Good ideas can come from anywhere
Knowledge should move in every direction. Communities participate as contributors with something to teach, not as case studies for somebody else.
A worldwide, open-source project that keeps evolving
The world needs
a better plan.
The Plan brings experts and communities together to rethink the systems we all depend on, from education and healthcare to energy, housing, and justice. A better plan has to work for people without making the planet less livable.
Share what we know. Build what works where you live.
Not one answer for the whole world. It is a better way to work on the future together.
THE MODEL
LIVING
THE
PLAN
v∞
01Imagine
02Test
03Connect
04Adapt
01
The premise
The systems we rely on are connected.
We should plan that way.
Education, healthcare, energy, housing, justice, food, technology, and the environment all shape one another. But we usually try to fix them one at a time, through institutions that rarely work together.
The Plan starts with a simple question: given what people around the world already know, what would we build if we were allowed to start fresh?
Connect the systems, because changing one almost always changes another.
Learn from the whole world, without treating any one country or culture as the default.
Keep decisions local, so communities help write their own future instead of receiving somebody else’s answer.
02
Working across the world
There isn’t one right future
for every place.
This is not about exporting one society’s version of progress. It is a way for people in different places to share what they know, learn from one another, and build plans that make sense in their own lives.
WHAT WE ARE REALLY PLANNING FOR
Not an institution.
Not a field.
Not a country. The whole system people live in.
Not a field.
Not a country. The whole system people live in.
Different places need different answers
People share many of the same needs, but that does not mean every place should end up with the same institutions, culture, or way of life.
The future shouldn’t be proprietary
The plans, evidence, tools, and lessons should be open for anyone to inspect, challenge, translate, adapt, and improve.
Don’t solve today by making tomorrow worse
Every proposal has to consider who pays the price later. That includes other communities, future generations, and the natural world.
Deep expertise without the silos
Experts need room to go deep. The work still has to connect.
Educators should be able to focus on education, and energy experts on energy. But before a plan moves forward, we need to know what it depends on, what else it changes, who could be harmed, and where it creates new possibilities.
Education
Health
Energy
Justice
Ecology
Housing
SHARED
FUTURES
FUTURES
01Start with the field
Experts begin with what they know best and show how they reached their conclusions.
02Find the overlaps
The process keeps track of what each plan needs, changes, risks, or makes possible elsewhere.
03Bring the right people together
When plans touch, people from those fields work through the overlap together.
04Check the whole
A major proposal is not finished until the other systems and communities it affects have had a chance to challenge it.
03
What we won’t compromise on
A few commitments
to keep us honest.
⌁
Look at the whole system
Nothing exists on its own. Every plan should show what it affects, what it relies on, and where feedback or unintended consequences could appear.
⊹
Start with deep expertise
Give people who know a field well enough room to work through it carefully before combining their ideas with everything else.
◌
Share what we know. Let places decide.
We can learn together without forcing every community to make the same choices. People should be able to adapt or reject what does not fit.
≋
Expertise has boundaries
Being an expert in one area does not make someone an expert in every area. Good experts know when to bring in other kinds of knowledge.
↻
Different paths can still be good paths
The goal is not to make the world the same. Different cultures and communities can build different versions of a good future.
◇
Keep disagreement on the page
Agreement can help us move forward, but disagreement tells us something too. Important minority views should remain easy to find.
⟲
Make the work public
The methods, evidence, changes, and failures should be open so other people can inspect them and build on them.
∞
Change when the evidence changes
No version is final. Better evidence and real-world results should be able to change what the network believes.
◈
Don’t push the cost somewhere else
A plan is not successful if it helps one group by pushing harm onto another place, the living world, or future generations. Sustainability is a requirement, not an extra.
04
How it works
Start fresh. Go deep.
Then see what it touches.
Starting from a blank slate does not mean ignoring evidence, ethics, culture, or practical limits. It means we do not assume today’s institutions are the only way things could work.
01
Imagine
Imagine it from scratch
Begin with the outcomes we actually want, then design the system around them instead of protecting the institutions we already have.
What is the system for?
What assumptions can we release?
02
Deepen
Let experts work it through
Let people with relevant experience work through the evidence, roles, resources, trade-offs, and open questions inside the field.
What does this field know deeply?
