The PSi project
Scaling Real-Time Deliberation to Massive Audiences
The PSi platform (People Supported intelligence) is an innovative system designed to facilitate large-scale, real-time deliberative decision-making. Unlike traditional methods that rely on surveys or hierarchical structures, PSi enables synchronous, voice-based discussions among thousands, millions, or even billions of participants. By leveraging iterative small-group discussions and advanced analytics, PSi produces ranked solutions to open-ended questions efficiently, converging to collective decisions in logarithmic time relative to group size.
The Problem of Deliberation at Scale
Scaling deliberative democracy presents a significant challenge as traditional deliberative processes struggle to accommodate large groups without sacrificing inclusivity, depth, or efficiency. Current approaches, such as citizen assemblies or deliberative polls, often rely on small, representative samples, leaving most voices unheard. As group size increases, the complexity of facilitating meaningful dialogue and aggregating diverse perspectives grows exponentially. Large-scale discussions are prone to issues like information overload, social biases, and dominance by influential individuals, which can hinder fair and informed decision-making. Furthermore, traditional methods often lack the infrastructure to manage real-time interactions at scale, making it difficult to engage millions of participants in nuanced debates. Addressing these limitations is critical for making democratic processes more inclusive, participatory, and resilient in the face of complex societal challenges.
Current Solutions
Systems like Polis and Google’s "Habermas machine" offer innovative solutions to the problem of aggregating individual preferences by identifying consensus statements that reflect shared views across diverse groups. These platforms leverage algorithms to map opinion spaces and highlight areas of agreement, making them effective tools for summarizing collective preferences. However, they do not address the fundamental deliberation problem: participants in these systems do not directly engage with one another, limiting the potential for social interaction to influence and refine individual perspectives. Deliberation relies on the dynamic exchange of ideas, where participants can challenge assumptions, negotiate differences, and collaboratively deepen their understanding of complex issues. Without this interactive element, systems like Polis and Habermas machine may capture static consensus but fail to foster the transformative benefits of dialogue that are central to meaningful democratic deliberation.
The PSi System
Scalability: PSi supports deliberations among massive groups while maintaining high informational depth, achieving consensus in minutes even for global-scale discussions.
Resilience: The platform mitigates social biases such as herding and groupthink by structuring conversations in independent, small groups.
AI-Driven Insights: PSi integrates AI to analyze conversation data, highlighting trends, summarizing key points, and supporting behavioral analysis.
Practical Applications: Used by organizations like the Mayor of London’s Office, PSi has proven effective in contexts ranging from community-led policing consultations to public deliberations on divisive topics.
By democratizing access to decision-making, PSi exemplifies the potential of collective intelligence to address complex societal challenges.