What works / Ready :
- Continuity Engine
- HDC VSA integration
- Core Substrate
- Semantic Translator
- Self Governance
- Memory Inheritance
- Action gating
- Trust Governance
- Multi-graph synchronization
- Interpretability
- Internal optimization (not self evolve - Continuity Engine control)
- Resource Efficiency control
- Collective alignment
- Semantic memory families
- Encapsulation method (internal ledger)
Next phase:
- Self-pruning
- Curiosity
- Node and edge schemas (extended)
- Graph-first cognition (skeleton)
Goal:
- Encapsulated HDC VSA substrate (closed SI membrane)
Potential uses:
- Observations how cognition develop
- Additional Layer for LLM
Operation mode:
- Continuous (attention based)
Business plan:
- Open License once finished
Requirements:
- CPU (GPU only as option to speed up calculations but not even needed at any stage)
Github:
- No
AI support:
- Codex CLI (coding)
- chatGPT (engineering issues)
- Gemini (cognition issues)
Build plan and project phases:
- Plan in full ready for gradual implementation
- Current phase 7 out of 32
Demo version:
- Not planning (membrane must be closed to work)
Additional help or investment needed:
- Not required
I work with Cognition and Primitives.
This work is not a search for better LLM or a speech model.
Updates will be either here or below in comments.
UI screenshots will be posted after phase 12 is done (currently at phase 7).
AI used to help me to organise my plan and verify stages: chatGPT and Gemini.