Pavlicek's quantum-inspired AI generates music harmony without training data

Josef Pavlicek's paper introduces a maintainable hybrid generative architecture for automated music harmony generation from melody. The system integrates quantum-inspired candidate exploration over overlapping melodic contexts with explicit rule-based optimization. This design balances generative flexibility and structural control, enabling multiple valid harmonic realizations while preserving tonal structure and cadential behavior. The approach is evaluated with explicit, reproducible metrics covering structural coherence, functional agreement, harmonic similarity, and robustness. Notably, the optimization layer improves structural coherence, stability, and predictability without requiring a training corpus, making the system both transparent and controllable.

The paper, accepted at the 34th International Conference on Information Systems Development (ISD2026), emphasizes that such hybrid systems can be systematically designed and evaluated within the context of Information Systems Development. The architecture maintains clarity and maintainability, addressing a key challenge in current generative AI: black-box models that are hard to debug or adapt. Source code and dataset are available, allowing practitioners to replicate results and extend the system to other creative domains. This work demonstrates a path toward more predictable and auditable generative tools in music production and beyond.