I want AI to know what I’m working on

Every time I open a new AI chat, I end up explaining the same things: what I’m building, what I tried, what changed, and why I’m asking this question in the first place.

That’s the problem I’m trying to solve with Transcripted.

Meetings and dictation are a start. They capture conversations and ideas I’d otherwise forget. But a meeting isn’t the whole story. I might write a follow-up that changes the plan, talk through it in a message, or spend the afternoon figuring out that my original idea doesn’t work. The useful context is spread across all of that.

So I’m experimenting with Writing as the next piece of Transcripted. Part of it helps me write. The part I’m more curious about is whether the writing I choose to keep can give an AI a better understanding of the work I’m doing. I’m also thinking about how a little context about my activity could connect the dots between a conversation, a decision, and what I actually did next.

I don’t want Transcripted to record everything just because it can. I want to be able to choose what’s kept, have it in files I own, and use it with whichever AI I want. If a better model comes along next month, my context shouldn’t be stuck with the last one.

I’m still figuring out what’s useful here. Writing is an experiment, and I don’t have the activity piece worked out yet. But the direction feels clear to me: I want to spend less time bringing an AI up to speed, and more time working with one that understands what I’m working on.