Surveillance Valley
Source: https://www.publicaffairsbooks.com/titles/yasha-levine/surveillance-valley/9781610398022/ ↗
Levine reconstructs the history of the internet that the Silicon Valley origin myth prefers to forget.
ARPANET was not a project to survive nuclear war — it was a counterinsurgency tool, funded by ARPA to help the U.S. military process data from the field in Vietnam and monitor domestic dissent at home.
The book follows this surveillance thread from the 1960s through Google's early contracts with intelligence agencies, the Tor network's origins as a Naval Research Lab project, and the structural entanglement between the tech industry and the national security state.
Levine's reporting is solidly documented and serves as a necessary antidote to the countercultural narrative popularized by Stewart Brand and Fred Turner.
For product leaders, the book raises uncomfortable questions about the institutional origins of the platforms they build on.
Understanding where your infrastructure comes from is part of understanding what it can and cannot become.
Central argument
Levine argues that the internet was not born from countercultural idealism or academic curiosity but from U.S. military counterinsurgency programs: ARPANET was funded by ARPA to help process battlefield data in Vietnam and monitor domestic dissent, not to survive nuclear war. This surveillance logic was not incidental but structural, and Levine traces it forward through Google's early intelligence agency contracts and the Tor network's origins at the Naval Research Lab. The tech industry's entanglement with the national security state is, in his account, a feature inherited at founding — not a later corruption of an innocent infrastructure.
Critique
Levine's argument risks a genetic fallacy: demonstrating that the internet originated in military surveillance does not, by itself, establish that surveillance is its necessary or permanent character. A thoughtful reader might ask whether the documented institutional origins actually constrain what the infrastructure can become, or whether that leap requires a stronger theory of technological determinism than Levine supplies. The book is rigorous as journalism but thinner as a structural argument about how founding conditions propagate forward through decades of layered actors, incentives, and competing uses.
Why it matters for product
For a CPO, the book's most direct challenge is to the assumption that the platforms and infrastructure you build on are neutral substrates — understanding that data collection architectures were designed with surveillance legibility as a primary goal changes how you should reason about what your product is actually doing when it 'just uses standard APIs.' It also reframes decisions about third-party integrations, analytics vendors, and cloud providers as choices with institutional and political valence, not purely technical or commercial ones. Teams that treat infrastructure as a given rather than as a set of inherited design decisions are likely to reproduce those decisions invisibly in their own product choices.