What I Believe

The wrong question

Everyone in my field is asking the same thing: will the machines replace the engineers? I think that's the wrong question. The agent can write the code now — that part is settled. So the work I care about isn't automating engineers out of existence.

Keep the human in the loop

I build agents that keep the craft, the judgment, and the accountability firmly human. That means real tool use, long-term memory, and multi-step reasoning — but also a check-with-me-first step before an agent does anything that actually matters. The engineer stays responsible for the outcome; the agent is the layer I work through, not a replacement for the thinking.

Local-first, by conviction

The AI I lean on most runs entirely on my own machine — one hundred percent on-device. My notes, my work, my second brain: none of it phones home. Local-first isn't a constraint I tolerate; it's a principle I build toward.

I actually believe this

This isn't positioning. I actually believe the stuff above — and I've built things because of it. See What I'm Building and the flagship, Candice.