Thriving Quest #001: No-stress email (with agents)

Email is the perfect place to start my questing.I like the term “quest” as a unit for a dedicated Thriving initiative because it invokes a sense of adventure while appropriately balancing the value of the search with a definitive win state at completion.

It’s a constant thorn in my side that I’ve tried to solve a half dozen different ways over more than a decade. I’ve tried auto-tagging, Superhuman, Cora, and even hiring a virtual executive assistant. I’ve never been happy with the results.Cumulatively, I’ve spent thousands of hours and many thousands of dollars on this problem. I always regress to some uneasy truce of ignoring most emails and accepting a constant, vague sensation of letting someone down. Like I know I missed something important, but it’s so buried in my emails I have no idea what for or to whom I should apologize.

Even as I struggle to navigate the inbox deluge, I know email is critical to cultivating my surface area for connection. Both personally and professionally, responding quickly and thoroughly to emails is a sign of respect and diligence. Over time, communication done well yields broader and richer connections with power-law compounding.For a fun read on power laws, check out Scale by Geoffrey West (opens in a new tab).

I’ve had glimmers of a brighter future with skillz I’ve created. Now it’s time to crystallize that vision into reality.

My first Thriving quest is to increase my capacity for connection with agent-native email. Specifically, I’m building (hiring?) a cast of agents to triage, route, draft, follow up, and maintain the system - with my oversight.

Five days in, this is already real: my new mailbox receives on hardware I control and sends through a self-managed relay. The quest hub is live, Slack is live as the orchestration channel, and the first agent seats are defined. The newsletter relaunch is next.

Agent-native gives my judgment better memory, better context, and better timing. Nothing sends without my approval.

From a Thriving with AI learnings perspective, I’m using this quest to practice:

• Slack as dispatch harnessI have a growing hypothesis that I need a centralized dispatch to receive agent updates and send new instructions or follow-ups. Slack works well enough for huge teams of humans, so let’s see how it works for a cast of agents.
• Hierarchical agent orchestrationI want to get better at having one smart, expensive model kicking out to the dumbest, cheapest model that will still do an excellent job.
• Autonomous background sessions
• Multi-vendor switchingSol (my orchestrator agent) + Codex (the harness) have got me back to paying OpenAI, and now I’m juggling Fable + Sol. In the not too distant future, I want to start working with open weight models that are a fraction of the cost. To get there, I need a decently seeded golden set so I can run rigorous evals to ensure I don’t have performance regression.
• Recursive self-improvementI’ve been thinking bunches about Anthropic’s June article (opens in a new tab) and the zeitgeist around “tending loops” right now. Very keen to start practicing those techniques because they have accelerating flywheel effects and apply to every agentic build going forward.

From a Thriving with AI learnings perspective, I’m using this quest to engage with:

• Tech independenceInspired by Derek Sivers’s tech independence page (opens in a new tab). A very worthwhile read.
• Big feels around how I show upFor all the practical tacticals, there is an equivalent amount of long runs and tea meditations needed to rewrite some limiting narratives about who and how I am.
• Foundations of household admin
• Build in publicI’ve wanted to build in public for a very long time, but I’d never found a project that felt inspired and worthy enough. Until now 🙂
• Relaunching my newsletter

I have succeeded at this quest when I feel resiliently independent in my email infra AND correctly handle and respond to all emails within 2 business days.In an unexpected way, I’ve started referring to this inbox as my “forever home” on the digital plane. I have never owned physical property. When I forecast into owning land, I imagine a very similar complex emotional landscape of unbridled joy and stability from independent ownership with a healthy dose of a dauntingly never-ending improvement to-do list. Hopefully, my cast of agents doesn’t become its own money pit (opens in a new tab).

I expect this quest to take me A LOT of time up front to set up the core, then a very long tail (3-6 months?) of constant tweaking and refinement.

When my agent-native email works as intended, I’ll experience less energetic drag and have more capacity to care for more relationships, more thoughtfully.

Email is foundational to how I show up for people. Getting it right compounds capacity for everything that comes next. Let’s get questing with #001!

If you’d like to follow along, live build notes are on the quest hub. There you can sign up for email updates or get an RSS feed to follow along without another algo-maniacal platform.