Working with AI shouldn’t feel like starting over
Introducing the Building a Better Context for AI series!
Hi friends,
I’ve been quietly working on something new—and it starts with a pain I think a lot of us have felt when working with AI: the awkward gap between what these tools could do and what they actually help us get done.
To explore what I mean by that, I’ve put together Building a Better Context for AI.
Building a Better Context for AI is a short series about a project I’ve started around AI context management—starting small with a web-based chat and a command line tools, but with bigger ideas in mind. Whether you’re a developer, writer, or just someone who’s tired of repeating yourself to your AI sidekick, I hope this resonates.
Why context, why now?
Like many people building and writing alongside AI, I’ve hit some friction points.
Smart tools that don’t remember what matters.
Conversations that feel productive in the moment but don’t stack up over time.
Workflows that feel more like hacking around limitations than designing with intention.
Instead of waiting for someone else to solve it, I’ve started exploring the idea myself—starting with a simple tool to help manage, persist, and reuse context across AI chats.
The idea isn’t to create another flashy AI app.
It’s to build a small, sharp tool that reduce friction, make collaboration smoother, and let us carry forward what we’ve already figured out.
Why I’m exploring context, one small tool at a time
I’ve been thinking a lot about context—not just as a buzzword in AI, but as something more foundational. A quiet, often invisible layer that shapes how we work, communicate, and build things.
Most of the time, context is something we manage intuitively.
We reference what came before.
We assume shared understanding.
We keep certain threads alive and let others fade.
But when we work with AI, a lot of that context disappears.
Even the best tools today make us repeat ourselves, lose track of ideas, and start over more often than we’d like.
That’s what this series is about:
Not how to make AI smarter—
But how to work with it in a way that feels more human, more grounded, and more useful.
What to expect from the series
Over the next few posts, I’ll walk through:
Why prompts aren’t enough
How I define “context” in this space
What I’m building first and why
My broader thinking about modular, composable context
How this might grow (or stay small and useful)
This series is a mix of idea sketching, tool-building, and systems thinking. I don’t have all the answers—but I think the questions are worth exploring out loud.
If you’ve ever wished your AI tools could just remember what you’re working on—or carry a little more of the load without so much copy-pasting—I think you’ll find something here that resonates.
Thanks for following along,
Adeline