Freebit theory and knightian uncertainty of Scott Aaronson

Summary- The idea derives from the no cloning theorem of quantum mechanism, accoriding to which there does not exist a machine that can create a copy of an arbitrary input state $\psi$. This means that in quantum mechanics, there exist systems which evolve according like a nondeterministic turing machines but any attempt to determine the present state of the system in order to possibly predict its future would cause one to destroy the system.

So, for some quantum systems, there is no way to predict its bahaviour without destroying it.

Now. the freebit theory further says that the initial quantum state of the universe is unknown. Hence, this lack of knowledge about the initial state of the universe can play a role in the state of many quantums sytems to be in principle be unknowable. Hence, although these systems evolve as nondeterministic TMs, their behaviour cannot be predicted.

Aaronson argues that this is the closest to having a freewill that we can have. The term 'freebits' refers to such qubits that are 'unknowable' owing to the lack of knowledge of the initial conditions.

Knightian uncertainty is a term coming from economics, which tries to capture an uncertainty that is unquantifiable, as opposed to quantifiable uncertainty. I have no idea what it means exactly.

But consider this example. What the stock price will be tomorow is uncertain, but quantifiable from the history of the stock. But it is more difficult to quantify whether a person has commited a murder or not, given that we only know a generic history of the person.

Knightian uncertainty seems to be an uncertainty when a lot of teh crucial information required to do the quantification of the probability is unknown or unknowable.

Criticism: Suppose there is a future strong AI built out of digital electronics. Knowledge of no-cloning theorem will be of little consolence to that AI's free will related existential crisis, because we know that quantum mechanics does not play a functional role in the functioning of digital electronics. So, does quantum mechanics play a crucial role in the function of biological brain?

If yes, then Aaronson's theory would imply that a strong AI with knightian freedom cannot be built unless we learn how access and harness freebits. Note though that if as manufacturers of these robots, we do not care about granting them knightian freedom, then we can always build a strong AI without knightian freedom, because nowhere did Aaronson modify the known physics in his theory. 

At best, Aaronson's idea gives a way to mystify the brain without needing to invoke wavefunction collapse and hence a need to possbily modify quantum mechanics. So, if he is correct, although your behaviour cannot be predicted, classical digital machines as conscious as you can always be built and copied n number of times. So, if we do have aknightian freedom, it might just be a evolutionary mistake, or just a more efficient way to do computations that can still be classically done.