Open Models
Foundation and task-specific models with meaningful access to training method, weights or both, released under usable licenses.
A coordination layer for the open AI commons
ORÆ Open Intelligence Initiative
Funding and coordinating the open AI commons.
- Program in formation
- Open AI commons
- 2026
The ORÆ Open Intelligence Initiative is a collaboration and funding program for open AI systems, datasets, evaluation tools, agents and governance infrastructure — designed to coordinate, finance and remember the public-good layers the open AI ecosystem still lacks.
- Funding turns intent into auditable infrastructure.
- Coordination connects maintainers, datasets and reviewers.
- Public memory keeps deliverables reusable, not forgotten.
Vision
An intelligence layer for the next civilization.
ORÆ connects intent, memory, simulation, creation and coordinated action — with every step open to inspection and challenge.
Long-term research and design vision. Not a deployed product.
ORÆ does not replace human will.
It gives it an architecture.
For humanity's new beginning — this time, with intelligence as an ally.
White Paper v0.1
A conceptual framework for intent, memory and coordinated action.
Not a launched system. A serious narrative for what a civilization-scale intelligence layer would need to become trustworthy.
Reading
ORÆ does not come from nowhere.
Its principles sit at the crossing of technical alignment, governance of complex systems, inspectable interfaces and non-dystopian narratives of civilizational intelligence. None of the works below endorse ORÆ — they are the library the framework is built against.
Why a commons
Open AI does not only need more models.
It needs better coordination, long-term maintenance, transparent funding, reusable datasets, trustworthy evaluation tools and public memory. The ORÆ Open Intelligence Initiative is built to coordinate these missing layers — not to replace open source projects, but to fund and connect the public-good infrastructure they depend on.
Models without maintainers fade. Datasets without licensing clarity rot. Evaluation tools without coordination diverge. Coordination is itself a public good — and one nobody is currently funding at the right scale.
ORÆ Grants — what we fund
Six layers of the open AI commons.
ORÆ Grants is the Initiative's funding arm. Concrete, fundable, auditable — each layer is a public good worth maintaining.
Open Datasets & Memory
Reusable, well-documented datasets and long-term memory infrastructure with clear provenance and usable licensing.
Evaluation, Safety & Governance
Benchmarks, evaluation harnesses, safety tooling and governance frameworks the broader ecosystem can rely on.
Open Agents & Coordination
Agent frameworks, protocols and coordination tools designed for interoperability rather than lock-in.
Maintenance & Critical Dependencies
Time-bound support for maintainers of the libraries, runtimes and infrastructure that hold the ecosystem together.
Documentation & Reproducibility
Documentation, tutorials, reproducible recipes and onboarding paths — the unsung infrastructure of a healthy commons.
How it works
From proposal to public deliverable.
Each supported project is turned into a public, fundable and auditable roadmap: purpose, architecture, license, risks, milestones, budget, deliverables and maintenance plan.
-
01
Apply
Submit a project with a clear purpose, openness level and rough milestones.
-
02
Structure
Genesis helps turn the proposal into an auditable roadmap with explicit licenses, risks and deliverables.
-
03
Review
Human review against public criteria. Conflicts of interest are declared. Decisions are logged.
-
04
Fund
Selected projects receive funding matched to their format — micro-grant, builder grant, fellowship or strategic grant.
-
05
Build
Builders ship against public milestones. Status updates are recorded — successes and slippages alike.
-
06
Publish & Remember
Deliverables, licenses and the project's full record become part of the ORÆ Atlas — usable, citable, maintainable.
ORÆ Genesis
The project-structuring layer of the ORÆ Initiative.
Genesis is the system designed to transform intentions into roadmaps, milestones, memory and accountable execution. Within the Initiative, it helps turn open AI proposals into clear, fundable and auditable project records.
Genesis is a prototype direction inside the Initiative. It is not currently a public product.
For builders & funders
Two sides of the same commons.
A coordination layer is only useful if both sides show up.
For builders
Turn your open AI project into a fundable, auditable roadmap.
If you maintain a library, train an open model, curate a dataset, write evaluation tools or build open agents — apply with a project. We help structure the work, secure funding and keep the public record clean.
Apply with a projectFor funders
Support open AI infrastructure with transparent selection and public reporting.
If you fund research, public-good infrastructure or AI safety — partner with ORÆ Open Intelligence Initiative to support thematic calls, maintenance, evaluation tooling or strategic initiatives, with measurable deliverables and a public record.
Partner with ORÆGovernance
Power requires audit, consent and limits.
A civilization-scale intelligence layer must be auditable before it is powerful.
- Coercive manipulation
- Mass surveillance without consent
- Autonomous destructive action
- Governance evasion
- False claims of operational AGI
- Auditability
- Traceable memory
- Human contestability
- Reversible decisions
- Declared uncertainty
- Staged deployment
Values designed in
Values are not slogans. They are constraints on what ORÆ should be allowed to become.
Æ also carries the symbolic reference to Adam & Eve — origin, duality, complementarity and balance.
Openness levels
Three explicit categories of openness.
"Open source" is too often used loosely. ORÆ distinguishes between three categories — so funders, builders and users know exactly what kind of openness they are dealing with.
Systems that can be used, studied, modified and shared.
With meaningful access to the preferred forms of modification — including training data, training code, weights and documentation where applicable.
Model weights are available.
Training data, full method or licensing may be partially restricted. Usable and inspectable, but not fully reproducible from scratch.
Datasets, benchmarks and protocols.
Documentation, tools and research artifacts that strengthen the open AI ecosystem without necessarily being models themselves.
Who should join
A civilizational project recruits across disciplines.
Each discipline carries part of the architecture.
Collaboration
Help build the intelligence layer.
Shape the research, product, governance and culture around advanced AI.
The profiles ORÆ is actively looking for. The disciplines are already named in the people section above — this is the shortlist of roles each discipline can fill.
Public log
Drafted in public means this exists.
Every entry is a draft decision, not a feature release.
Anticipated objections
Hard questions, answered without pirouette.
Anticipated, not avoided.
Funders & strategic partners
Fund the open AI infrastructure your ecosystem depends on.
The ORÆ Open Intelligence Initiative is being prepared as a collaboration and funding program for open AI. The first phase focuses on partner alignment, public roadmap design and project intake.
See partnership formatsHelp build
Help build the open intelligence commons.
The Initiative is in formation. The work ahead is concrete: open project intake, public selection criteria, transparent decision logs and a Public Atlas of the systems, datasets and tools that hold the open AI ecosystem together.
Contact
Register your interest.
Submissions stored locally for now. Ready for future API integration.
