Plural by design
Disagreement across schools, cultures, languages, and generations remains visible instead of being flattened into a single synthetic answer.
Independent initiative · Design stage
A governed intelligence commons for the Muslim world.
IDO is a proposal for a public-interest institution that helps scholars, civic leaders, and communities work across sources, languages, and disagreement. Its AI tools would keep evidence visible and remain answerable to people.
Design principle
“The model is replaceable. The institution's memory is not.”
StatusInstitutional concept
Technical foundationWorking knowledge graph in MindGraph
Next stepOne governed pilot on a specific question
The thesis
The challenge is not a shortage of information. It is the absence of shared, durable context.
Islamic intellectual and civic life spans centuries of sources, many languages, multiple legal and theological traditions, and radically different political contexts. Ordinary search can find fragments. A chatbot can summarize them. Neither creates an accountable institutional memory.
IDO asks whether plural communities can build a shared memory that keeps evidence, disagreement, decisions, and change connected over time. AI could then support human deliberation without being treated as the authority.
Constitutional principles
Governance is part of the architecture.
The institution should be designed around the limits of AI, the legitimacy of disagreement, and the right of communities to inspect and correct what is held about them.
Evidence before synthesis
Sources, claims, interpretations, and counterarguments remain inspectable so every briefing can show how it was assembled.
Memory belongs to the institution
Knowledge and judgment live in a governed, portable substrate. Models may change without taking the institution's memory with them.
People retain authority
AI may organize, retrieve, compare, and draft. Scholars and accountable institutions decide what is trusted, published, or acted upon.
Technical foundation
A living memory, not another all-knowing model.
MindGraph provides a working system for keeping linked sources, interpretations, decisions, and revisions outside any one AI model. That makes a governed, model-independent institutional memory technically credible for IDO.
Knowledge stays outside the model. A general-purpose model can be swapped without losing institutional context.
Maintenance stays accountable. Any AI system that organizes or updates the graph must be versioned, audited, replaceable, and unable to approve its own changes.
Complexity must earn its place. Small collections should stay simple. Structured memory becomes useful when sources, state, and revision history outgrow flat files.
Institutional memoryGoverned
01SourcesTexts, testimony, policy, scholarship
02InterpretationClaims, evidence, dissent, provenance
03DeliberationQuestions, options, constraints, decisions
04ContinuityWhat changed, why, and what comes next
Reasoning model AReasoning model BHuman review
One durable memory · Multiple replaceable readers
Evidence boundary
MindGraph provides a working technical substrate and research on specific memory tasks. This does not show that AI can resolve conflict, adjudicate doctrine, or improve public institutions. IDO must earn those claims through domain-led pilots and real-world evaluation.
A credible first mandate
Begin with tools that help people make better judgments.
The first IDO pilot should be useful even if a person reviews every generated sentence.
Source-grounded research
For example, a briefing on one policy question could connect primary material, scholarship, and competing readings while preserving citations.
Institutional memory
A durable record of decisions, open questions, commitments, and evidence that survives staff turnover and model changes.
Deliberation support
Tools for mapping arguments, finding precedents, and exposing unresolved assumptions while leaving judgment to people.
Proposed sequence
Legitimacy before scale.
No country list, adoption target, or geopolitical promise substitutes for a trusted process. The work should advance only as evidence and governance mature.
Convene
Form a small, representative design council spanning scholarship, civil society, technical practice, and youth.
Choose one specific question
Start with a research or record-keeping workflow where the sources, permissions, and intended outcome can be evaluated clearly.
Build under review
Use only material provided with consent, preserve its sources, expose uncertainty, and require human approval before adding anything to the institutional record.
Earn the right to scale
Publish the evaluation, failures, governance record, and correction process before expanding the collection or mandate.
Non-negotiable boundaries
Ambition needs a constitution.
IDO should be judged as much by what it refuses to automate as by what it builds.
- Not a fatwa engine or substitute for religious authority
- Not an autonomous diplomatic or policy decision-maker
- Not a single model trained to contain a community's private memory
- Not a claim to erase legitimate disagreement
- Not a data collection program without consent, rights, and redress
An invitation
Help shape the charter before we shape the technology.
We want to hear from scholars, librarians, civic institutions, technologists, funders, and community organizers who care about plural knowledge, accountable AI, and institutional memory.
Start a conversation