Hey everyone, long time no see.
Some of you probably noticed I’ve been pretty quiet for a while. I haven’t really been posting new models, releases, or much AI content lately, so I wanted to explain why.
I didn’t disappear because I lost interest in AI.
If anything, I went deeper into it.
Over time, I started feeling that a lot of AI today is very strong at generating output, but still weak when it comes to deeper structure: how it observes, how it decides, how it tracks state, and how it acts in a more controlled and understandable way.
So instead of focusing only on models and outputs, I’ve been spending a lot of time working on something more foundational behind the scenes.
It’s called OCP.
The simplest way to explain it is this: I’m exploring a deeper runtime/system layer for AI and software that is built around a clearer flow of observation, decision, and controlled action.
Not just “input in, output out.”
More like: observe what is happening, match the right path, then commit actions in a more explicit and governed way.
So in a way, I’ve been shifting from mostly working around AI content and generation toward building part of the deeper system layer I think future AI will need.
This is obviously a lot less flashy than posting images, models, or demos people can instantly feel.
But to me, it matters a lot, because I think this layer is part of what separates impressive generation from something more coherent, reliable, and structured.
I know this is different from what most people here usually follow me for, but I wanted to share honestly what I’ve been building and why I’ve been away.
If you’re curious, the project is here:
GitHub: https://github.com/DucHaiten/OCP
Even if you’re not a developer, I still appreciate anyone who wants to take a look, just to see the direction I’ve been moving in.
This is a different phase for me, but it still comes from the same bigger interest: pushing AI toward something deeper than just generation.