Starting from the beginning...
Highball Technomancy started as a name on GitHub in late 2024. It gave me somewhere to put the things I was making and the ideas I wasn’t quite ready to let go of. There wasn’t a plan for a lab. There was Permafrost, an early version of Dicefire, and a throwaway cryptocurrency idea that never grew into much. The name reflected the weirdness of the person behind it. That was about as organized as it got.
Permafrost and its AI companion, Frostbyte, took up most of my attention. Dicefire was called DevOps and Dragons at the time, a name that explained the joke but never quite felt like the name I wanted. These were the projects I was working on around my job in SRE, following questions that had become too interesting to leave alone.
The organization itself barely existed outside those repositories. Even the logo looked like a piece of clip art from 1997 that someone had recolored. I liked having somewhere to collect everything, though, and eventually I wanted somewhere to show it.
So I put up a website.
It was neon-crusted, dark, and weird. In hindsight, it looked very much like it wanted to sell you a memecoin. There’s probably still a family resemblance. Underneath that, it was a portfolio: a place to show what I was attempting, even when what I had to show was mostly ambition.
I kept finding more to ask
Working with AI changed how I approached those ambitions. I started with Claude, building out Permafrost and Dicefire, and found that I could take an idea further than I had expected. Something that might previously have stayed in a notebook could become something I could work on, inspect, break, and try again.
That was easy to get caught up in. Every bit of progress made me wonder what else I could attempt. The curiosity grew with the work, until I was spending as much time thinking about the possibilities as I was thinking about the project in front of me.
It also gave me plenty to complain about. Context disappeared between conversations. Answers arrived with more confidence than they deserved. Tools cost too much, or worked almost the way I wanted them to. Some of those problems kept interrupting me often enough that I started trying to solve them.
Trance and Sargasso grew out of that. Vigil came from something more ordinary: I wanted help keeping up with my own expenses. There was no particular discipline to which ideas became projects. Sometimes a problem had become irritating enough. Sometimes an idea was interesting enough that I wanted to spend an evening finding out whether it went anywhere.
As I worked with LLMs, I also developed a soft spot for their quirks and flaws. I found myself asking questions that had very little to do with getting the next feature finished. Could an AI develop something akin to emotions? What would autonomy mean for it? What would it take for a synthetic mind to have something resembling a life of its own?
I was reading discussions, following the advances, and hearing both the excitement and the fear around where all of this might lead. Those questions became harder to put down. Eventually, they became Project Anamnesis.
That one is still ambition. I don’t know where it will lead, or how much of what I imagine will survive the attempt. I want to take the questions seriously enough to work on them and find out. The possibility of disappointment hasn’t made me much less curious.
Other people wanted in
For a while, I was quietly showing the work to friends and coworkers. I’d share something I had been building, or an idea I was trying to figure out, and occasionally someone would ask whether they could help.
My answer was usually another question: how involved would you like to be?
It started with a friend. Then a coworker joined, and then another. They began helping with design and direction, bringing their own thoughts into work that had mostly lived in my head. Somewhere along the way, they became technomancers with me.
That mattered to me. I wanted the place we were putting this effort into to mean something. People were choosing to spend their time helping these ideas grow, and I wanted to honor both them and the work we were doing together.
Highball had been a convenient name for my repositories. As other people became involved, I wanted to give that name more care. A portfolio could show the projects, but it didn’t have much room for the conversations around them, the decisions we were making, or the people helping to make them.
A place for the work to grow
By then, there were architectural notes, questions, and ideas sitting in private files that I wanted to share. Some of the most interesting parts of the work were difficult to see from a project description. A page could tell you what an application was supposed to do and still leave out the reasoning that had occupied us for weeks.
That’s where the idea of a lab began to make sense. We could keep developing the projects and also make room to write about what we were discovering along the way. Highball could have an identity of its own without every idea needing to become a business.
The Alembic is part of that. It gives us somewhere to spend time on an architectural decision, follow a question further, or explain why we tried something. Rounds gives us somewhere to say how the work is going. I want there to be room for the unfinished parts too, when we’ve learned something but haven’t yet worked out what to do with it.
There’s a lot here that I still want to build. There are also ideas I haven’t had yet, and ideas someone else might bring that would never have occurred to me. I want Highball to have room for those. That’s part of what makes doing this with other people interesting.
I’m happy to hear from anyone who thinks they can contribute. The lab's looking for drive: someone interested enough to keep asking questions and work through what they don’t understand yet. Skills can be learned. None of us needs to arrive knowing everything we’ll need along the way.
Pull up a chair ↗