A user-authorized, evidence-based framework for connecting AI, software, physical-world knowledge, resources, products, supply chains, and historical context.
This repository is the evolving home of two connected projects:
- ORBIT Language β a provider-neutral language and semantic layer for describing computation, data, locations, objects, relationships, evidence and AI tasks.
- Global World Knowledge β a structured world-observation system connecting coordinates, buildings, roads, products, prices, resources, production, transport, recycling and history.
The project is designed to let humans and authorized AI systems observe the same world through a common, versioned vocabulary while preserving evidence and uncertainty.
ORBIT can maintain a User Knowledge Graph from information that the user explicitly authorizes or imports.
Authorized Sources
β
Permission / Consent Layer
β
Import / Connector
β
Privacy & Redaction
β
User Knowledge Graph
β
ORBIT Semantic Layer
β
AI Models / Tools
Possible sources include:
- ChatGPT data exports
- GitHub repositories
- local documents
- project files
- notes
- calendars
- photos and media metadata
- other AI conversation exports
- explicitly connected services
No component should silently read a user's entire device, account, private messages or cloud storage. Data must be supplied through an explicit export, connector, permission or import process.
The system should record the state of every source as:
NOT_CONNECTED β REQUESTED β AUTHORIZED β IMPORTED β INDEXED β REVOKED
Revocation must prevent future ingestion and should support deletion of derived indexes where applicable.
The project now includes registry specifications for:
- UN / ISO / WIPO / IANA / ICANN / ITU / IEEE source layers
- countries, jurisdictions and official-source registries
- banks and banking regulators
- currencies and exchange infrastructure
- exchanges, regulators, main boards and securities identifiers
- global payment rails
- enterprises and corporate hierarchies, including the Top-100,000 coverage target
- elements, materials and manufacturing
- electronics brands and the EMBED device language
- genomics, proteins, species, biodiversity and biobanks
- CRISPR and adjacent life-science fields
- products, prices and supply chains
- GIS / physical-world observations
- evidence, provenance and historical change
See docs/PROJECT_DOCUMENT_STATUS_V0.3.md for the canonical document/status index.
Sensitive information should be minimized and never committed to a public repository without an explicit decision to publish it.
The world model connects:
coordinate β space β building β people β roads β commerce β product β price β production β supply chain β resources β lifecycle β recycling β history
See GLOBAL_WORLD_KNOWLEDGE.md and docs/GLOBAL_KNOWLEDGE_MASTER_PLAN_V0.2.md.
ORBIT aims to unify concepts rather than merge unrelated source code or tokenizers. Provider-specific languages remain native behind adapters.
Swift / C / C++ / Rust / Python / JavaScript / SQL / ML systems / GIS
β
ORBIT semantic layer
β
AST β ORBIT-IR β backends
The executable prototype is under orbit/:
orbit/lexer.pyβ lexer prototypeorbit/parser.pyβ parser prototypeorbit/ast.pyβ normalized ASTorbit/checker.pyβ identifier/evidence checks
The current parser grammar is docs/ORBIT_GRAMMAR_V0.2.md. The older v0.1 grammar is retained as a historical draft.
See docs/ORBIT_LANGUAGE_SPEC_V0.1.md and docs/ORBIT_ARCHITECTURE_V0.2.md.
AI providers connect through adapters rather than receiving unrestricted access to each other or to user data.
Siri / Apple Intelligence / Local AI / Cloud AI
β
ORBIT Bridge
β
Semantic Message
β
Router + Evidence
β
Global Knowledge
See docs/SIRI_AND_MULTI_MODEL_BRIDGE.md.
The project distinguishes provider billing tokens from the cross-provider OSU (ORBIT Semantic Unit) workload measure.
User task β quote β authorization β AI work β usage β cost β ledger β settlement
No token is treated as money merely because software assigns it a name. Monetary settlement requires appropriate authorization, contracts, payment infrastructure and applicable laws.
See docs/ORBIT_ROUTER_LEDGER_V0.1.md.
ORBIT is intended to optimize for the hardware actually available rather than pretending software can turn one physical chip into another.
The first hardware research target is Intel x86-64, with future Apple Silicon and other backends.
See docs/ORBIT_INTEL_ACCELERATION.md.
The repository contains a recovered thematic index of long-running ChatGPT discussions. It is not a complete account export.
See docs/CONVERSATION_ARCHIVE_INDEX.md and docs/CHATGPT_ARCHIVE_PIPELINE.md.
