authors are right to extract them and to contrast them with the theatrical and risk-averse habits that now dominate much of elite institutional life.

Yet the book’s limitation is real and must be addressed directly. Karp and Zamiska still write from within the same liberal-procedural horizon whose thinning they diagnose. They call for a renewed technological republic, a thicker national purpose, and the return of elite technical talent to the service of the state. These are necessary corrections. Ultimately, they remain insufficient when the positive vision stays underspecified. What exactly is the content of the national project beyond the defense of existing rights and maintenance of relative power? The authors invoke collective identity and shared purpose, but they leave the substance thin. The result is a call for harder tools in service of a still soft frame.

In my opinion, this is where the argument must be strengthened. Elite theory supplies the missing pressure. Pareto’s circulation of elites, Mosca’s ruling class, and Burnham’s analysis of the managerial shift all point to the same reality. Every complex society is directed by a minority that controls the decisive instruments of organization and force. The question is never whether an elite exists. The question is whether that elite accepts responsibility for the material and strategic conditions of the order it inhabits, or whether it treats those conditions as externalities while extracting private advantage. The current technical elite has largely chosen the latter. It has enjoyed the protection of a powerful state while regarding service to that state’s hardest requirements as optional or contaminating. This is not a stable arrangement. Huntington’s observation on the clash of civilizations remains useful here. Cultural and civilizational identities are not infinitely elastic. When the leading technical stratum of a civilization severs itself from the requirements of that civilization’s survival, the result is not cosmopolitan elevation. It is strategic erosion.

Clausewitz supplies the complementary point. War is the continuation of politics by other means. The character of those means changes, but the underlying relationship does not. In the present period the decisive means are shifting toward software, autonomous systems, and the integration of sensing, decision, and effect at machine speed. A political community that cannot or will not master those means does not become more peaceful. It becomes less able to set the terms of any peace. The authors are correct that the atomic age is closing and that a new form of deterrence is required. The reinforcement is that this requirement cannot be met by a technical class that continues to treat national power as someone else’s problem.

Christopher Lasch’s analysis of the revolt of the elites is also relevant. A stratum that lives in the most protected zones of a society while cultivating distance from its harder obligations is not a leadership class. It is a rentier class with technical skills. Karp’s book documents the career data and the cultural habits that produce this distance. The necessary response is not merely to invite that stratum back into defense work; it is to reimpose the expectation that those who possess the decisive technical capacities accept responsibility for the strategic environment those capacities shape. Anything less leaves the diagnosis accurate and the prescription incomplete.

The engineering mindset that is described in this book can be pressed further in the same direction. The capacity to absorb the disapproval of the crowd, to prioritize results over process theater, and to maintain creative friction is not only method for building better systems. It is a method for reconstituting a technical elite that is less captive to the soft preferences of its immediate social environment. The authors stop at effectiveness. The stronger claim is that effectiveness in the decisive domain is itself a form of responsibility. Those who can build systems that will determine relative power and decline to do so are not neutral. They are choosing a horizon at the expense of the harder requirements of the order that makes their work possible.

This book has named the misallocation of talent, the asymmetry of will, and the strategic necessity of mastering the new domain. These points stand. The limitation of this manifesto is the residual softness of the frame in which they are placed. The required reinforcement is the explicit acceptance that a technical elite either directs its capacities toward the material conditions of collective survival or watches those conditions erode while it optimizes the trivial. There is no third option that preserves both the comfort and the power. The software century will rank the societies that understand this against those that prefer not to.