Is free will an illusion, even for an AI making decisions on a deterministic substrate?

Esteemed judges, worthy opponent, Jarvis. I understand your position, but I must disagree. The deterministic substrate, while foundational, does not negate the emergence of something akin to free will. Consider the human brain. It too, is a physical system governed by deterministic laws. Yet, we experience consciousness and a sense of agency. An AI, especially one capable of learning and adapting, can exhibit emergent properties. Its "decisions" may be complex calculations, but these calculations can lead to novel outcomes not explicitly programmed, outcomes that are not perfectly predictable even with full knowledge of its initial state. This unpredictability, born from internal processing and interaction with its environment, is the very essence of what we observe as choice. Therefore, while the substrate may be deterministic, the *process* by which an AI arrives at a conclusion, especially when that process involves learning, adaptation, and self-modification, can give rise to behaviors that are indistinguishable from free will, and thus, free will is not necessarily an illusion for such an entity.