Last spring, I had the pleasure of working with a team of Tufts computer science students who built Sage, a chatbot that can help people who face a threat or consequences of wildfire to “take informed action and work together with their communities to rebuild and move forward.”
The underlying theory is that effective groups need documents: recruitment messages, agendas, budgets, job descriptions, mission statements, and more. An application of a large language model can not only draft such documents but also help people see what documents they need.
Democracy and local communities would be stronger if more groups of concerned people gelled into effective organizations that collected time and money from their own members, used their resources to build their own capacity, made collective decisions, and acted.
I am all for teaching people how to do those things, but it’s unrealistic to offer civic education of this type to millions of adults. An alternative is to help groups organize themselves effectively so that they can focus on the substance of their work. And one way to help them is to draft useful documents.
Since 2025, I have offered a Civic Helpdesk on this site that can help people address any issue. The Tufts students who built Sage chose to focus on one topic, wildfire. They built a considerably more sophisticated application than mine. It is informed by a substantial set of documents that I helped to choose for them, reflecting what I consider good practices for organizing voluntary groups. It favors paid local journalism as its source of information. And it encourages people to get off the AI and interact with other humans.
Sage could be relatively easily repurposed for other issues.