So here’s the trick: first, the system gets a random bit from some other source. Then it adds the bit from the other source with the bit from the quantum bitstream and uses the result to decide if you move left or right. Now, when you want to continually move left, half the time you’ll have to make the quantum bitstream return zeroes and the other half the time ones — exactly what the laws of quantum mechanics require.
With a little additional mathematical complexity, the scenario is generalizable to much more complicated quantum functions and human results. But the basic principle is the same: one can use quantum randomness to exercise free will without violating any statistical laws.
Of course, this still leaves one key problem. What is picking the results of this quantum bitstream? And how does it do it? I have to admit I cannot really think of a sensible way. But this seems like a problem for neurobiology to figure out and report back to us. I merely aim to prove that its doing so is consistent with what we know about the laws of physics.