Technically Wrong: Sexist Apps, Biased Algorithms, and Other Threats of Toxic Tech
A catalogue of how the assumptions, demographics, and blind spots of design teams crystallise into products that harm the people they claim to serve — from name-validation forms that reject non-Western characters to algorithms that reinforce racial profiling.
Wachter-Boettcher grounds each case in specific, documented product decisions rather than abstract ethics, making the book unusually actionable for practitioners.
The argument is not that technology is inherently biased but that the homogeneity and insularity of the teams building it produces predictable failures that could be caught by broader perspectives at the design stage.
For product directors, the book makes a concrete business case for diversity that goes well beyond compliance or optics.
Central argument
Wachter-Boettcher argues that harmful product outcomes — name fields that reject non-Western characters, algorithms that reinforce racial profiling — are not accidents or edge cases but predictable consequences of homogeneous, insular design teams whose shared assumptions go unchallenged. The core thesis is that bias is not intrinsic to technology but is crystallised into it through specific, traceable design decisions made by people who lack the perspective to recognise their own blind spots. The book's method is deliberately case-based: by anchoring the argument in documented product failures rather than abstract ethics, it positions bias as a quality and risk problem, not merely a moral one.
Critique
The book's strength — its reliance on specific documented cases — is also a structural limitation: the examples tend to cluster around consumer-facing, high-profile tech companies, which may make it easier for product leaders at enterprise, B2B, or resource-constrained organisations to treat the failures as someone else's problem. More substantively, Wachter-Boettcher's prescription — broader perspectives at the design stage — risks becoming a checklist solution if not paired with an account of the structural and economic incentives that produced homogeneous teams in the first place; diagnosing the symptom more rigorously than the cause.
Why it matters for product
For a product director, the book reframes diversity in team composition as a discovery and risk-mitigation problem: homogeneous teams produce systematically narrower user models, which means whole categories of failure modes never surface until post-launch, when remediation costs are highest. This has direct implications for how product leaders structure research panels, define personas, and set acceptance criteria — the book suggests that edge cases are often only 'edge' from the perspective of whoever defined normal, which is a framing that should inform how discovery is scoped and who has sign-off on it.