Four Propositions
The four nails BP, press, and stage talks must hammer in together
These four propositions hold each other up: #1 is the philosophical premise, #2 and #3 are the battlefields, #4 is why we can hold the ground. A pitch deck, press piece, or keynote must surface all four — never cherry-pick one.
Pinned. No single proposition forms a complete narrative on its own.
Proposition 01
"I don't know"
Not a bigger model — the first world model OS that can say "I don't know"
Scale players promise "bigger means more accurate". We promise "we know where we are not accurate, and we report it."
Core points
- Default output format
- Moat
- Anti-thesis pairing
Proposition 02
Full-stack health loop
The first full-stack health closed loop (TCM modernisation + Western-medicine auditability)
Evidence provenance + individual state space + risk gating — fused into one.
Core points
- Five-stage closed loop
- Moat
- First battlefield
Proposition 03
A brain that says "No"
The first brain that can say "No" (a refusal-capable brain any robot can mount)
Refusal is not silent striking. It is collaboration with an auditable account.
Core points
- Boundary Gate — five dimensions
- Refusal four-tuple
- Moat
Proposition 04
Research = the moat
Research is not a revenue line — it is the strategic moat
Base models are swappable; the "Genmount OS layer" is not. That is why the scale camp burns capital and we don't — and why we move across industries.
Core points
- Formula
- Four public-grade standard formats
- Transfer logic
Methodological boundaries · Four Antis
The Four Antis
Four boundaries pinned at the moment the model was conceived. This is not anti-science — it is anti the single path of pure engineering stacking.
Anti-token
Do not treat life and medicine as text or symbol sequences to fit. Build the state-mechanism-feedback structure first.
Anti-scale
Do not treat "bigger" as the only answer. Reach for stronger interpretability and transferability at smaller scale, given a structurally correct base.
Anti-stacking
Do not trade more data and more parameters for performance. Use structured modelling to organise the complexity.
Anti-compute-monopoly
Key capabilities must not depend on a handful of supercompute centres. Push for reproducible, deployable, accessible intelligence infrastructure.
Set the law before the compute; set the structure before the scale.
Philosophical bridge · Modelled on Yin-Yang, harmonised with Shu-Shu
Modelled on Yin-Yang, harmonised with Shu-Shu.
Modern mathematics, statistics, and computer science give us a powerful Shu — computation and optimisation. But Shu alone tends to settle for correlational fitting, never touching mechanism or intervention. We use Eastern image-thinking and systems-thinking as the constraint, and pin the Dao to computable structure.
Yin-Yang
Dual mechanisms and dynamic balance inside a system — inner/outer, cold/hot, deficiency/excess, ascending/descending; equilibria and perturbations.
Shu-Shu
Computable inference and decision — probability, logic, dynamical systems, causal inference.
Turn experience into evidence, turn syndrome differentiation into a traceable reasoning chain, turn therapeutic effect into a comparable outcome.