Wayfinder journal
Questions, wrong turns, fragments, and things worth carrying forward.
Brian's thoughts
I don't want to write only when I have an answer.
Some of the things worth leaving behind begin as questions, mistakes, frustrations, or something I can't quite explain yet.
This is where I'm going to keep some of those thoughts.
Some will become projects. Some will change completely. Some may turn out to be wrong.
That is part of the point.
If something I learn can give the next person a better place to begin, I want to leave it here.
Working thoughts
Illustrative journal sketches — based on Brian's current thinking and created for this page.
First substantive note
The projects did not stop at the questions that started them. As we worked, some of those questions became more specific. That is worth writing down.
When we first wrote about the Wayfinder projects, most of the descriptions were about why they existed.
That was the right place to start.
But projects change when you actually work on them.
The questions got sharper · continued
The questions get sharper.
Sometimes the thing you thought you were building turns out to be only one part of the problem.
Sometimes an annoyance becomes a much bigger question.
And sometimes the best sign of progress is that you can finally describe the problem more clearly than you could before.
The questions got sharper · continued
Waymorrow started with the belief that a good idea should have a chance.
That is still true.
What has become clearer is the kind of help I want it to provide.
Waymorrow is being developed around Abriella as a human-assistance platform: a place where someone can bring an idea that is still messy, find the words for what they mean, understand what is in front of them, and work toward a next step without handing away the decision.
The questions got sharper · continued
Wayforge began because I wanted our own way to build.
That has grown into a broader question:
How do we organize complicated work so research, planning, building, checking, and review can work together without hiding responsibility?
It is still an internal Wayfinder tool.
The questions got sharper · continued
NCS began with information overload.
That problem is still there, but “less noise” was not specific enough.
Information also needs a trail.
Where did it come from?
What supports it?
What contradicts it?
What are we still unsure about?
If we cannot answer those questions later, then organizing the information was only part of the job.
The questions got sharper · continued
SOVARITAS started because repetition and cost bothered me.
Then the question grew.
Instead of only asking how to carry less repeated context, I started asking what it would take to understand more of the model foundation itself.
Tokenizer.
Architecture.
Weights.
Corpus.
Runtime.
Evaluation.
Reasoning.
Not because writing down those words means the problem is solved.
It means the work has become more specific.
The questions got sharper · continued
A project description should not become permanent simply because it was true when we wrote it.
If the work changes what we understand, the public description should get better too.
Otherwise we are doing exactly what Wayfinder is supposed to avoid:
leaving the next person an outdated map.
What I think we learned · continued
The projects are not finished.
The questions are simply better than they were when we started.
And that is worth carrying forward.