Where is its uncertainty?
03
Weave
Check what else it changes
Map what the idea depends on and what it changes across other fields, communities, and the natural world.
Which systems does this touch?
Who is missing from the room?
04
Localize & learn
Make it work somewhere real
Adapt the plan to a real place, try it, and share what worked and what did not with everyone else.
What must be true here?
What can everyone learn?
↔
The zoom rule
Every group has to move back and forth between its piece and the bigger picture.
Go inward far enough to earn rigor. Zoom outward far enough to see consequence. Return inward with new constraints and possibilities. Repeat until the plan can explain both itself and its place in the larger world.
What do we depend on?
What do we change?
What could we damage?
What could we unlock?
Which places differ?
Who carries the cost?
05
The plan artifact
A real plan is more
than another report.
DESIGN QUESTION 01
What should learning make possible?
CONFIDENCE82%
Define the human outcomes the system exists to serve before discussing schools, schedules, curriculum, or credentials.
PROPOSED OUTCOMEAgency and lifelong adaptabilityHigh agreement · 84 expert reviews
KEY ASSUMPTIONLearning happens beyond institutionsUnder active challenge · 12 responses
MINORITY VIEWPrioritize shared civic knowledge18% support · Read the case →
Version controlled
Evidence linked
Systems mapped
Dissent preserved
Regionally forkable
Openly licensed
06
Participation system · early PRD 0.1
Experts test the work.
Affected people ground it.
The phone app is meant to help people make better decisions, not give everyone another place to post. Verified experts review the parts they know well. Affected people and public participants share needs, experiences, priorities, and concerns in a structured way. Local decision makers still decide what to adopt, and authorized organizations remain responsible for carrying it out.
The whole platform in one view
From a panel’s first draft to locally authorized action.
The mobile experience is one part of a larger loop that separates authorship, review, lived experience, local choice, and execution.
- 01 · FRAMEA provisional panel begins
Field experts publish the scope, outcomes, boundaries, gaps, and first draft.
- 02 · OPENThe expert network grows
Invitations, applications, and nominations feed independent, scoped verification.
- 03 · REVIEWVerified experts challenge the work
The app supports evidence, amendments, structured judgment, and visible dissent.
- 04 · LISTENAffected people ground it
People share needs and experience, test trade-offs, and request plans for unmet needs.
- 05 · RATIFYA credible reference emerges
Relevant expert communities establish confidence without creating a global mandate.
- 06 · ADOPTA place chooses its own version
A community or institution adapts the plan and authorizes accountable local owners.
- 07 · EXECUTESoftware helps organize reality
Locally controlled tools coordinate roles, budgets, schedules, approvals, and work.
- 08 · LEARNResults change the shared plan
Real-world evidence returns to the commons and may confirm, amend, or retire the plan.
PRODUCT CONTRACT 06.1
Different kinds of legitimacy should never be mashed into one score. Expert judgment, lived experience, local consent, and operational authority answer different questions and must remain visible on their own terms.
LANE 01Identity + expertise verified
E
For relevant practitioners and researchers
Expert peer review
Let qualified people challenge the work in a focused, evidence-based way, even if they were not part of the original panel.
01
MatchReceive review requests within a verified scope of competence.
02
ExamineSee the claim, reasoning, evidence, dependencies, and affected places.
03
JudgeScore multiple dimensions, state confidence, flag conflicts, and explain concerns.
04
AmendPropose a specific change or alternative that authors must address.
05
RevisitReview the response and record whether the revision resolved the issue.
PRIMARY OUTPUTAn inspectable peer-review record
LANE 02Open, safe, place-aware
P
For people living with the consequences
Public listening and lived experience
Hear from people at a meaningful scale without creating posts, followers, pile-ons, viral arguments, or a contest for attention.
01
GroundChoose the place, system, or life experience relevant to the prompt.
02
ExpressShare needs, experiences, priorities, trade-offs, and optional narrative context.
03
ProtectUse clear consent and anonymous, protected, or attributed participation as appropriate.
04
AggregateSee patterns, distributions, minority needs, and regional differences rather than individual posts.
05
Close the loopLearn what authors heard, what changed, and what they declined to change.
PRIMARY OUTPUTA structured public-needs signal
Four questions, four kinds of legitimacy
Be clear about who knows what and who gets to decide.
