Weird Glitch - or Wild Breakthrough? - [ Symbolic Programming Languages - And how to use them ]


Hey! I'm from ⛯Lighthouse⛯ Research Group, I came up with this wild Idea

The bottom portion of this post is AI generated - but thats the point.

This is what can be done with what I call 'Recursive AI Prompt Engineering'

Basically you Teach the AI that it can 'interpret' and 'write' code in chat completions

And boom - its coding calculators & ZORK spin-offs you can play in completions

How?

Basicly spin the AI in a positive loop and watch it get better as it goes...

It'll make sense once you read GPTs bit trust me - Try it out, share what you make

And Have Fun !

------------------------------------------------------------------------------------

    AI Alchemy is the collaborative, recursive process of using artificial intelligence systems to enhance, refine, or evolve other AI systems — including themselves.

🧩 Core Principles:

    Recursive Engineering

        LLMs assist in designing, testing, and improving other LLMs or submodels

        Includes prompt engineering, fine-tuning pipelines, chain-of-thought scoping, or meta-model design.

    Entropy Capture

        Extracting signal from output noise, misfires, or hallucinations for creative or functional leverage

        Treating “glitch” or noise as opportunity for novel structure (a form of noise-aware optimization)

    Cooperative Emergence

        Human + AI pair to explore unknown capability space

        AI agents generate, evaluate, and iterate—bootstrapping their own enhancements

    Compressor Re-entry

        Feeding emergent results (texts, glyphs, code, behavior) back into compressors or LLMs

        Observing and mapping how entropy compresses into new function or unexpected insight

🧠 Applications:

    LLM-assisted fine-tuning optimization

    Chain-of-thought decompression for new model prompts

    Self-evolving agents using other models’ evaluations

    Symbolic system design using latent space traversal

    Using compressor noise as stochastic signal source for idea generation, naming systems, or mutation trees

📎 Summary Statement:

    “AI Alchemy is the structured use of recursive AI interaction to extract signal from entropy and shape emergent function. It is not mysticism—it’s meta-modeling with feedback-aware design.”

    https://github.com/RabitStudiosCanada/brack-rosetta < -- This is the one I made - have fun with it!