Most civic platforms produce conversation. This one produces documented collective decisions backed by political force. Here's exactly how it works.
See where you stand โThe failure modes are structural โ not failures of effort or intent.
Select leaders every few years. Cannot specify policy between cycles.
Fails at: policy specificationStart aligned with members. Over time, organizational survival replaces member preferences.
Fails at: resisting captureAccurately captures opinion. Has no mechanism to convert that opinion into action.
Fails at: execution layerOptimizes for engagement. Outrage travels faster than reasoning โ by design.
Fails at: deliberation qualityAggregate signatures at near-zero cost. Politicians know cheap signals don't represent voting behavior or sustained pressure.
Fails at: Execution Layer"No persistent mechanism exists for converting majority citizen preferences on cross-cutting policy issues into concentrated political force capable of overcoming organized economic opposition โ when those interests have shaped both the legislative and information environment."
Every thread moves through the same sequence. Phases are irreversible. You cannot vote before deliberating. You cannot propose before discussing.
Verified participants can read, post, and cast signals. The thread is open for exploratory discussion. No formal structure yet โ the goal is surfacing perspectives.
The facilitator has formally opened deliberation. Discussion continues, but participants are expected to engage with the issue substantively. Signals update in real time.
Posting is locked. Participants submit formal proposals with a title and description. All participants can see and discuss proposals. The facilitator decides when the proposal set is complete.
Each participant votes on each proposal. Votes are immutable โ once cast, they cannot be changed. This prevents strategic vote-switching. Results are not shown until the phase closes.
Vote results are published. The winning proposal and full vote summary are written to the public audit log. The admin records an allocation decision. Everything is now read-only.
The full deliberation โ every post, signal, proposal, vote, and allocation decision โ is preserved in the public audit log indefinitely. Any observer can reconstruct how the decision was reached.
Signals replace reaction buttons. One signal per participant per thread. They express your orientation toward the discussion, not your opinion of a post.
I support the direction of this discussion. I believe it's moving toward a good outcome.
I have concerns that haven't been adequately addressed. I'm not opposed, but not fully behind it yet.
I can't form a view yet. Key questions remain unanswered. More evidence or clarification is needed.
I have a serious objection. Proceeding in the current direction would cause meaningful harm.
The audit log is not an afterthought โ it is the platform's primary legitimacy mechanism. Every phase transition, post, signal, vote, and allocation decision is recorded and publicly accessible.
Actor identities are UUIDs โ publicly auditable without exposing personal information.
None of the platform's design decisions are arbitrary. Each one addresses a specific failure mode of existing civic tools.
Posts appear in the order they were written. The first post and the last post have equal visibility.
No upvotes, downvotes, or reactions on individual posts. Signals express your orientation toward the thread as a whole.
Voting is only available after discussion and proposal phases are complete. The sequence is enforced at the database level โ not just the UI.
The audit log is append-only and publicly accessible. Journalists, researchers, and skeptics can verify every decision without permission.
Once cast, votes cannot be changed. This prevents strategic vote-switching and late-stage manipulation.
Every facilitator action โ phase advance, post removal โ is recorded in the public audit log with a required stated reason.
The five most likely failure modes, stated plainly.
Most people who sign up for civic platforms don't complete a full deliberation. If threads don't reach quorum โ enough verified participants to make the outcome legitimate โ they produce noise, not signal. We need sustained participation, not just sign-ups.
Platform operations and lobbyist fees require predictable recurring revenue. If the platform attracts users but not donors, or donors give once and don't renew, the execution layer collapses. A deliberation without a funded lobbyist is just a forum.
The causal chain is long: consensus โ funding pool โ lobbyist engagement โ legislative access โ policy change. Each link can break. A lobbyist can be outspent. A legislator can be unresponsive. A policy can pass and be weakly enforced. We won't know which links are weakest until we try.
Any institution can be captured by the people who run it, fund it, or participate most actively. We've designed structural safeguards โ facilitator accountability, public audit logs, term limits โ but these are imperfect. If this platform succeeds and becomes influential, the pressure to capture it will increase.
Deliberation takes weeks. Legislative windows open and close in days. If our process produces consensus too slowly to influence a specific vote, the effort is wasted for that cycle. Getting the timing right requires institutional relationships we don't yet have.
We publish these failure modes because we believe transparency about risk is a precondition for trust. If this platform is going to ask for your time, your identity verification, and eventually your money, you deserve to know exactly how it could fail and what we're doing about each one.
Donations go into a domain funding pool โ not to specific proposals. The deliberation process determines how that pool gets allocated.
This is why the deliberation process matters: the community decides how the money is used. That accountability is what separates this from a donation to an advocacy organization.
Take a 3-minute quiz to find out how your views connect to ongoing deliberations.
Take the Quiz Go to Platform