Real-World Models¶
BNI projects are not built in a vacuum. Every BNI initiative has prior art — community programs, tokenization experiments, AI deployments, or hardware-recovery projects that already exist somewhere in the world. This section catalogs the real-world models BNI Foundation studies, learns from, and is honest about diverging from.
The purpose is two-fold:
- Avoid reinventing wheels. If a model exists and works, BNI's job may be translation and adaptation, not from-scratch design. If it exists and fails, BNI's job is to know why before repeating the failure.
- Ground claims in evidence. Real-world models with measured outcomes give BNI's grant applications and product claims a foundation that pure design hypotheses do not.
The Epistemic Honesty directive applies here strongly: a model that worked in Kenya in 2018 with chamas as the trust substrate is not automatically validation that the same model works in West Virginia in 2026 with recovery cohorts as the trust substrate. Real-world evidence supports the thesis; transferring the implementation is its own [HYPOTHESIS].
Models under study¶
MPowerUP¶
| Model | Geography | Status | Deep entry |
|---|---|---|---|
| Sarafu Network (Grassroots Economics) | Kenya, since 2010 | [EMPIRICALLY VALIDATED] — community currency operational at scale; RCT-measured economic impact |
MPowerUP → Sarafu Network |
Second Boot¶
No real-world models documented yet. Candidates: Free Geek (Portland), PCs for People (national US), Computer Reach (Ohio). Each runs a different variant of the hardware-recovery-plus-digital-literacy model. To be researched.
RlivN¶
No real-world models documented yet. Candidates: ElliQ (Intuition Robotics), Embodied Moxie, Care.coach. Each runs a different variant of the conversational AI companion for elderly / cognitively impaired users. To be researched.
Toaster Chef (game / simulation)¶
No real-world models documented yet. Candidates: Free Rice (UN-backed educational game with measurable food-aid impact), Foldit (scientific crowdsourcing), McKinsey-style policy simulations used in development economics.
How BNI uses these models¶
For each model under study, the deep-entry page (in the relevant project's docs) covers:
- Origins & History — when it started, who started it, how it evolved.
- How It Works — the underlying mechanism, technical or social.
- Measured Impact — what outcomes have been documented, with citations and validation markers.
- Direct Applications for the relevant BNI project — what specifically translates.
- What NOT to Copy — failure modes, criticisms, context-specific limitations.
- Launch / Adaptation Plan — how BNI would actually deploy a learned-from-this variant.
- Conclusion — the synthesis: what's confirmed, what remains hypothesis.
This structure is the same across model entries for comparability.
Contributing a model¶
If you know of a real-world program or project that informs BNI's work — drop the reference URL into a new GitHub issue on the relevant project repo or this umbrella repo. Models with measured outcomes (RCT, longitudinal observational, large-N field data) are highest priority. Anecdotal models are useful too — flagged as such.