Skip to main content
THE THESIS

Modern democracy is not failing
because people disagree.

It's failing because the infrastructure for agreementdoesn't exist.

THE PROBLEM

Every existing tool is architecturally broken

⚡

Engagement through conflict

Social media

Algorithms surface the most divisive opinion because outrage drives clicks. Latent agreement — often the majority of what a community actually holds in common — stays invisible.

▣

Low-fidelity signal capture

Surveys, polls, votes

A thumbs-up, a 1-5 rating, a yes/no vote. These flatten the richness of human values into noise. A "No" could mean a dozen different things.

⚐

Access-biased testimony

Town halls, public comment

Biased toward people with time, confidence, and access. The loudest voice wins, not the most representative one.

“Most disagreements are actually the same underlying needexpressed through different cultural or political vocabularies.”

“Defund the police” and “Back the blue” both contain people who want safe neighborhoods with accountable public servants. The Civic Filament finds that signal.

WHY NOW

Five forces converging

LLMs are finally capable enough

Conducting nuanced, psychologically-grounded interviews at scale was impossible 3 years ago. Now it's an API call.

Institutional trust at historic lows

Broad demand for tools that make people feel genuinely heard. Trust in institutions, in most polling, sits near historic lows — the reservoir is unmet demand.

DAOs proved the appetite

Decentralized governance proved people want to participate — and that "voting on everything" doesn't scale. They need synthesis.

The civic loneliness epidemic

People feel disconnected from their communities. Meaningful participation in shared decisions is a partial antidote.

AI regulation is coming

Governments need to demonstrate AI is used for democratic processes, not against them. Civic Filament is the poster child.

LANDSCAPE

What exists today — and what's missing

Pol.isReal-time survey + clustering◇ No AI mediator. Low-fidelity input. Clusters positions, not values.
Consider.itPro/con deliberation◇ Still position-based. No synthesis into actionable documents.
LoomioCooperative decision-making◇ Process tool, not intelligence tool. No AI synthesis.
SurveyMonkeyTraditional surveys◇ Fundamentally low-fidelity. Questions are static, not adaptive.
Town HallsIn-person testimony◇ Biased toward access. No synthesis. No scale.
THE MOAT

Five compounding advantages

1

High-fidelity input

AI interviews, not surveys

2

Value-level analysis

Needs, not positions

3

Actionable output

LRDs, not reports

4

Growing perspective graph

Network effects compound

5

Privacy by architecture

Trust as competitive advantage

THE BUILDERS

Built by Moonlit Social Labs

Moonlit Social Labs

FOUNDER & PRINCIPAL ARCHITECT

Moonlit Social Labs builds tools at the intersection of AI, civic technology, and collective intelligence. The Consensus Infrastructure was born from the belief that the infrastructure for human agreement is the most important unsolved problem in governance — and that AI can finally help solve it.

SUPPORT ON KO-FI
THE VISION

In five years, the question“What does the community actually want?”has an answer.

Not a poll. Not a protest. Not a comment section full of the loudest voices. An answer — nuanced, representative, actionable, and continuously updated.

We are not building a social network. We are building civic infrastructure — as fundamental to 21st-century governance as roads are to transportation.