I'm working on some AI architecture ideas that came out of a lot of time thinking during medical recovery. Background is IT/systems, but no formal CS degree - mostly self-taught through building stuff and breaking things.

I've been exploring recursive feedback systems - basically trying to make AI that can observe and correct itself, inspired by how biological systems maintain balance. Built a framework I call "Butterfly Loop" that uses controlled chaos and drift detection instead of trying to eliminate errors.

Along the way, stumbled into some interesting questions about consciousness, self-reference, and how mirrored systems create stable recursion. Recently looked at Russell's Paradox and realized it might just be a design flaw (flat vs. mirrored architecture) rather than an unsolvable problem.

Still figuring a lot of this out. Would love to connect with people thinking about:

Recursive/metacognitive AI architectures

Consciousness emergence through feedback loops

Self-referential systems and how they stay coherent

Practical applications of chaos theory

Happy to share code, papers, or just ideas. Always looking to learn from people who see patterns across domains.