Civic Hygiene – avoid building technologies that could be used by a police state (2013)

Original Article ↗ Hacker News Discussion ↗

Modern digital systems—from telemetry and ad tech to nationwide databases and smartphone platforms—can greatly enhance state surveillance and control, raising the question of whether technologists have a duty to avoid building tools that might “facilitate a police state.” Commenters argue over how realistic that constraint is in a world where almost all powerful technologies are dual use, with some advocating for new licensing models, data minimization, and privacy-by-design, while others insist that authoritarianism must be addressed primarily through politics rather than engineering. There is broad, if uneasy, agreement that individual ethics alone are insufficient safeguards, but that refusing to build the most obviously abusable systems and constraining how data is collected and centralized are still meaningful forms of resistance.

Scope of “civic hygiene”

  • Many agree with the core idea: don’t build systems that obviously empower persecution (e.g., centralized databases of religion/sexuality), even if current leaders seem benign.
  • Others say focusing on tools is shallow if you ignore the political conditions that enable a police state; the Nazis and East Germany repressed people without modern tech.

Dual‑use technology & inevitability debates

  • Repeated theme: “everything is dual use” – cars, the internet, cameras, cryptography, even rocks and sticks.
  • Some argue this makes strict avoidance impossible; what matters is who has power and how they use it.
  • Others counter that dual‑use doesn’t mean neutral: some technologies structurally favor centralization and surveillance (cheap cameras, centralized cloud telemetry), while others favor decentralization (BitTorrent, some cryptocurrencies).

Licensing, open source, and use restrictions

  • Proposals: introduce “no military / no surveillance / first do no harm” licenses for software.
  • Pushback:
    • Conflicts with the OSI principle of “no discrimination by field of endeavor.”
    • Likely unenforceable against states or secret programs; bad actors will ignore it.
    • Creates endless boundary fights over “dual use” and “commercial” similar to existing “non‑commercial” clauses.
  • Some still see expressive value in such clauses (publicly stating norms) even if they’re not enforceable.

Ethics of working on military & surveillance tech

  • Lived examples: game telemetry repurposed for artillery, medical‑device code reused in missiles.
  • Split views:
    • Some say “if you don’t do it, someone else will,” and prefer to “be at the helm” to steer or whistleblow.
    • Others see that logic as identical to “might as well work at a concentration camp” and insist on personal responsibility and conscientious refusal.
  • Distinction drawn between helping legitimate defense vs enabling domestic police states; but critics note military tools often flow to policing.

Politics, power, and preventing police states

  • Strong theme: you cannot “just focus on tech.” If you don’t want your work used in a police state, you must help prevent such a state: voting, local civic engagement, FOIAs, advocacy, protest, supporting privacy‑preserving projects.
  • Counter‑view: everyday people can “avoid politics” by voting and being decent; deep engagement is seen by some as the domain of “fanatics.”
  • Others reply that politics already shapes nearly everything in life; apathy simply delegates choices to others.

Surveillance infrastructure & data collection

  • Smartphones and major platforms already form a massive, semi‑voluntary surveillance network (location, contacts, behavior) that states can tap via warrants or purchases.
  • Flock license-plate cameras and similar systems are cited as qualitatively worse because:
    • They’re non‑opt‑in and pervasive in public space.
    • They normalize mass location tracking for routine policing.
    • Integration with other databases could make targeting dissidents trivial.
  • Some recount early webcam days, where they consciously refused to build generalized remote‑viewing networks after considering long‑term abuse risks.

Design heuristics & partial mitigations

  • Suggested “civic hygiene” heuristics:
    • Minimize data collection and retention; avoid building datasets whose abuse would be catastrophic.
    • Don’t build non‑consensual telemetry or “spying by default”; avoid engagement‑maximizing “Skinner box” designs.
    • Architect systems so users control roots of trust, cryptographic keys, and update mechanisms; favor peer‑to‑peer and open APIs to reduce central chokepoints.
  • Critics note this can rule out a large share of today’s web, and doesn’t fully stop determined regimes, but supporters argue it raises the cost and visibility of abuse.

Ethics education & professional norms

  • Many call for mandatory CS ethics courses and stronger professional norms akin to licensed engineering fields, to prepare developers for these decisions.
  • Others are skeptical: ethics classes exist but are often treated as checkboxes, may be ideologically skewed, and don’t reliably change behavior.
  • Some frame ethics education as useful not to “make people good,” but to give already‑ethical people concepts and tools to act more effectively.