I’ve been experimenting with how a personal AI assistant could help me follow through on the things I care about. That means giving it some understanding of my goals, my preferences, and the direction I want my life to go.
In this video, I ask my dot, Shaine, to start a ChatGPT thread and use my personal notes in Obsidian to propose a life vision document. The idea is to give us something we can refer back to as we work together.
Along the way, I discover something I hadn’t realized it could do: my dot sends additional context into the thread while it’s still working. You can see the moment I notice it in the recording.
The document itself also surprised me. It drew on journal entries from 2024 and reminded me how much progress I’d made on things that mattered to me. When you’re always thinking about the next milestone, it can be easy to lose sight of that.
I read through the proposal and made some edits. Then I asked how we could use it to guide our ongoing work together. I also raised a question about whether computer history could help identify when my time and attention drift away from my priorities. That access wasn’t confirmed in this recording.
This is an early experiment. Whether it helps me follow through over time is still an open question. But the reflection alone felt worthwhile, and it’s a direction I want to keep exploring.
What would you want a personal AI assistant to understand about you before asking it to help with your goals?