IS IT CREDIBLE?Expert legitimacy
Relevant verified experts test evidence, causal logic, feasibility, and risk.
Expert reviewWHAT WILL IT DO?Lived legitimacy
Affected people reveal needs, harms, constraints, and realities formal evidence can miss.
Public listeningSHOULD WE CHOOSE IT HERE?Local legitimacy
A community or institution weighs values, trade-offs, fit, and consent in its own context.
Local adoptionWHO MAY ACT?Operational legitimacy
The implementing organization assigns permissions, accepts responsibility, and authorizes spending and action.
Execution authority
THE GOVERNED BRIDGE
Asking people only matters if their answers can change the work.
- 1
AskA decision-linked prompt states what is open to influence.
- 2
CollectExpert and public signals are gathered independently.
- 3
InterpretMethods, uncertainty, missing groups, and distributions remain visible.
- 4
RespondAuthors revise, investigate, or explain why they will not change course.
- 5
AuditParticipants can see the path from input to outcome.
Deliberate non-goals
The app is not trying to keep people scrolling.
Success means better decisions, better representation, and clear follow-through, not more time on screen.
- No public posts, replies, follower counts, or viral sharing loops
- No popularity totals before a person contributes their own judgment
- No single score that blends expertise, preference, and lived experience
- No advertising, sale of participant data, or behavioral targeting
- No AI deciding what the public “really means” or manufacturing consensus
- No promise that every input becomes policy. Material input still receives a visible response.
FIRST PRODUCT TESTOne education proposal
EXPERT SIDEVerify · review · amend · judge
PUBLIC SIDENeed · experience · priority · trade-off
SHARED PROOFA visible response ledger
07
Governance
Different people know different things.
The process should make that clear.
Broad participation matters, but not every question should be settled by one giant vote. Experts should lead where deep knowledge matters. Communities should lead on what works and feels legitimate where they live. The larger network should help everyone learn from the results.
01Provisional steward
The core field team
A deliberately balanced starting group responsible for producing depth and coherence without claiming ownership of the field.
- Frame the design question
- Synthesize contributions
- Publish draft versions
- Declare cross-system dependencies
PRIMARY POWERStewardship
02Verified network
The wider expert network
Practitioners and researchers across regions whose authority is scoped to demonstrated areas of competence.
- Review specific sections
- Submit evidence and amendments
- Join domain and interface groups
- Cast domain-relevant judgments
PRIMARY POWERChallenge
03Lived expertise
People living with the system
People who experience the system directly and can expose needs, effects, and blind spots.
- Define priorities and harms
- Contribute lived evidence
- Evaluate acceptability
- Author local adaptations
PRIMARY POWERGround truth
04Open access
Anyone who wants to take part
Anyone can follow the work and contribute structured signals about needs, experience, priorities, and acceptability.
- Read and follow plans
- Request a missing plan or unanswered question
- Answer decision-linked prompts
- Signal needs and priorities safely
- See how input affected the work
PRIMARY POWERLegitimacy & accountability
Panel formation
Invitations can start the network.
They cannot define its electorate.
The first panel is a provisional steward. It can invite trusted peers, but participation must also open through applications, nominations, independent verification, and legitimate competing branches.
01
Open more than one doorInvitations, applications, and nominations all feed the expert network.
02
Verify independentlyThe panel does not decide by itself who qualifies to review its work.
03
Publish the shape of the roomFields, regions, practice settings, affiliations, and missing perspectives remain visible.
04
Allow legitimate branchesDissenting schools of thought can develop inspectable alternatives instead of fighting for one official answer.
A layered trust model
Verification tells us who someone is.
Their work tells us how much to trust their judgment.
Expertise is not a permanent yes-or-no label. It should be tied to a specific field, kind of decision, practice context, place, and review date, then strengthened or challenged by the person’s actual contributions.
✓
Identity verifiedPersonhood and professional identity
BASE✓
Expertise scopedField, subfield, decisions, place, and practice context
DOMAIN↻
Verification currentEvidence and credentials reviewed on a published schedule
TIME↗
Contribution recordEvidence quality and peer assessment
DYNAMIC◎
Conflicts disclosedFinancial and institutional interests
PUBLICThe connective tissue
When two plans touch, the right people work through it together.
