Privacy is architecture, not a feature

Key distinction from the manifesto: privacy is not a settings checkbox but an architectural decision made before the first line of code.

In practice: on-device processing, local storage, no accounts when possible. “If we can’t see your data, we can’t leak it, sell it, or be forced to hand it over.”

This isn’t altruism — it’s a competitive advantage. When Apple makes privacy a marketing moment and GDPR/CCPA tighten regulation, privacy-first architecture saves compliance costs and builds trust.

For a solo founder this is especially valuable: no legal department, no compliance team. If data isn’t on the server — there’s nothing to protect.

Implementation: MLX embeddings (Apple Silicon, on-device), Supabase Auth + RLS (minimal cloud), PostHog EU (GDPR-compliant analytics).

• manifest-summary — origin: manifesto principle #2
• privacy-first — hub page for privacy-first architecture
• antifragile-life-design — privacy = antifragility: less data = smaller attack surface
• bonsai-ternary-llm — concrete path to privacy-architecture for web AI: 500M LLM runs in browser via WebGPU, data never leaves the tab
• openai-privacy-filter — the “strip PII before sending to LLM provider” primitive: OpenAI’s 1.5B PII detection model (Apache-2.0), runs on laptop. Makes the strip-before-send pattern cheap