Free will as structured unpredictability: toward a symbiotic human–AI relationship

Human history has been shaped by revolutions of varying pace, with artificial intelligence (AI) emerging in mere decades. This paper introduces a preliminary framework for fostering a symbiotic human-AI relationship by integrating human free will into AI systems. We conceptualize free will as 'structured unpredictability' and propose a speculative extension of Shannon's information theory to model its informational contributions. By framing free will as an informational surplus, we envision AI as a mirror and amplifier of human creativity. While theoretical, this framework lays the foundation for future empirical and computational research to preserve human autonomy and diversity in AI-driven systems.