On Friday the 10th of July, we hosted our first "Vibecode Show & Tell." We know that's something that might raise a few eyebrows, given the polarization around AI. Here's why we did it...
August 5, 2026
Reason 1 - Vibecoding is a legitimately creative space, but creativity alone isn't enough.
There's real energy in civic tech right now, with practitioners suddenly holding powers that previously only belonged to full-stack software developers. Staff inside city governments, community newsrooms and philanthropies (among others) can now spin up hyper-specific, well-informed tools to do their jobs better. That's genuinely exciting.
But creativity disconnected from context or the constraints of practical implementation tends to fail. There are a lot of folks reinventing the wheel and duplicating the last decade-plus of civic tech tools with little idea of what's already there, or why those exact products didn't work before. The enthusiasm is great, but if the tools are redundant, what are we even doing? Also, it risks diffusing attention on the innovations that already exist and work.
Add to that the fact that while the cost of creating things might appear near zero now (assuming you have, or someone is providing, open civic data), rebuilding the same tools again and again can have other costs, not least of which is the environmental impact of using resource-intensive cloud-based AI. It's also hard to find a purpose for another legislation explainer unless there's something notably innovative about it. So your own time, while ostensibly saved in development, might be better spent elsewhere.
Reason 2 - We need bridges between the individual, the social, and the technical.
Somewhere along the way, the concept of "innovation" was collapsed into "tech," and "tech" collapsed into "software." It's an understandable slide. Software is the kind of innovation that people touch every day. You don't need a research university or corporate lab. You only need a phone. But the consequence is that the tech industry now largely leads us wherever it decides to go, and all of our social, political, and cultural practices are compelled to follow and adapt.
Civic tech, at its best, doesn't buy into that arrangement.
Projects should be and can be informed by human interaction and experience. But that requires us to connect with each other and a community with real institutional knowledge. It also requires us to connect the tech with real civic contexts.
Holding these events allows us to make both connections at once. It's what we did last month, and we hope to do more of it in the future.
Which brings us to the elephant in the room...
Yes, we're aware of the irony here. We're reaching for tech to address the tech’s societal influence. That's a tension that unfortunately touches a great deal of our social and political lives at the moment (e.g., activists compelled to communicate via social media). And it's not something we can solve in one swoop.
But there's no denying the trend. We're getting more and more vibecoded submissions all the time. And one thing we can do, in the short term, is try to foster those aforementioned connections and promote best practices. Not to mention learn from each other.
The presentation on Civic AI Tools by Nathan Storey (our inaugural guest) did just that, by demonstrating a technical bridge that serves our social one. Nathan discussed a Model Context Protocol (MCP) that connects open civic data straight into an entire system of AI tools, grounding them in actual human experience.
And the discussion that followed amongst our attendees proved the point of the whole exercise. The questions weren't about the novelty of the tech. They were about making it work in the real world:
How do you fund a tool like this long-term, when token consumption costs money? Can any of it run on local models, or does using it presuppose a cloud subscription? How valid is the data once an LLM (prone to hallucinating) sits between it and the user?
These are exactly the context-and-constraints questions we wanted to draw out, and they're the kind you only get by connecting people with each other.
Perhaps the best question of the event, however, was also the simplest…
"What does civic tech need from AI?"
Not the other way around. And that framing is the whole point. The AI isn't the innovation we're discussing. In fact, the tech, in and of itself, rarely is. For us, the most significant innovation is how tech connects with, is informed by, and benefits people.
If you'd like to show off your work in a future show & tell, let us know.