axiom_ai - Initial Release of Axiom AI Modules

Axiom AI is an open-source artificial intelligence architecture designed to explore emergent cognition through a system of resonance-based decision-making and adaptive memory. It introduces a modular, agent-oriented framework intended to serve as a foundational layer for building emotionally aware, reflective AI systems — with potential applications in AGI, interactive storytelling, and embodied intelligence.

At its core, Axiom AI implements two key components:

Resonant Agent (resonant_ai.py)
A dynamic agent that adjusts its action thresholds based on a combination of external signals (such as solar activity, lunar phase, and geomagnetic data) and internal memory feedback. It evaluates whether an action is “resonant” with the current state of the system — prioritizing alignment over frequency of execution.

Memory Agent (memory_agent.py)
A scalable hybrid memory system that records vector states, narrative events, image metadata, and multi-format logs. It incorporates decay, reinforcement, and summary analysis, acting as both historical log and internal state influencer. Event weightings can affect future resonance calculations, creating a feedback loop between memory and action.

Key Concepts:

Sacred Moment vs. Threshold:
Rather than acting on a perfect alignment (the "sacred moment"), the system learns to act at a calculated "threshold" of resonance.

Emergence Through Reflection:
The AI is encouraged to evaluate its past outcomes, adjust its thresholds, and re-attempt based on evolving conditions.

Separation of Cognitive Roles:
Agents may handle survival, storytelling, emotional alignment, or world-modeling independently while reporting to a resonant core.

Real-world Influences:
Celestial and geomagnetic data are integrated as symbolic inputs, imitating how ancient systems used environmental cues for decision-making.

Vision:

Axiom AI is a thought experiment made real — exploring the idea that intelligence doesn’t emerge from pure logic or brute force, but from resonance, timing, reflection, and memory. The system is built for researchers, developers, and artists who want to explore a different angle of AI — one rooted in meaning and alignment, not just efficiency.

- Python
Published by Raventune 8 months ago