Kia ora {{ First name | friend }},
We’re back in the flow of regular emails. Welcome to the party.
Last week, we gathered in community for a community call about AI. 20 people showed up from Aotearoa, Australia & Singapore. We discussed new opportunities for working with AI in a way that works for you, rather than telling you what to do.
Here’s a few things we covered;
AI on your team - shifting from copy-pasting into ChatGPT to building automated workflows that run in the background. AI as a team member, not a search engine.
AI Brain - A sovereign knowledge base that lives on your machine. Stores decisions, voice notes, project context — compounds over time, readable by any AI. Ties in to the AI on your team above.
AI privacy and ethical concerns - how we navigate the fast-paced evolution of AI, and maintain a sense of humanity, creativity, privacy and inner authority. Basically, how do we use AI without losing our humanity?
I’m feeling really stoked about the opportunity to gather as real humans and find coherence together. If you’re stoked too, the next call is happening on Tuesday 2 June 7pm - 8:30pm NZT. Join here.
Build your own sovereign AI brain
Gif by Nazoom1 on Giphy
I’m gonna be harping on about this for a while, until the general public catches up. If you got the download already, skip this section.
The most valuable AI action you can take right now is to build yourself some brain infrastructure. A permanent memory that lives on your machine, accessible anywhere you can access internet, that knows your work across multiple contexts (I have four brands built into my brain ecosystem), and grows it’s knowledge over time.
Bonus round, you own your data, and can edit/delete it whenever you like.
This is the single most important step in shifting from “Advisory AI” to “Autonomous AI”.
To do this, you need to
Get started with Codex or Claude Code (don’t get wigged out, I literally have not written a single line of code since beginning this)
Make a file on your desktop to work out of (mine is just called “workspace”)
Download the AI Brain Starter Kit and run through the onboarding steps I set up
Open Codex/Claude Code and tell it your first project you want to build
Start feeding context and watch it grow over time.
PS: If you want to try setting up your own brain from scratch without the frameworks I’ve set up, you can download the GitHub repo and give it a go yourself here. I set up the Starter Kit to make it a lot easier to install and set yourself up for success, adding in autosave memory, project-specific context and a few other fun bits you’ll see inside.
Give your brain some limbs (and a heart, ideally)
You can significantly increase the value of your AI brain by connecting it to your tools. So it’ll remember things, AND ALSO it becomes the command centre of your business. Guiding it with heart (your mission/intention/kaupapa), and taking action (building websites, activating workflows, etc).
Here’s some ideas from my workflows
Build new landing pages from a voice note, like this one
Set up autonomous actions that work on a trigger (when email received etc), or on regular intervals (every Monday 9am)
Connect apps together for seamless website experiences, like this one with Zoom, Airtable and Stripe as key components
Take voice notes from my phone and immediately ingest them into the brain context to feed my other mahi
Receive clips (videos/images) and schedule them in draft for posting on social
I’ve created this AI Brain Starter Kit to create a smooth experience for you to get started with building your own brain. I humbly stand on the shoulders of giants here, and have just upgraded the experience based on my own fumbling lessons.
Cohort based learning
I just want to be real here. These systems are not perfect. They can be clunky and downright stupid.
In my follow-up email from the call, I sent out an expression of interest form to join a cohort program to build together. I didn’t realise that the AI hadn’t connected the form responses to my database. So anyone’s form response ended up in the infinite void of the formless internet. Oops. Partly my bad for not double checking, and partly the AI’s bad for not using existing database infrastructure to run the form through.
If you did fill it out (sorry about that), or you happen to be interested otherwise, you can express your interest here. I’ve made sure my AI brain never makes this mistake again… And will make sure to integrate this lesson into the cohort… :)
Here’s a few observations from the AI space this week
OpenAI and Anthropic are both racing toward persistent memory features. They basically want to be your “brain” rather than have you own your data. The direction is clear - your personal AI should know your context. The difference: theirs lives on their servers. Yours can live on your machine. If you don’t own your data, you don’t own your system…
The gap between "uses ChatGPT for life admin" and "AI runs their business" is narrower than it looks. It's not intelligence or coding knowledge. It's an internal shift towards seeing AI as something that can actually work for you. I’ve got whole websites and workflows running on code, having never written a line of code myself.
Bonus section: Voice to Action
Use your voice to talk to AI. I cannot recommend this enough. Superwhisper enables you to hit a keyboard shortcut on your laptop and vocalise what you want to do. This enables you to voice dump (I’ll literally go on a 10min yarn) and have your AI ingest it all, organise your thoughts and reflect it back to you - or better yet, take action for you.
Keeping it real
It’s becoming increasingly clear to me that the human, connected, grounded, relational, ancestral intelligence (the real AI) is fundamental to establishing right relationship with AI in 2026. If you’re down for this, come join the next call Tuesday 2 June 7pm - 8:30pm NZT. Join here.
See you next time
Billy