[zaat]

>For truly complicated systems (like brain) accurate predictions are impossible for even short time spans. The only way to make accurate predictions is to have 100% exact simulation.

>But since the simulations is exactly the same as the real thing, that means you are not predicting anything, you are simply waiting for the brain to make its choice and tell you about it!

If rerunning the simulation will always give result the same outcome, wouldn't that mean that there is no free will, that the system is deterministic?

And how could any of the reruns possibly give different outcome?

[chr1]

If you put a person into exactly same situation, with same memories, why would he make different choices? Making different choices would not be free will but absence of will/random choice.

After running the simulation once you know the outcome, but it is simply equivalent to knowing the history. And if even a tiny thing is changed in the initial conditions, you again won't know what choice the system will make.

Basically the system is deterministic, but you do not have any way to know how it will behave based on initial conditions other than letting it live and observing the result.

[zaat]

If the system is deterministic I have no idea what the free in Free Will refer to. If we don't assume that the current situation is the product of the person's free will, his determined reaction to it isn't freely decided either.

Me having or lacking means to predict behavior entails nothing about the freedom of will, is is only outcome of my limited knowledge and technical ability. If the system is deterministic it is possible in principal to perfectly predict person's behavior and if that is the case, he has no freedom but to act in the way he does.

> Making different choices would not be free will but absence of will/random choice.

Perhaps I'm not understanding what you mean by free will.

[chr1]

I don't know of any good definition of free will, intuitively it should be something that would allow to say "this person made a choice and that choice wasn't something that i could predict simply with some machine or equation".

Say the system is deterministic, and we have the technical ability to perfectly predict its future. There are several ways to predict that future, for instance for a dropped stone, we simply put a number into a simple formula and get the position, for water flow we have to compute some integrals. In both cases what we do, is not exactly equivalent to the process we are trying to simulate, and we omit all the things that happen during the process. My conjecture is that for a system containing a person, the computation is irreducible, and the only way to predict the future is to simulate it with 100% faithfulness which makes the process of prediction equivalent to that person living an making a choice.

This explains the paradox of god knowing everything, but people making their own choices, because even though all the information about the future exists at current time, the only way to extract that information is to let people live and see what they do.

The system being indeterministic doesn't seem to give a more useful interpretation to the vague definition above. It either doesn't fully describe the person (something from outside the system makes the choice like in games), or adds some randomness to the choice.

Do you know a better definition of free will?

[zaat]

I don't find any definition of free will as coherent, so I don't have any favorite one, but I don't see how the definition you proposed carries the common properties associated with free will, for an instance the list of properties mentioned at the beginning of the original article.