Pieces of the Action
Bush's memoir of directing the US wartime science effort — the Office of Scientific Research and Development, the coordination of the Manhattan Project, and the institutional design that made American science dominant for half a century.
This is the companion to his visionary 1945 essay "As We May Think": not the speculative vision of hypertext and personal computing but the operational story of how one person coordinated thousands of scientists, navigated military bureaucracies, and built the organisational structures that turned basic research into decisive advantage.
Bush writes with the directness of an engineer who had to make systems work under existential pressure.
For product directors, the book is a case study in the hardest version of their job — aligning autonomous, brilliant people toward a shared outcome without destroying the autonomy that makes them effective.
The Stripe Press edition rescued a book that had been out of print and underappreciated for decades.
Central argument
Bush argues that decisive institutional design — not just scientific genius — was the decisive factor in American wartime and postwar technological dominance. His central thesis is that the Office of Scientific Research and Development succeeded because it preserved the autonomy of individual scientists and university labs while imposing enough coordination structure to direct their output toward concrete military objectives. The lesson is organizational: the architecture of the system, including how authority was delegated, how civilian and military bureaucracies were kept in productive tension, and how basic research was connected to applied outcomes, mattered as much as any single breakthrough.
Critique
Bush writes as a protagonist, not a historian, which means the account is structurally blind to his own institutional power and the exclusions it rested on. The OSRD's model of 'autonomous brilliant people' coordinated by a trusted broker assumed a homogeneous scientific elite — overwhelmingly white, male, drawn from a handful of elite universities — and the book offers no reckoning with whether the organizational genius he describes was partly a function of that homogeneity rather than transferable to more diverse or distributed contexts. A reader who takes the model seriously has to do the work of separating the structural insights from the sociological conditions that made them work.
Why it matters for product
The core challenge Bush navigated — how to give expert contributors enough direction to be useful without enough control to become compliant — is the daily tension in any product organization that depends on engineers and designers thinking independently. His answer was to hold firm on outcomes and interfaces (what the military needed, by when) while leaving the scientific method entirely to the scientists; for product directors, this maps precisely onto the distinction between owning the problem space and resisting the urge to own the solution space. The OSRD's structure also illustrates why organizational design is a product decision: Bush had to ship an institution before he could ship technology.