Public engagement requires more than just amplifying citizen voice. The digital democracy movement of the 1990s promised that if everyone could speak up, then meaningful democratic engagement would naturally follow. But as Beth Simone Noveck and Danielle Allen explored in their recent workshop launching the Reboot Democracy: Designing Democratic Engagement for the AI Era series, real public engagement requires infrastructure for both listening to, and acting on, public input.

This insight cuts to the heart of why 25 years of "e-democracy" experiments haven't transformed governance the way we hoped. As Noveck emphasized throughout the workshop discussion, we ignored the equally important challenge of listening. We built tools for talking, not tools for hearing what people actually had to say.

Why Democracy's Technology Stack is Broken

Political philosopher Danielle Allen offered a provocative reframe during the conversation: democracy itself is a technology. It was "discovered in antiquity, disappeared, then rediscovered in the 18th century when people glued the idea of democracy together with representation."

But here's the problem: 18th-century technology was designed for radically different conditions. It assumed slow information flows fragmented by geographic distance. It was built for a world where circulating knowledge beyond official institutions was genuinely difficult.

Those conditions have vanished. The result is what Allen calls a crisis in the “spinal cord” of democracy—with vertebrae that connect citizen’s voice to decision-making to implementation. Instead of a functioning backbone, we have three disconnected pieces that rarely talk to each other.

If we want to upgrade democracy for the 21st century, Allen argued, we need to strengthen all three vertebrae. Technology—and AI in particular—can help us build better infrastructure for each area. But the real opportunity lies in connecting these enhanced participation processes to decision-makers who can act on what they learn, and to implementation systems that can deliver results people can see.

There's another design flaw we inherited from the web era. Democracy is fundamentally a place-based technology—that's why we have districts, wards, counties, congressional boundaries. But in the 1990s, we saw digital tools as explicitly not place-based. We were building for a borderless world, which meant our democratic innovations often floated free from the geographic communities where governance actually happens.

The AI Opportunity: Beyond Better Megaphones

But what if we could build tools for listening, not just talking? Recent examples  demonstrate strong potential for improving public engagement. In Bowling Green, Kentucky, city leaders used Polis and Google's Jigsaw Sensemaker to develop the city’s 25-year plan. Instead of the usual town halls that attract the same dozen vocal residents, they engaged thousands of people, identified minority perspectives that would have been drowned out, and discovered points of consensus that weren't previously visible. The Engaged California platform is taking a different approach, beginning with the issue of LA wildfire recovery, where citizens are providing input on resource allocation and rebuilding plans that will feed into state-level policy decisions

This points to what Noveck has coined "combinatorial democracy," instead of the old approach where resource constraints forced us to pick one method off the menu—either a deliberative dialogue OR an expert consultation OR a public survey—AI tools are making it possible to knit together multiple forms of engagement. As Noveck explained in the workshop, we no longer have to choose between different approaches to participation; we can combine them strategically.

New Jersey's recent AI and workforce task force demonstrated this approach in practice. The team could simultaneously ask thousands of people about the problems they were experiencing, run deliberative dialogues with representative citizen samples, and consult domain experts—all feeding into a single decision-making process. The traditional choice between legitimacy and expertise, between scale and depth, was no longer necessary.

From the Margins to the Mainstream

Allen made a crucial point about how institutional change actually happens: "transformation often moves from the margins to the center." The innovators are rarely the people at the center of power—they're often those facing crisis, those with bandwidth to experiment, those who need something different because the status quo isn't working.

In the upcoming workshop sessions, we’ll continue to learn from these experiments. Claudius Lieven  will share how they built the DIPAS platform for citizen engagement in urban planning—starting locally but now deployed in nine different European cities. The Belgian company Go Vocal will share their open-source platform which has worked with 500 different communities. We'll also hear from places like Copenhagen and Vienna, which are pioneering new ways to combine online and offline engagement that actually connects to municipal decision-making. 

What's different about this moment is the shift from abstract theory to nitty-gritty practice. Because the real barrier isn't technological anymore. We have the tools. The barrier is knowing how to use them in ways that connect citizen voices to institutional power.

The Infrastructure Challenge

Multiple workshop participants raised concerns about depending on private, unaccountable platforms, with the potential for CEOs who might change the rules on a whim. Others worried about AI systems amplifying existing biases rather than correcting them.

These are legitimate concerns that require ongoing attention. The communities using these tools—government agencies, civic organizations, and citizen groups—need to actively shape how AI platforms are developed and deployed for democratic purposes. This means advocating for transparency, pushing for algorithmic accountability, and demanding that platforms serve democratic values rather than just commercial interests.

This is why we need more experimentation across different tools and approaches. More use cases, more communities trying different methods, more practitioners sharing what works and what doesn't. The goal isn't to find the perfect platform, but to build collective knowledge about how to use these technologies responsibly for democratic engagement.

Making Participation Routine

The ultimate goal remains what we outlined in our first post: moving from episodic engagement to systematic democratic governance. Instead of citizen participation as special events, we want it integrated into institutional practice, as routine as budget planning or performance evaluation.

AI tools can make large-scale public engagement far more feasible by reducing the time, cost, and complexity that usually make it impractical. For example, the UK Cabinet Office estimates that analyzing 30,000 consultation responses would normally require 25 analysts working for three months. With such resource demands, it’s understandable that many institutions limit themselves to small-scale or symbolic consultations.

But when AI can help synthesize thousands of responses in hours rather than months, when translation barriers disappear, when we can combine multiple engagement methods without multiplying costs—then systematic participation becomes feasible.

The Reboot Democracy Workshop Series is designed as a community of practice for this transformation. We're bringing together the doers—not just the thinkers—who are figuring out how to make this work in practice.

A final note: If you're experimenting with AI-enhanced engagement in your community, your agency, or your organization, we want to learn from you. The "plus" in our "11+ workshops" is an invitation. This infrastructure won't build itself, and it won't be built by technologists alone. Please reach out to [email protected]  We're building this community of practice together, and your insights could shape future sessions.

The Reboot Democracy Workshop Series continues September 24th with People Powered discussing "Which Tools for Which Democratic Purposes." [Registration and full schedule available here.]