Rethinking public engagement in state government with generative AI

In a recent InnovateUS workshop, Eric Gordon, Professor of Civic Media at Emerson College and director of the Engagement Lab, discussed how to rethink public engagement in state government using generative AI.

Gordon focuses on the role of storytelling and emerging technology (such as AI)  in trust and governance. He specializes in collaborative research and design processes, and has served as an expert advisor for local and national governments, as well as NGOs around the world, designing responsive processes that help organizations transform to meet their stated values. He is the author of over 50 articles and chapters on media and urbanism, and the author of two books on the subject.

Gordon began the workshop with a story about a current project he is working on. This project aimed to answer the question: “What if honest stories from students in any language could be collected with a trusted platform, effectively summarized in real-time, and then shared with decisionmakers in such a way that makes participants feel heard and that holds institutions accountable.” Following an exercise that involved students speaking with AI chatbots, students expressed feeling validated and heard in a way they hadn’t been before. 

Gordon revealed that shifting this model to a government context can present difficulties. 

“When government is in crisis, we tend to talk about listening better. But the problem is, we mostly just invite more speech,” he said.

He stressed that listening matters now more than ever, as while speech is relatively easy and accessible to anyone, it is making sense of what is said and creating actionable change that is difficult. 

The process starts with trust. Trust involves a multi-faceted approach where different alternatives are considered, the prioritization and selection of risks, and non-cognitive, non-verifiable factors. Unfortunately, Gordon said, many democracies are experiencing a trust deficit, caused by values and expectation misalignment. 

To address this trust deficit, democratic governments need to start listening and incorporating data of all kinds, not just the most recent information.

“Listening is the top trust-building action that government and other institutions can take,” he said. “ … We’re stuck in our analog ways. We’re stuck with a perception that data has a short shelf life and that we need fresh data to protect against community blowback.”

Gordon stressed that trust with communities is lost when listening isn’t taken seriously. Effort is put into soliciting more speech, but the processing of that speech is ignored. 

Governments need to add a data layer to their process. This allows for reimagination of current institutions, and includes the collection, analysis, and use of data. This focus requires reworking the traditional way governments engage with the public. Gordon suggests governments use generative listening, which is an institutional commitment to collaborate with constituents on shaping and making decisions through the critical collection, analysis, and use of data.

“I think engagement has run its course. We need to listen,” Gordon said. 

Gordon introduced five qualities of governments to show how generative listening is different from public engagement.  These include:

• Intake

• Speech

• Transparency

• Feedback

• Decisionmaking

Intake

There are both close-ended and open-ended types of listening. While public engagement encourages more forms of close-ended listening, generative listening is more open-ended, allowing for researchers to understand new perspectives and inform future questions. 

Speech

Gordon emphasized that speech is overabundant and overprioritized. Public engagement emphasizes motivating speech from as many stakeholders as possible. However, generative listening emphasizes both dialogue between stakeholders and the interaction between publics and organizations.

“The difference is subtle, but it’s important,” he said. “With dialogue it’s about speech that builds.”

Transparency

With public engagement, attention is focused on the transparency of specific transactions, including rules of engagement in town meetings, or participation in surveys. Generative listening is more systemic, and focuses on transparency in how the organization collects, analyzes, and uses data, and how values are used to justify organizational decision-making. 

Gordon highlights data walks to visualize this concept. Data walks are a means of creating transparency. For example, in a Chicago suburb the city is using the city health dashboard to access community health and equity data. Instead of just using novel data sources in the dashboard and visualizations as a decision support tool, the city partnered with NGOs and activists in a neighborhood where life expectancy was slightly lower than the rest of the city’s wealthiest neighborhoods. Together they have arranged regular data walks where community members walk the neighborhood to check and verify data sources including asthma rates, eviction rates, access to healthy foods, and more. The walks are led by community members with the representatives from the city present and listening. The result of these walks has been positive for both the community and the government’s trust in the data they have, as they provide a human check on the data source.

Feedback

For the public engagement paradigm, feedback on specific contributions is given after a period of analysis. Generative listening is beneficial in that it provides immediate feedback and acknowledges the listening relationship. 

Decision-making

 
Generative listening differs from public engagement in that data storage and decision-making is distributed across multiple agencies or departments. This is significant, as institutions are held accountable by publicly accessible data traces and conversations that transcend specific transactions.

Gordon highlighted the Data Collaborative for Justice, an organization dedicated to criminal justice reform as an example of how to properly engage in deliberative and consenus-oriented decision-making. 

To conclude, Gordon emphasized that generative listening is a crucial shift in our thinking about democratic governance.

“Generative AI is not the solution, but it’s a call to action. We need to stop engaging and start listening,” he said.

To watch the full recording of the hour-long workshop, click here. Visit innovate-us.org/workshops to sign up for future sessions.