AI State of the Union

At the beginning of February I put together a small map of the AI models and tools I was actually using. I wanted one place where I could always see the latest and greatest parts of my AI setup. A couple of days later it had a name: State of the Union.

The name is somewhat serious. The project is not.

I had a simple problem. Every AI company has model families, variants and product names, and every coding tool supports a different part of that list. The names change quickly, new models replace old ones, and a bookmark says nothing about which tool I would actually choose today.

State of the Union is my answer to that. It is a living map of the AI tools and models I currently consider useful enough to keep in my own setup. The tools are connected back to the models they use, so I can see not only what is available, but how everything relates.

It is deliberately opinionated. Bright nodes are things I currently use, while darker nodes are alternatives I am watching or keeping for context. There is no score, benchmark or attempt to include every model announced on X yesterday. The point is not to document the whole AI industry. It is to show the current state of my part of it.

Clicking a node shows its provider, current variants and a link to the original product. New or renamed nodes are marked after an update, giving returning visitors a quick hint about what changed.

Since February I have already refreshed the graph several times, removed models that no longer made sense and added new tools. I plan to keep updating it regularly. When a better tool becomes part of my workflow, it goes onto the map. When something is no longer useful, it gets darker or disappears.

I do not know if a radial graph is the best way to describe the whole AI ecosystem, but it has been a useful way to show where my own AI stack stands now.

State of the Union is available online, and the source is on GitHub.

The tools will change. The map will change with them.