Suppose you program a simple robot (say, like a Neato) to try to move around while vacuuming and avoiding obstacles. It follows a purely deterministic program. Now suppose you have another robot controlled purely with a quantum-level source of randomness (say, it turns to the left each time a particle decays). Would the second robot really be more free? I think not. Intuitively, what we call "free will" seems proportional to the ability of an entity to interact with its environment in a complex way.