A true historical archive requires the user's own ChatGPT data export or conversation files.
Every important claim should distinguish:
- Observed
- Sourced
- Derived
- Inferred
- Unknown
Forensic-inspired reasoning follows:
observation β evidence β hypothesis β corroboration β alternatives β conclusion β revision
AI agreement is not proof. Correlation is not automatically causation.
.
βββ README.md
βββ GLOBAL_WORLD_KNOWLEDGE.md
βββ examples/
βββ schemas/
βββ docs/
β βββ PROJECT_DOCUMENT_STATUS_V0.3.md
β βββ GLOBAL_KNOWLEDGE_MASTER_PLAN_V0.2.md
β βββ GLOBAL_COMPLETION_AUDIT_V0.1.md
β βββ GLOBAL_DATABASE_UPDATE_V0.2.md
β βββ GLOBAL_COUNTRY_EXCHANGE_CURRENCY_REGISTRY_V0.1.md
β βββ GLOBAL_BANK_REGISTRY_V0.1.md
β βββ GLOBAL_EXCHANGE_REGULATOR_MAINBOARD_REGISTRY_V0.1.md
β βββ GLOBAL_PAYMENT_RAILS_REGISTRY_V0.1.md
β βββ GLOBAL_TOP_100000_ENTERPRISE_REGISTRY_V0.1.md
β βββ COUNTRY_LICENSE_REGISTRY_SCHEMA_V0.1.md
β βββ GLOBAL_BIOLOGICAL_KNOWLEDGE_REGISTRY_V0.1.md
β βββ BIOTECH_CRISPR_AND_LIFE_SCIENCE_REGISTRY_V0.1.md
β βββ ELECTRONICS_BRAND_REGISTRY_AND_EMBEDDED_LANGUAGE_V0.1.md
β βββ AI_OBSERVATION_PROTOCOL.md
β βββ ORBIT_LANGUAGE_SPEC_V0.1.md
β βββ ORBIT_ARCHITECTURE_V0.2.md
β βββ ORBIT_GRAMMAR_V0.1.md
β βββ ORBIT_GRAMMAR_V0.2.md
β βββ ORBIT_INTEL_ACCELERATION.md
β βββ SIRI_AND_MULTI_MODEL_BRIDGE.md
β βββ ORBIT_ROUTER_LEDGER_V0.1.md
β βββ CONVERSATION_ARCHIVE_INDEX.md
β βββ CHATGPT_ARCHIVE_PIPELINE.md
βββ orbit/
β βββ lexer.py
β βββ parser.py
β βββ ast.py
β βββ checker.py
β βββ README.md
βββ tools/
βββ import_chatgpt_export.py
- World Knowledge specification
- Observation schema
- Product/price schema
- AI observation protocol
- ORBIT language specification
- Intel optimization strategy
- Multi-model bridge concept
- Token/value ledger concept
- User Knowledge permission model
- Global registry architecture
- ORBIT lexer prototype
- ORBIT parser prototype
- Normalized AST prototype
- Basic identifier/evidence checker prototype
- type and unit checker
- entity resolver
- full evidence/provenance checker
- permission engine
- ORBIT-IR
- provider adapters
- hardware capability probe
- token/work meter implementation
- ledger implementation
- conformance test suite
- image observation pipeline
- GIS layer
- product recognition
- price normalization
- supply-chain graph
- historical timeline
- knowledge graph
- data-gap engine
- conflict-resolution engine
- UN/ISO country master
- official-source registry for all jurisdictions
- central-bank / banking-regulator adapters
- exchange/regulator/main-board adapters
- payment-rail adapters
- enterprise ranking/source adapters
- materials/electronics sources
- biological/scientific sources
- per-country coverage reports
- ChatGPT export ingestion
- additional AI export adapters
- Apple/Siri App Intents prototype
- local knowledge index
- consent dashboard
- revocation and deletion workflow
- dashboards
- machine-readable APIs
- source/provenance viewer
- global data-gap map
- security/legal audit
- reproducible releases
- User control before automation.
- Evidence before certainty.
- Open standards before vendor lock-in.
- Interoperability without unrestricted access.
- Privacy by design.
- Version everything that changes.
- Separate observation from inference.
- Separate workload accounting from money.
- Build experimentally and measure real performance.
- Make the system useful to people before making it powerful.
- Never claim global completeness until the underlying records are actually verified.
Project licensing and third-party component notices must be defined before redistributing third-party source code. See each upstream project's license and NOTICE requirements.