Polis is described on the website as a real-time system for gathering, analysing, and understanding what large groups of people think in their own words, enabled by advanced statistics and machine learning. It allows anyone to create a conversation in the form of a Wikisurvey and invite participants to submit comments expressing their views on the topic. They can also vote on other people’s opinions by clicking agree, disagree, or pass. As more statements and votes accumulate, the algorithm sorts participants into clusters of like-minded people and identifies what opinions distinguish them and if there is a rough consensus among groups. Participants are shown the visualisation of the entire landscape of opinions as it evolves and can create new comments to the conversation.
Advantages
- Proven Success in Public Deliberation: Polis has been successfully utilised by activists, politicians, and researchers to crowdsource ideas and facilitate public deliberation. A notable example is its use by the Taiwanese government to crowdsource a set of regulations for Uber, demonstrating its effectiveness in consensus building and policy development.
- Reduction of Trolling: By removing the ability to reply directly to others' comments, Polis minimises the chances of trolling and unproductive arguments, fostering a more respectful and focused discussion environment.
- Representation of Minority Opinions: The clustering algorithm used in Polis ensures that even minority opinions are represented within the opinion landscape.
- Engaging: Unlike traditional government questionnaires, which can be tedious and passive, a wiki-survey offers a more engaging experience. Participants actively contribute by shaping the dimensions of the discussions, making the process more dynamic and enjoyable. By allowing participants to influence the conversation's direction, a wiki-survey fosters a sense of ownership and investment in the outcome.
- Open Source and Aligned with Common Good: The solution is open source and developed by The Computational Democracy Project, a non-profit organisation dedicated to creating tools for the common good. This alignment with values of transparency, inclusivity, and public benefit further enhances its credibility and adoption in civic participation efforts.
Limitations
- Sybil Attacks: The current solution is vulnerable to Sybil attacks, where individuals could use multiple devices or VPNs to create multiple identities, allowing them to manipulate discussions or voting outcomes. This could lead to a distorted representation of opinions and outcomes. It also prevents Polis from scaling outside of trusted communities that receive specific Polis conversations links.
- Algorithmic Bias: The semi-random algorithm that distributes comments prioritises those likely to predict participants' positions in the opinion landscape, while demoting others that might not be passed on. This approach could inadvertently filter out expert opinions that use technical language, thereby skewing the conversation and limiting the diversity of viewpoints.
- Centralised Moderation: The current design places all moderation power in the hands of the conversation initiator. This centralization of authority allows the moderator to potentially censor opinions or manipulate the conversation to align with their own views. This could undermine the platform's goal of fostering open and unbiased civic participation.
- Technical Limitations: The software is divided into subsystems that are highly-coupled with the database component. The source code was designed to be used as stand-alone software, with only the possibility to embed links to the pol.is website into third-party websites.
Could it be useful for our requirements?
- Polis is a powerful building-block to provide bridging-based ranking algorithms. Tightly integrating the core Polis algorithms directly with our product without linking to the Polis website will however require lots of innovative work, because Polis wasn’t designed with this use-case in mind. In particular, we’d need to extract the core Polis algorithm, and plug into it our sybil-resistant authentication algorithm as well as a more transparent moderation system (Verifiable Moderation), then store the corresponding data in our own unified database.
