Day One with Clawdbot


I spent a full day tinkering with a system that runs on my server. Not an app. Not a dashboard. A terminal. And by the end of the day, I had built something that thinks alongside me.

The structure

Morning — A briefing arrives. A quote from Marcus Aurelius or Montaigne. Weather. My calendar. Five news items I actually care about. My priorities. Not everything. Just what matters. The day starts with clarity.

Throughout the day — Ten minutes before each meeting, a ping. But not just a reminder. Context. Who's attending, what we last discussed, relevant threads from email and Slack. The system distills the meeting's narrative and purpose so I walk in prepared — not scrambling to remember what this is about.

The philosophy

Someone asked me: why use AI for things you could do yourself?

Because I'd rather spend that mental energy on the person across from me. On the idea I'm presenting. On listening — actually listening — instead of half-remembering context. AI doesn't replace presence. It creates the conditions for it.

I tell my team that learning to work with AI is like building new reflexes. Looking at every workflow and asking: where's the friction? It's genuinely fun.

A larger thought

We're going back to terminals.

GUIs were an abstraction — visual interfaces built so humans could point and click their way to getting things done. But now? You type what you want. The system does it. No buttons. No menus. Just language in, action out.

The black screen is back. Not because we're regressing — because we've evolved past needing the visual layer. The computer reads, studies, gathers context. We direct.

This is day one. I'm curious what day thirty looks like.