plain_text for active_inference: Modeling, Learning, and Exploration

All right, we're live. Welcome everyone. It's February 7th, 2025. And we are in ActiveInfra Ant Stream 11. Looking forward to a lot of feedback, interactivity, and plain text fun and learning. All the pushes during the stream are going to be going to the repo doxology slash cognitive. And let's get right into it. All right, so previously on ActiveInfra Ant Streams, we've heavily featured the cursor development environment tool. And that's going to continue to be used as well as seen a little bit with CodeViz. And that's going to come into play again. And in this stream, for plain text, surprise, education, and delight, we're going to be working a lot with Obsidian. So Obsidian at the repo or website obsidian.md is a plain text note linking software, open source. And it is going to facilitate us doing some truly awesome things with plain text repos. And playing back and forth between Cursor and Obsidian is going to go really far. So here in the Obsidian window, on the left side, we're looking at the same file that on GitHub can be found here. And in Cursor can be found here. So we have two local versions of Infra Ant Stream 10.md. So if we delete something, like starting with a GitHub push, save it in Cursor, that's going to be gone in Obsidian. Obsidian uses a linking style where two brackets open up the ability to make a linker. So that is going to be reflected visually in this graph version. So before we go too deep in, this is some incipient, exciting, relevant development work. It's semi-structured, so don't let the graph and the weave make too much light of the fact that if we zoom out, of course there's more to paint into than not. And looking forward to people's collaboration and contributions. You can head over to the repo while the stream's happening or afterwards. Or if you're watching live, definitely put some funny questions and ideas into the live chat. And we'll bring them into the repo and that will weave that way. Big picture, we're going to be in some ways that were not really even possible a few years ago, integrating theoretical foundations, writing them on the fly, and connecting with practical implementations. So we can delete that and then we can see, okay, now we just updated a plain text file. So let's just make that push just for fun. Okay, where do you even begin? So welcome an overview of Obsidian usage. So this could have been typed in manually like this. Two brackets, Obsidian usage. When we click on that link, it's like Wikipedia. We head over to this article. And there's going to be links in the article. The ones that are brightened are the ones where there is a page. And the ones that are dim, there's no page yet. And so if we click on that, it makes that page at the top level of the repo. So we don't need that at this point. But let's look a little bit at Obsidian usage. And this is a markdown format. Everything is these plain text markdown files. Okay, so file names link directly. File name, vertical line, alias uses custom text. If you want it to say something different than the file name. And here we have some machine readable information. And then over on the graph side, we can filter down and try to see what's in the neighborhood of Obsidian usage. There's going to be a lot of active inference generative models in this stream as well. But first, I just wanted to start this way. So here we are. And when we have our cursor there, we get this colorful highlight of Obsidian usage. And everything that it's linked to in the plain text gets rendered as an edge here. So we can drag around that, play with that. Okay, Obsidian usage, plain text benefits. Let's check out plain text benefits. Wow. Version control, knowledge management, machine readability, research benefits. All these different benefits. Which cursor wrote entirely for me? And we'll get to how and where that happened. Okay. Back to the top level and let's carry on with the stream. So, preview streams, we've been using different…