These are temporary working groups, not another permanent layer of bureaucracy. They form when, for example, an education proposal affects health or an energy plan changes housing costs.
DEPENDENCY
One system cannot function without capacity from another.
EXTERNALITY
A benefit or cost is displaced into another system or place.
TENSION
Two otherwise sound plans pull shared resources or values in different directions.
OPPORTUNITY
Coordinated action can create an outcome no discipline can reach alone.
08
Decision-making
Don’t just count votes.
Show why people voted that way.
Simple upvotes tend to reward confidence, popularity, and team loyalty. Here, an amendment has to make its reasoning inspectable, and relevant experts judge it on several questions instead of reducing everything to one reaction.
- 1The amendment names the problem, evidence, assumptions, dependencies, expected effects, and risks.
- 2Objections, alternatives, conflicts of interest, and missing perspectives are collected.
- 3Authors revise, withdraw, branch, or respond on the record.
- 4Relevant domain and interface experts record scoped judgments and confidence.
- 5Consensus, uncertainty, regional difference, and minority positions are published together.
PROPOSAL E-17
128 reviews
Replace seat time with demonstrated mastery
76expert confidence
EMERGING EXPERT CONSENSUS
Strong technical support, with material implementation concerns.
18%
Minority report attachedAssessment reliability may recreate existing inequities.
RULE 07.1
Relevant expertise can help us understand a result. But the original, unweighted numbers should always be visible too.
Ratification is a ladder, not a stamp
Credible enough to share is not the same as authorized to act.
The worldwide network can establish confidence in a reference plan. A community or institution must separately decide whether it fits, adapt it, and authorize the people who will carry it out.
- 01DraftOpen claims, evidence, and alternatives
- 02Expert-ratifiedRelevant technical review completed
- 03System and impact reviewedDependencies and affected people addressed
- 04Available for local adoptionA credible reference, not an instruction
- 05Locally adapted and authorizedFit, consent, ownership, and responsibility established
- 06In use and evaluatedResults can confirm, challenge, or change the plan
THE AUTHORITY BOUNDARY
Expert ratification establishes confidence. Local adoption establishes legitimacy in a place. Operational authorization establishes who may spend and act.
09
From plan to practice
Expertise should not end as advice.
It should become useful infrastructure.
An expert-ratified reference plan is not automatically ready to run. A community or institution must adapt it, decide that it belongs in its context, and authorize the people responsible for carrying it out. The Implementation Commons would then turn that locally approved version into open-source software for coordinating the work and learning from what happens.
01Build the plan
→
Experts and communities work through the problem.
02Ratify the reference
→
The network establishes technical confidence.
03Adapt it locally
→
A community fits the plan to its own reality.
04Authorize the work
→
Local owners accept responsibility and grant authority.
05Generate and use tools
→
Software turns the approved plan into usable workflows.
06Share what we learn
Results flow back into the plan for everyone.
An implementation contract
The plan needs a structured version that software can understand, not just a long document written for people to read.
- Outcomes and measures
- Roles and responsibilities
- Required steps and dependencies
- Local choices and variables
- Safety rules and stop conditions
Tools that fit the work
The interface can change based on the plan, the place, the user’s role, and the resources available. People get the tools they need instead of one giant application built for everyone.
- Step-by-step workflows
- Training and readiness checks
- Schedules, forms, and budgets
- Decision support and risk alerts
- Progress and learning dashboards
Authority and judgment
Generative UI is the interface layer, not the authority. The software helps people carry out a locally approved plan. It does not ratify the plan, grant itself permission, move money on its own, or quietly change the rules.
- Local approval and accountability
- Clear permissions
- Human review of consequential actions
- Visible sources and plan versions
- The ability to stop, inspect, and override
Federated by design
One shared standard.
Many locally controlled systems of record.
The worldwide platform should not hold every community’s money, personnel, schedules, and sensitive operational data in one central system. Each implementation needs an authoritative local record that can still exchange plans and lessons through open standards.
Reference plans, evidence, public reasoning, version lineage, reusable modules, and appropriately aggregated results.
Common schemas, provenance, exports, APIs, permissions, and connectors that prevent lock-in.
