Consensus Engine: Collective Intelligence for Participatory Budgeting
AI that amplifies collective wisdom instead of individual productivity
The Opportunity
Exploits: Individual Productivity Obsession
Their Blind Spot: “AI should make individuals more productive”
Our Opportunity: AI that enhances group decision-making, not individual output
Status: ✅ Stage 5 - Design Complete
Problem Space
What Capitalism Built
Current “participatory” systems are just voting apps with extra steps:
- Winner/loser dynamics that divide communities
- Vote counting that reduces complex views to numbers
- Gamification that turns civic engagement into competition
- Individual choice prioritized over collective wisdom
- Efficiency metrics that rush decisions
What Communities Need
True participatory budgeting requires:
- Deep understanding across different perspectives
- Creative synthesis of seemingly opposing views
- Minority protection from majority dominance
- Collective wisdom emerging from deliberation
- Authentic consensus not just vote tallies
The Exploitation
Capitalism literally cannot comprehend success metrics based on:
- Collective understanding over individual speed
- Process quality over decision efficiency
- Emergent wisdom over predetermined outcomes
- Synthesis over competition
- Group intelligence over individual productivity
Our Solution
ConsensusEngine: A Collective Intelligence Amplifier
Instead of counting votes, ConsensusEngine amplifies the collective intelligence of communities making decisions together. It treats consensus-building as an emergent process where the whole becomes greater than the sum of individual opinions.
Core Philosophy
- Emergence over reduction: Complex collective wisdom can’t be reduced to vote counts
- Process over product: How we decide matters as much as what we decide
- Synthesis over compromise: Seek creative integration, not middle ground
- Understanding over agreement: Measure comprehension, not just consensus
Key Features
Not vote counting, but perspective weaving
Instead of reducing input to votes, the system:
- Preserves the full richness of diverse perspectives
- Amplifies quiet and marginalized voices
- Identifies connections between seemingly different ideas
- Resists reduction to individual preferences
Example: When discussing playground funding, parents might emphasize safety, kids want fun equipment, and elders seek shaded seating. Rather than voting between options, the system helps discover a design incorporating all needs.
2. Emergent Theme Discovery
Finding patterns humans might miss
The AI identifies:
- Unexpected connections between different proposals
- Shared values underlying surface disagreements
- Creative possibilities from combining ideas
- Patterns across multiple decision cycles
Example: Noticing that proposals for youth programs, elder care, and job training all share themes of intergenerational knowledge transfer, suggesting a unified community mentorship initiative.
Turning tension into creative energy
Rather than avoiding or suppressing conflict:
- Maps the creative tension between different visions
- Identifies synthesis opportunities
- Generates bridging proposals
- Facilitates constructive dialogue
Example: When transit advocates clash with parking supporters, the system might surface shared concerns about accessibility and suggest demand-responsive transit that serves both needs.
4. Consensus Quality Analysis
Measuring depth, not just agreement
Goes beyond “yes/no” to assess:
- How well people understand each other’s perspectives
- Whether agreement is genuine or coerced
- If minority views are truly integrated
- The stability and resilience of decisions
Example: A 60% vote might hide deep division, while a thoughtful 80% consensus with understood dissent might be healthier.
5. Collective Memory System
Learning from every decision
The system maintains:
- Why decisions were made, not just what
- Patterns of successful consensus building
- Community-specific facilitation wisdom
- Lessons from past processes
Example: Remembering that evening meetings exclude parents, or that certain facilitation styles work better for this community.
