Use this for: Quick reference, stakeholder conversations, design reviews.
The Core Insight
Two frames, one architecture:
| Frame | Purpose | Question It Answers |
|---|---|---|
| Layers | Organization | "Where does this code belong?" |
| Invariants | Mission | "Does this prevent the worst day?" |
The Five Invariants
These must hold everywhere. They prevent the structural failures SIL committed to fix.
| Invariant | Prevents | Enforced By | Status |
|---|---|---|---|
| Everything has lineage | Epistemic collapse | GenesisGraph | Production |
| Reasoning is inspectable | Black box dictatorship | Reveal | Production |
| Computation is grounded | Hallucinated science | Morphogen | Prototype |
| Contracts are explicit | Brittle complexity | Agent Ether | Gap |
| Efficiency is sustainable | Compute as privilege | Beth + Reveal | Production |
The Chief Scientist Test
Before approving any architectural decision, ask:
- Lineage: Does this maintain traceable provenance?
- Transparency: Is reasoning inspectable?
- Grounding: Is computation connected to reality?
- Contracts: Are assumptions explicit?
- Efficiency: Is this sustainable at scale?
- Agency: Does this keep humans as conductors?
If any answer is "no" or "unclear," revise the architecture.
The Provenance-First Stack
For organizational questions ("where does this go?"):
L6: Reflection - Learning from execution (observability)
L5: Execution - Doing work under constraints (agents)
L4: Composition - Cross-domain integration (Pantheon IR)
L3: Intent - What we're accomplishing (contracts)
L2: Trust - Who can do what (authorization)
L1: Meaning - Embeddings, types, similarity (Beth)
L0: Provenance - Everything has lineage (GenesisGraph)
───────────────────────────────────────────────────────
L-1: Substrate - Physical/computational reality (optional)
Key Phrases
For funders/partners:
"We build transparent infrastructure for AI—semantic foundations that make reasoning visible, provenance verifiable, and intelligence grounded in reality."
For technical audiences:
"Use layers for organization, invariants for mission. Five guarantees that must hold everywhere."
The differentiator:
"AGI as colleague, not god. Integrated into human trust frameworks, not transcending them."
Quick Reference
| If you need to... | Go to... |
|---|---|
| Understand why SIL exists | THE_FORK.md |
| See the full invariants proposal | INVARIANTS_OVER_LAYERS.md |
| Navigate reading paths | SYNTHESIS_MAP.md |
| See layer model comparison | LAYER_MODELS_COMPARISON.md |
| Check implementation status | models/README.md |
The One Sentence
SIL builds the semantic substrate that makes AI transparent, grounded, and trustworthy—preventing epistemic collapse, not by stopping AI, but by building the steel beneath the wood.
Last updated: 2025-12-15 (turbulent-current-1215)