The authoritative budget, schedule, roles, approvals, actions, payments, and sensitive records for one implementation.
ARCHITECTURE PRINCIPLE
Centralized within each implementation where coordination requires it. Federated across communities where sovereignty requires it.
What this could look like
A community puts an education plan to work.
Suppose an expert-ratified education plan replaces seat time with demonstrated mastery. A local community reviews the trade-offs, adapts the plan, authorizes its own version, then uses open-source software to generate the tools needed by school leaders, teachers, students, families, and community partners.
One shared plan. Different tools for the people doing different parts of the work.
LEADERSReadiness assessmentCapacity, policy, staffing, and timeline
TEACHERSMastery workflowEvidence, feedback, and learner progress
FAMILIESClear progress viewWhat a learner knows and what comes next
COMMUNITYEquity and results checkWho benefits, who is missing, and what changed
WHY A COMMONS?
The tools should belong to the people using them.
Communities should be able to inspect the software, host it themselves, translate it, protect sensitive data, replace the underlying AI, and share useful improvements with other places.
10
Safeguards
Think through the ways
this could fail.
DESIGN RESPONSE
Transparent selection criteria, affiliation limits, complete funding disclosure, rotating seats, public vote records, and independent appeals.
DESIGN RESPONSE
Reserve participation for frontline and lived expertise, distinguish experience types, and assess contribution quality independently of prestige.
DESIGN RESPONSE
Multi-dimensional judgments, confidence reporting, minority reports, dissent prompts, and visible distributions instead of winner badges.
DESIGN RESPONSE
Time-boxed drafting cycles, defined decision thresholds, scoped questions, version releases, and a required path from recommendation to pilot.
DESIGN RESPONSE
Source provenance, citation verification, evidence quality ratings, rate limits, anomaly detection, and clear labeling of AI-assisted contributions.
DESIGN RESPONSE
Graduated identity visibility, protected review modes, rigorous conduct enforcement, threat response, and careful defaults in sensitive domains.
DESIGN RESPONSE
Mandatory dependency maps, interface assemblies, displaced-cost accounting, whole-system review, and veto-grade warnings for unresolved cross-system harm.
DESIGN RESPONSE
Polycentric governance, multilingual participation, regional authorship, explicit cultural assumptions, protected plural variants, and no single canonical implementation.
DESIGN RESPONSE
Every prompt names the decision it can influence, every material signal receives a response, and a public ledger connects input to investigation, revision, explained refusal, or later testing.
DESIGN RESPONSE
Personhood checks proportionate to risk, blinded contribution before result viewing, sampling disclosures, anomaly review, rate limits, conflict declarations, and no raw majority count presented as representative truth.
DESIGN RESPONSE
Data minimization, explicit purpose and retention rules, meaningful consent, deletion and export rights, aggregate defaults, protected narratives, community data stewardship, and no advertising or behavioral targeting.
11
Pilot roadmap
Test the process
before building the whole platform.
NEW WORKSTREAMS · GOVERNANCE + IMPLEMENTATION COMMONS
The next layer needs its own design documents and PRD sets.
We should not jump from this direction straight into product development. The governance system and the Implementation Commons each need a durable working corpus that can evolve alongside the public concept.
That work will need to define:
- Panel formation, legitimacy, and competing branches
- Scoped expertise verification and conflicts
- Amendments, expert judgment, and minority reports
- Public listening and the response ledger
- Ratification, local adoption, and plan versioning
- The machine-readable implementation contract
- Federated records, generative UI, and human authority
- Privacy, security, accessibility, and open-source stewardship
STATUSDirection establishedDedicated design documents and PRDs not yet started
PILOT DESIGN QUESTION
“If a city were creating a public learning system for ages 14–18 from scratch, what outcomes, learning model, staffing structure, and accountability system should it use?”
HOME DOMAIN Education
CYCLE 12 weeks
CORE PANEL 8–12
INTERFACES Health · Mobility · Work
Frame
Form a provisional panel. Publish scope, outcomes, boundaries, selection gaps, and decision rules.
Weeks 1–2Draft
Build the ideal model and publish assumptions as structured claims.
Weeks 3–5Connect
Invite community evidence, cross-system review, alternatives, and red-team analysis.
Weeks 6–8Decide
Revise, record structured judgments, and publish the first expert-ratified reference version.
