"Coherent Mode" for the World's Public Square (2211.12571v1)

Abstract: Systems for large scale deliberation have resolved polarized issues and shifted agenda setting into the public's hands. These systems integrate bridging-based ranking algorithms - including group informed consensus implemented in Polis and the continuous matrix factorization approach implemented by Twitter Birdwatch - making it possible to highlight statements which enjoy broad support from a diversity of opinion groups. Polis has been productively employed to foster more constructive political deliberation at nation scale in law making exercises. Twitter Birdwatch is implemented with the intention of addressing misinformation in the global public square. From one perspective, Twitter Birdwatch can be viewed as an anti-misinformation system which has deliberative aspects. But it can also be viewed as a first step towards a generalized deliberative system, using Twitter's misinformation problem as a proving ground. In this paper, we propose that Twitter could adapt Birdwatch to produce maps of public opinion. We describe a system in five parts for generalizing Birdwatch: activation of a deliberative system and topic selection, population sampling and the role of expert networks, deliberation, reporting interpretable results and finally distribution of the results to the public and those in power.

• The paper proposes a framework to generalize Twitter Birdwatch by using bridging-based ranking algorithms to foster consensus and constructive public dialogue.
• It outlines a five-part system covering topic activation, representative sampling, user interface adaptations, interpretable reporting, and results distribution.
• The study highlights potential integrations with LLMs and decentralized consensus methods to make public deliberations more scalable and impactful.

An Expert Overview of "Coherent Mode" for the World's Public Square

This paper, authored by Colin Megill, Elizabeth Barry, and Christopher Small of The Computational Democracy Project, presents a proposition for enhancing Twitter’s deliberative capacities by generalizing its Birdwatch system. The foundation of the paper is built upon the bridging-based ranking algorithms exemplified by the systems of Polis and Twitter’s Birdwatch, aiming to foster coherent and constructive public discourse on large-scale platforms like Twitter.

Bridging-Based Ranking Algorithms

The authors highlight the use and benefits of bridging-based ranking algorithms, such as those employed by Polis and Birdwatch. Polis, recognized for its effectiveness in nationwide political deliberations, is noted for its ability to galvanize constructive dialogue by utilizing Matrix Factorization (MF) to group and rank public comments based on consensus. Such capabilities enable Polis to surface opinions that enjoy majority support across different opinion groups. Similarly, Birdwatch engages users to address misinformation by crowdsourcing notes that are ranked using MF algorithms, thus leveraging the strengths of collective intelligence to combat false information on Twitter.

Generalizing Twitter Birdwatch

The core proposition of this work is to expand Birdwatch's capabilities from a misinformation-focused tool to a versatile deliberation platform. The authors delineate a five-part system designed to extend this utility:

1. Activation and Topic Selection: The paper suggests leveraging Twitter's dynamic interaction properties for topic emergence and system activation.
2. Population Sampling and Expert Networks: The proposition includes a multilayered, demographically representative sampling strategy, which includes an integration of expert networks, to ensure informed deliberation.
3. Deliberation through User Interfaces: Suggestions are made for adapting the current Birdwatch interface to accommodate general deliberation tasks, including allowing for replied tweets to act as native deliberative prompts.
4. Interpretable Results Reporting: The paper discusses the necessity for results to be reported in an interpretable format to both the public and decision-makers, facilitating broad understanding and alignment with deliberative insights.
5. Results Distribution: Highlighting the potential impact, results are designed to be publicly accessible and to promote accountability among public officials.

Practical and Theoretical Implications

The implications of this proposal are manifold, with potential impacts on both democratizing public deliberation and improving the synthesis between citizen perspectives and policymaking. The extension of Birdwatch into a broad deliberative mechanism could serve to highlight consensus that would be otherwise obscured by the polarized nature often experienced on social media platforms. The authors posit that the potential for widespread application of bridge-based ranking methods might improve public discourse and understanding if deployed on a global scale, as they offer a structure for coherent expression of public will.

Future Directions in AI and Deliberative Systems

Future work in this area includes further integration of LLMs for more nuanced topic analysis and readiness checks for large-scale implementation. The paper speculates on the application of machine learning augmented deliberation systems in distributed social media protocols, anticipating a shift towards decentralized implementations that leverage distributed consensus algorithms for public discourse.

Conclusion

This paper presents a compelling proposal for transforming Twitter into a more deliberatively coherent social media platform by generalizing Birdwatch. Through structured engagement and advanced ranking methodologies, the proposed system positions itself as a potential vehicle for enhancing the quality and impact of public debates and their reception by those in power. As such, this research contributes to the ongoing discourse on leveraging computational methods to fortify democratic processes and public engagement.