A starting proposal for who may take part, how choices are counted, and how people can challenge the result.
A proposal to test, not rules already in force. No constitution has been adopted. There are no verified constituents, binding elections, or common treasury operating through this site. AI would help people do the work; people would hold the authority.
One eligible person. Equal standing.
Separate support from civic standing
Proposal v0.1
Give each future verified constituent one equal base vote. A separate process must establish eligibility and check that a person is counted only once.
Purpose
Stop wealth, duplicate accounts, or access to many email addresses from multiplying someone’s voice.
Tradeoff
Identity checks can collect too much information or shut out people without common documents. Global rules may conflict with local laws.
Open question
Which eligibility rules and privacy-preserving checks can include people fairly, with safe appeals for mistaken rejection?
A supporter confirms an email address and chooses to support the idea. That confirms control of an inbox, not a unique human. It gives no voting rights. A newsletter reader may not support the idea at all. These categories must stay separate.
A future verification service should keep only the information it needs, avoid public identity lists, and let an independent process hear challenges. No identity documents are collected by this site.
Represent more than one winning side
Multi-member proportional ranked-choice elections
Proposal v0.1
As a starting point, elect several representatives in each constituency. Voters rank candidates. Use a published single transferable vote rule, or STV, with its exact tally and audit method settled before voting.
Purpose
Give different groups a chance at seats and reduce votes being wasted on a single winner.
Tradeoff
Transfers are harder to explain and audit than a one-choice tally. Results depend on seat numbers, constituency boundaries, and the exact counting rule.
Open question
Which seat size, transfer rule, and boundaries give fair representation without making the count hard to understand?
A constituency is a group of voters that elects a shared set of representatives. In plain terms: you mark your first choice, then your next choices. Candidates who reach the published threshold win a seat. Parts of votes beyond that threshold can move to the next available choice; votes for a candidate who is eliminated can move too. Counting continues until the seats are filled. Exact transfer weights and tie rules must be public in advance.
This aims for broadly proportional seats; it does not promise perfect fairness. Terms would be limited. Elected officials would disclose conflicts, cast public roll-call votes, and face defined recall or removal processes. Election staff would remain separate from the Assembly.
Different rules for different decisions
Proposal v0.1
Use majority voting for routine yes-or-no decisions. Use a higher approval threshold, a waiting period, and independent review for constitutional decisions.
Purpose
Let ordinary work move forward while making changes to basic rules harder to rush.
Tradeoff
Higher thresholds can let a minority block a needed change. Simple majority choices can leave important alternatives off the ballot.
Open question
What thresholds, turnout rules, abstention rules, and rights that cannot be changed are appropriate for each decision?
Protected rights cannot be swept aside by an ordinary majority. The decision-rights table sets out the proposed bodies and checks. Thresholds remain open, not adopted rules.
Private choices. Independent checks.
Keep eligibility apart from ballot choices
Proposal v0.1
Use secret ballots for constituent decisions where pressure or retaliation is a risk. Check eligibility separately from selections. Let a voter check inclusion without a public record that reveals their choices.
Purpose
Protect a person’s choice while making missing or changed ballots easier to detect.
Tradeoff
Secrecy, checking the count, access, and protection from pressure can conflict. A compromised device or a person watching at home can defeat privacy.
Open question
How can inclusion checks work without creating a receipt that lets someone prove a vote to a buyer or coercer?
An inclusion check should show that a ballot entered the count. It must not provide a transferable proof of how someone voted. That distinction is a design goal, not a claim that we have built or validated a solution.
Official votes by representatives would be public so people can hold officeholders accountable. Private constituent ballots and public officeholder votes serve different needs.
Independent administration and human certification
Proposal v0.1
Have a separate election body publish the rules, administer ballots, support access, commission audits, hear challenges, and submit the result for final independent human certification.
Purpose
Keep a candidate, donor, administrator, or AI system from declaring its own victory.
Tradeoff
Independent review takes time and money. Reviewers need their own conflict checks and can still make mistakes.
Open question
Who appoints and removes election staff and certifiers, what evidence must audits inspect, and when is a rerun required?
Before a vote, publish eligibility rules, collection methods, tally rules, deadlines, complaint paths, and conditions for a recount or rerun. After it, publish the audit findings and reasons for certification while protecting ballot secrecy.
Council of Europe e-voting standards are a useful benchmark for review. Meeting a checklist is not independent certification of this proposal or a promise of security.
Start small and match checks to the stakes
Proposal v0.1
Test discussion and nonbinding priorities before any binding election. Require independent validation of security, legality, access, privacy, and resistance to coercion before high-stakes use. Keep safer alternatives open.
Purpose
Avoid handing real power to an untested voting system.
Tradeoff
A slower rollout postpones participation and may need in-person or other channels that cost more.
Open question
What failures stop a rollout, who makes that call, and what safe options work across countries?
Low-stakes trials cannot prove that a system is safe for high-stakes elections. AI may help inspect records, but it cannot issue the final certification, vote for a person, or decide an election dispute on its own.
Show priorities without buying power
Equal priority credits
Proposal v0.1
Give each eligible constituent the same number of credits in a defined cycle. People can divide them across agenda items or budget priorities. Credits cannot be bought, sold, gifted, or transferred.
Purpose
Show the strength of people’s priorities without letting money buy more influence.
Tradeoff
People can coordinate or pressure others. Credit totals can hide why people disagree and are not a full budget decision.
Open question
How many credits should a cycle use, what limits prevent gaming, and what access support makes participation fair?
These are points for expressing priorities, not cash, a token, or a claim on treasury funds. They advise the published decision process; they do not override rights, reviews, or the Assembly’s accountable budget vote.
Optional, revocable topic delegation
Proposal v0.1
Let a constituent delegate a limited choice to another eligible person for a named topic and set period. Make withdrawal easy. A direct ballot replaces the delegated choice so the person is never counted twice.
Purpose
Let people get help on issues they cannot follow closely, while keeping control of their voice.
Tradeoff
Delegates can gain too much influence. Hidden pressure, confusing chains, or late revocation can distort a result.
Open question
Should delegation chains be prohibited, what caps are needed, and how can revocation be clear before a tally closes?
Delegation would not transfer ownership of a vote. No permanent proxies or purchased delegation. Scope, expiry, conflicts, and the point at which a ballot becomes final must be explained before participation.
Give everyday people time to think
Sortition: choosing a panel by lot
Proposal v0.1
Choose temporary panels randomly from verified constituents, then use a published selection method to include different backgrounds or jurisdictions. Provide accessible evidence, expert questioning, opposing briefs, and majority and minority reports.
Purpose
Bring in people outside the usual political, donor, and expert circles without asking everyone to study every complex issue.
Tradeoff
A small panel cannot mirror every person. Group categories, refusals, unequal free time, and who writes the briefing can bias the result.
Open question
Which categories are fair to use, how should refusals be handled, and what pay or support lets more people serve?
Sortition means selection by lot. Stratification means balancing the draw across stated groups, rather than relying on a simple draw that might miss some of them. Neither guarantees that every view is represented.
OECD good-practice guidance informs the selection and process design. Panels advise and scrutinize; they do not replace the full constituent body or gain a general veto over it.