Common Good AI

Building common ground for the common good.

Dear Subscriber First Name,

Welcome back to our monthly newsletter. We continue to receive messages from readers, who are enthusiastic about community-based problem solving using digital collaboration at scale. We’ve connected with a wide range of individuals, from strategists designing inclusive global governance systems to grassroots organizers hosting citizen dialogues to mitigate US polarization. It validates the need for new tools to promote inclusive civic engagement to find an actionable path forward together.

Based on several readers’ requests, we compiled three cases that highlight the application of the CrowdSmart tool. We share our initial learning and insights from these cases. We complement our learning with reflections on other global case studies.  

Keep sending us your thoughts, ideas, and suggestions, which you can email to info@commongoodai.org.

Sincerely,

Victoria Stanski, Managing Director

Founders, John Seely Brown, John Clippinger, John Cordier, Tom Kehler, Kim Polese

Application Case: Bridging citizens and leaders using digital tools

Village Media is a leader in community-based journalism who reaches five million subscribers in over 150 small communities across North America. Earlier this year, Village Media used the CrowdSmart tool for two engagements:

1. Municipal budget planning in Sault Ste Marie Municipality, Michigan

2. Identifying health care barriers in Ontario, Canada

We document the two engagements and how citizens generated new insights collected through online collaborations that were shared with local policy makers.

Deeper Case Study: Identifying root causes of Cincinnati gun violence

Before starting Common Good AI, the CrowdSmart technology was used by a Cincinnati community to explore gun violence. We interviewed the doctor who initiated the conversation between parents, parole officers, former incarcerated persons, and gun violence victims. You can learn more by reading one of our first case studies, which inspired the launch of Common Good AI.

Moving forward: What we're thinking

As we identify and design more digital engagements, Common Good AI is learning from other case studies and digital tools. We are thinking about four critical issues to inform future collaborations at scale: 

1. Success: What does “success” look like for an AI-supported democratic deliberation?

Massachusetts Institute of Technology (MIT) with Project Liberty conducted a comprehensive study on Generative AI for ProDemocracy Platforms and reflected on 27 different deliberative tools. It explores multiple aspects of a deliberative process to enable success, including preserving user agency and autonomy, encouraging mutual respect, promoting equality and inclusiveness, and augmenting human capacities—not substituting them. 

2. Incentives: How do we best incentivize people to participate in online dialogues, especially within the current ‘attention economy’?

For some, the opportunity to influence public policy or to “have their say” is enough to motivate participation, but this could result in a biased sample, made up of only those who are already engaged. We applaud Brazil’s effort to reach 1.4 million citizens through their online deliberations to set priorities on national policies to combat hunger, deforestation, and climate change. Numerous strategies were used to motivate constituents, tailored to demographic and topic issue. 

3. Languages: How do we ensure an online engagement is accessible to all people?

We want Common Good AI to operate in a range of languages to ensure we’re reaching across diverse communities. It requires investing time to build translation capacity to support genuine inclusive approaches. We recently learned about the Alliance for Middle East Peace and Remesh-supported peacebuilding in Israel-Palestine.

4. Human touch: When we design a deliberation process, what actions should be human vs. machine generated? 

We believe that a successful engagement is one that is human designed and AI supported, both to maximize impact and reduce potential bias. In Germany, 400,000 citizens identified misinformation issues using the online tool Pol.is and then deliberated in a 120 person Citizen Assembly to generate recommendations to tackle disinformation that were shared with the government.