Weeks 9–10Localize
Test whether a real place can adapt and legitimately authorize its own expression without losing lineage.
Weeks 11–12The first test
What must the pilot prove?
01
Participation qualityCan busy experts make focused, useful contributions?
02
LegitimacyDo participants trust the process even when they disagree?
03
Output qualityIs the result more coherent and actionable than a typical report?
04
System awarenessDid the process reveal consequences and possibilities beyond education?
05
Implementation pullDoes a real community want to adapt or test any part of it?
12
Open questions
The big questions
we still need to answer.
Your reactions are saved only in this browser.
Who has legitimate authority to appoint the first panel?
Selection will shape both the substance of the plan and whether outsiders regard it as credible.
Decision needed before pilot
How can a founding panel invite experts without choosing its own electorate?
Invitations can start the network, but open applications, nominations, independent verification, and appeals must prevent the panel from reproducing itself.
Needed before recruitment
When should affected communities be able to override expert judgment?
Technical feasibility and community legitimacy can conflict. The process needs an explicit rule for that boundary.
Needs structured debate
What does each ratification state actually authorize?
Technical confidence, public legitimacy, local adoption, permission to spend, and authority to act are separate decisions that need separate owners.
Define before workflow design
What is the smallest unit people should review and vote on?
A whole plan is too broad; an isolated sentence may lose the system context that gives it meaning.
Prototype during pilot
Should The Plan be a nonprofit, public-benefit company, or hybrid?
The structure must support durable funding without allowing funders to influence outcomes.
Explore before fundraising
How should relevant expertise affect the interpretation of a vote?
Weighting can improve signal, but opaque weighting could recreate the hierarchies the platform hopes to correct.
Test multiple models
When can expert identity be hidden from the public?
Transparency builds trust, but experts in politicized or sensitive fields may face real personal risk.
Policy required before launch
Who pays for a panel without purchasing influence?
Sponsors can fund questions, but governance must prevent them from funding preferred answers.
Funding charter needed
How does a shared blueprint become a legitimate local expression?
Local authorship is essential, but the network still needs to understand what changed, why, and what others can learn.
Learn from first partner
Who is responsible for consequences that cross disciplinary boundaries?
Domain assemblies can identify interdependence, but the platform needs legitimate authority to resolve cross-system harms and tradeoffs.
Design interface governance
How can one worldwide network remain truly polycentric?
Language, connectivity, funding, prestige, and institutional power could quietly recreate a center even when the platform rejects one in principle.
Prototype regional federation
When does a public signal obligate plan authors to act?
Listening becomes theater unless the product defines when patterns require acknowledgement, investigation, revision, a new test, or an explained refusal.
Draft response thresholds
How can public input be representative without becoming surveillance?
Understanding who is missing may require contextual data, but collecting identity and demographics can exclude people, expose them, or enable misuse.
Design a privacy-preserving sample
Which implementation data stays local, and what may flow back to the worldwide network?
The system needs to learn across places without centralizing sensitive budgets, payments, personnel, identity, or operational records.
Federation model required
This document is alive
This is a working idea.
It should keep changing.
This is not a finished doctrine or a polished pitch pretending to have every answer. It is a starting point. We are being clear about what we believe so far, and we are open to having it rewritten by people and places we have not met yet.
CHANGE LOGLatest first
v0.5
From expertise to legitimate action
Separated expert, lived, local, and operational legitimacy; added panel formation guardrails, layered ratification, local authorization, and a federated execution architecture.
August 4, 2026v0.4
From plan to practice
Added the Implementation Commons, the role of generative UI, the real-world learning loop, and the roadmap commitment for a separate execution-platform design document and PRD set.
August 4, 2026v0.3
The participation system
Defined separate expert-review and public-listening lanes, the governed bridge between them, and the first app product requirements.
August 4, 2026v0.2
The world layer
Expanded the ethos to worldwide, polycentric collaboration and made cross-system intersections foundational.
August 4, 2026v0.1
Concept architecture published
Premise, participation model, plan anatomy, decision framework, safeguards, and pilot path.
August 4, 2026v0.6
Global participation charter
Language, regional federation, representation, knowledge equity, funding boundaries, and shared licensing.
Planned