Projects / 01 · Ongoing work
Building the
collaboration.
A model is one part of the work. I'm building the environment around it: the context, tools, memory, and handoffs that help a person carry a project forward.
- Focus
- AI harnesses
- Stage
- Building and iterating
- Perspective
- Human direction, inspectable work
The work
Give the collaboration somewhere to live.
I'm working on tools for keeping track of projects, carrying context between sessions, and turning conversations into something concrete that can be inspected.
Parts of that environment include a local inbox and project board, a harness for scoped work, and ways to bring captured notes back into a session. They are working tools I continue to develop.
Where it lives
On my own computer, where I can read it.
The inbox and the project board are files on my own computer. I can open them and see what a session was given and what it did. I think that matters because the direction of a project should stay with the person, even when the model or the tool changes.
The questions
What makes this useful in practice?
Context that survives.
How much can a new session pick up without making the person explain the whole project again?
Work you can examine.
What lets a person see what changed, what was checked, and what still needs a decision?
Room for human judgment.
Where can the system keep working, and where does it need the person to choose?
One consequence
Sometimes the product leaves AI out.
The Return Authorship app grew out of my experiences building with AI. It is deliberately a no-AI space for attention and choosing a direction.
The Return Authorship app ↗How it runs now
Two chains and a ledger.
Building moves through two named chains of work that call on four agents, and nothing goes into the ledger as kept or discarded until I say so. I wrote down how that works today.
How the work moves now ↗Follow the work
More of this belongs in public.
This is where I'll share the tools, experiments, and lessons from working at this frontier. It starts with the story of how the harness came to be, told in three parts. Part one: The repo that lasted three days. Part two: Letting it run.
I’m putting this work to use for small businesses in the Hudson Valley. A free assessment finds where AI can help, then I build one workflow and hand it over. See how we can work together.