But in this case, the unpredictability is a measurement limitation. i.e. It is unpredictable in practice - but not in principle. Taking the thought experiment to the extreme, if we simulated two identical minds, with identical memories in a matrix-like environment, they would both make the identical decisions when confronted with identical stimuli. How then, can they have free will, if we can use the first one to predict perfectly what the second one will do? In the real (yet farflung future) world, Syne has a valid point. We could do the same experiment by duplicating a real human, in a real, closed environment (to keep external stimuli identical). How could the two humans do anything but make identical decisions?