Every morning at 7am, I get a Telegram message telling me exactly what to work on today.
It’s from my OpenClaw agent running on AWS. It picks 1 task from my Linear backlog and sends me a bulleted action plan breaking it down into simple steps. This has been one of my most successful experiments in integrating agentic intelligence into my life outside of coding.
The task is always something small — 15 to 30 minutes of real work. I’ll often just have the agent handle the implementation too. But even when I do, it sets the tone: I’m making progress on things that matter before the day gets loud. Getting in the habit of winning early in the day translates to winning the rest of the day in my opinion.
All it takes is one task.
It’s also nice to not have to use the strategic part of my brain in the morning. As someone who likes to plan a lot, I appreciate the certainty of having at least one decision made for me.
I think people get distracted by lofty grand fantasies of AI completely taking over their lives, but this to me is a dystopian nightmare. I prefer gentle guidance and assistance. I want to be in control and in-the-loop. I want to be the agent of change. The AI is a commodified, specialized tool, and not a magic pill to blindly administer to all my problems.
New and featured essays sent directly to you. No spam, no fluff.