Technical Approach
Architecture Overview
graph TD
A[Community Input] --> B[Synthesis Engine]
B --> C[Theme Discovery]
C --> D[Conflict Transformation]
D --> E[Consensus Building]
E --> F[Decision Crystallization]
F --> G[Collective Memory]
G --> B
H[Community Governance] -.-> B
H -.-> D
H -.-> E
Key Technologies
- Multi-modal input: Text, voice, video, drawings - meeting people where they are
- Semantic analysis: Understanding meaning, not just counting keywords
- Network analysis: Mapping relationships between ideas and perspectives
- Federated architecture: Each community runs their own instance
- Privacy-first design: Strong anonymization and consent controls
Communities govern their own consensus process:
- Define what consensus means for them
- Choose facilitation styles and intensity
- Set minority protection thresholds
- Control pacing and phases
- Access and interpret their collective memory
Use Cases
Municipal Participatory Budgeting
Transform city budget allocation from voting on pre-set options to genuine collective visioning:
- Surface community priorities organically
- Synthesize neighborhood needs with city-wide goals
- Build understanding across different districts
- Create innovative funding proposals
Worker Cooperative Decisions
Support democratic workplaces with complex decisions:
- Strategic planning with all voices heard
- Resource allocation balancing multiple needs
- Policy development through collective wisdom
- Conflict resolution preserving relationships
Enable grassroots groups to make inclusive decisions:
- Campaign strategy incorporating diverse tactics
- Resource sharing agreements
- Partnership negotiations
- Vision and values alignment
Neighborhood Associations
Facilitate local decision-making:
- Development proposals affecting everyone
- Community garden management
- Event planning and resource allocation
- Dispute resolution between neighbors
Implementation Roadmap
Phase 1: Core Engine (4 months)
What we’ll build first:
- Basic input gathering and synthesis
- Simple theme emergence detection
- Consensus quality metrics
- Facilitation prompts
Community involvement:
- Alpha testing with 3 partner communities
- Weekly feedback sessions
- Co-design of core features
Phase 2: Collective Intelligence (4 months)
Enhanced capabilities:
- Advanced emergence detection
- Conflict transformation tools
- Real-time visualizations
- Collective memory system
Scaling up:
- Beta testing with 10 communities
- Cross-community learning
- Facilitator training programs
Phase 3: Federation Network (6 months)
Building the network:
- Full federation protocol
- Wisdom sharing between communities
- Pattern libraries
- Distributed governance
Going wide:
- 20+ federated communities
- Municipal partnerships
- Open source release
Get Involved
For Communities
Pilot Program
- Join our alpha testing cohort
- Shape the development with your needs
- Get support implementing participatory processes
- Build facilitation capacity
Requirements:
- Active participatory decision-making
- Commitment to feedback and co-design
- Willingness to experiment
- Basic technical infrastructure
For Developers
Technical Contributions
- Distributed systems architecture
- Machine learning for emergence detection
- Accessibility and internationalization
- Privacy and security features
Open Source Development
- GitHub: [Coming soon]
- Tech stack: Rust, Python, JAX
- License: AGPLv3
For Funders
Support Community Democracy
- Fund development costs
- Sponsor community pilots
- Support facilitator training
- Enable municipal partnerships
Investment Needs:
- Development team: $200K/year
- Community coordination: $100K/year
- Infrastructure: $50K/year
For Researchers
Collaborate on:
- Collective intelligence metrics
- Emergence detection algorithms
- Consensus quality assessment
- Cross-cultural facilitation patterns
Why This Matters
Beyond Voting
Voting systems, even ranked choice, reduce complex community wisdom to numbers. ConsensusEngine preserves and amplifies the full richness of collective deliberation.
Challenging Core Assumptions
By measuring collective understanding rather than individual productivity, we challenge capitalism’s fundamental framework for evaluating AI success.
Building Commons
Every decision process adds to a commons of facilitation wisdom that communities share and build upon, rather than proprietary algorithms owned by corporations.
Real Democracy
Democracy is more than voting—it’s the capacity for communities to think together, understand each other, and make decisions that reflect collective wisdom.
Current Status
Completed
- ✅ Full technical specification (460 lines)
- ✅ Architecture design
- ✅ Data models and flows
- ✅ Community governance framework
- ✅ Implementation phases planned
In Progress
- 🔄 Community partnership development
- 🔄 Funding acquisition
- 🔄 Technical team assembly
Next Steps
The Vision
Imagine city budget meetings where:
- Every voice shapes the outcome, not just the loudest
- Creative solutions emerge that no one imagined alone
- Understanding grows even through disagreement
- Decisions reflect genuine collective wisdom
- Communities own and control their democratic tools
This isn’t just better voting software. It’s infrastructure for collective intelligence—tools that help communities think together in ways capitalism’s individual-focused systems never could.
“The master’s tools will never dismantle the master’s house—unless we reprogram them for collective liberation.”
View full technical specification