Why does this page exist?

I’ve been following AI closely since the first major wave of generative AI started moving into everyday use. Along the way I’ve built my own systems, tested models and tried to understand what is actually happening around the technology, not just what is being said about it.

That has produced an absurd amount of information. Much of it is easy to lose under technical jargon, marketing language and whatever the algorithms happen to amplify that day.

That is why this page exists.

Dispatches are a translation layer

The point is not to showcase every project I’ve built or document every line of code. A more useful question is: what can any of this actually do for a person?

One article might look at why an AI agent should not be given unlimited permissions. Another might examine how to recognize a weak claim. Elsewhere I might write about algorithms, investment risk, automation at work, disinformation, language and thinking, or the way organizational metrics can start steering an entire system in the wrong direction.

Those subjects can look unrelated from the outside, but I often end up asking the same question: how does this system actually work, and what does it do to people?

You do not need the vocabulary first

Technology discussions often carry a strange assumption that you should already know the terminology before you are allowed to understand the subject. I think the order should be reversed: understand the phenomenon first, then give it a technical name if that name is useful.

If AI agent permissions can be explained in ordinary language, I’ll use ordinary language. If something needs a diagram, I’ll make a diagram. Complexity does not need to disappear, but it does not need to be dropped on the reader all at once either.

Why now?

AI is no longer a small technology topic sitting in its own corner. It affects work, media, education, software, economics, creative practice and the way information is produced and distributed.

At the same time, the public conversation around it is shaped by algorithms, commercial interests and constant competition for attention. In one direction, AI is supposedly going to solve everything tomorrow. In the other, the whole thing is dismissed as a bubble. Reality is usually somewhere in the much more interesting space between those extremes, and that is the space I want to explore here.

Why Anomancer?

Anomancer is an entrance. It is one of several alter egos I use, but I chose this one for this role because the name describes what I do surprisingly well.

I am interested in anomalies: places where something does not fit the expected model, behaves unexpectedly or reveals something about a system that would otherwise remain hidden. I have also spent a fair amount of my own life slightly outside the usual boxes.

That is why Anomancer is not only a name for technology or research. There is another idea inside it: an anomaly is not automatically an error.

Not everyone has had an environment, school, workplace or group of people that knew how to recognize their particular way of thinking, learning or creating as a strength. Sometimes an ability stays invisible simply because it appears in the wrong place, or in a form the surrounding system does not know how to recognize or reward.

Part of what I want to do with my work is push against that. Useful things can come from unusual starting points, and a real path does not have to resemble a ready-made career template. If something I build gives a little more confidence to someone who has felt out of place, late, overlooked or simply wrong for the available boxes, that matters to me.

Maybe that is why the name fits: I study anomalies, build things around them, and I am one myself.

Who am I?

I have followed AI closely since the first major wave of generative AI began breaking into everyday use. The last three years have been unusually intense. Models, tools, interfaces and the conversation around them have sometimes changed in months, sometimes in weeks.

I have followed the global development, but also the way the same phenomenon appears in Finland. I am especially interested in the gap between what the technology can actually do and the way people talk about it. That gap fills quickly with uncertainty, exaggeration, dismissal, bad information and straightforward disinformation. Algorithms make it stranger by rewarding what produces a reaction, not necessarily what helps people understand what is happening.

These years have been intense on a personal level as well. I have spent a huge amount of time building, testing, reading, comparing models and trying to keep up with a field that does not wait around. It has been fascinating, but it has also been difficult at times. That is probably one reason I have become increasingly interested in making all of this easier for other people to understand.

My original background is not in computer science or AI research. I trained as a practical nurse and disability support worker, and I also trained as a ventilation installer. There was no ready-made route from my education into AI systems, language models or software architecture. I mostly ended up here by making things, studying them and connecting areas that do not necessarily look related on paper.

That probably explains something about the way I look at technology too. I am rarely interested only in how a system works. I am just as interested in what it does to people, what kinds of behavior it creates and what happens when a technical system collides with the real world.

And Core?

Anomancer is the entrance and the Dispatches are the human-readable layer. Core is the private machine room where agent contracts, orchestras, runs and authority boundaries are visible in one control plane.

You do not need to see all of it at once. Finding one useful thing is enough.