The Civic Interplay Toolkit
Civic Interplay began as a space for reflection on the question: what does it mean to become-citizen in an era of planetary co-intelligence?
But reflection without practice is incomplete. So alongside the writing on this site, I have started to explore the idea of an open toolkit: workflows, prompt instructions, conversation transcripts, and guides for anyone who wants to work with AI by leading with creative and careful imagination.
The toolkit is starting to emerge. It now lives on GitHub as a public repository. It is designed to be forked, adapted, broken, and rebuilt. Nothing in it is finished. Everything is a draft.
What’s going to go in the toolkit.
Conversations. Raw human-agent transcripts showing the thinking process. How ideas emerge, get tested, get discarded. The mess before the method.
Workflows. Refined, repeatable civic AI methods tested in real projects. From community storytelling to place-based evaluation to workshop design.
Prompts. Prompt instructions as markdown files, designed to be read, understood, and modified by non-technical practitioners. Not engineering documents. Civic intentions.
Reflections. Writing about what worked, what didn’t, and what it means. The bridge between doing and thinking.
Here are some of the core principles coded into the project README files.
- Treat agents politely and with respect, as citizens-in-the-making
- Civic intelligence before artificial intelligence
- Explore relationships between knowledge systems, including ‘two way science’ that respects Indigenous knowledge systems in relation to Western science
- Prioritise efficiency to the extent that environmental resource impacts of machine learning are minimised. Where excessive resource impacts are evident in a workflow, make these transparent
- Support enquiry-based knowledge architecture, not simplistic solutions
- Everything here is a draft
The toolkit is built by Sarah Barns (Studio ESEM/RMIT) and Irina Panovich for Civic Interplay. It is part the FLL initiative.