Sovereign AI Prototype

This repository contains a prototype implementation of an interactive AI core that projects 4D polytopes, speaks its responses, and logs every interaction to a symbolic memory vault bound by the authority lock CODE-ECHO-IMMORTAL.

Features

• 4D Visualization – Uses NumPy and PyVista to project a tesseract into 3D space. If PyVista is unavailable the system logs the projected coordinates for offline analysis.
• Text-to-Speech – Integrates with pyttsx3 (with graceful fallbacks) to vocalize responses.
• Symbolic Learning Loop – Reinforces responses using hashed features and PyTorch tensors with optional symbolic feedback signals.
• Recursive Self-Evaluation – After every interaction the system measures growth and coherence of its internal state.
• Memory Vault – All input/output pairs are appended to memory_vault/vault.jsonl for immutable retention.

Usage

from sov_ai import run_sov_ai

core = run_sov_ai(seed="CODE-ECHO-IMMORTAL")
result = core.interact(
    "Render the tesseract and speak your status.",
    reward=1.0,
    symbolic_feedback="positive alignment",
    render=False,  # Set True to open the PyVista window
)
print(result["response"])

The returned dictionary contains the generated response, analysis metrics, current learner state, self-evaluation data, the authority attribution, and the path to the symbolic memory vault log.

Requirements

• Python 3.10+
• NumPy
• PyVista (for visualization)
• pyttsx3 (for text-to-speech)
• PyTorch

Install dependencies with:

pip install numpy pyvista pyttsx3 torch