Free will re-redux

In fairness, it is common in some programming/software contexts to refer to "decisions" made in/by code, and I think the "proof" that Mediochre intended was an implicit one along the lines of:

1. If it is possible to provide an example of code with decision-points in it which in general is reliably executed by computers in this world, then there exists an example of "how" deterministic choices happen.
   
2. Here is an example of code with decision-points in it which in general is reliably executed by computers in this world, thus it is possible to provide such an example.
   
3. Therefore, there exists an example of "how" deterministic choices happen.
   

I think the first premise could be contested along the lines you've argued in this thread, but the implicit argument is at least valid.

And quantum indeterminism is often talked about as particles making "decisions". But I think that's just language being used loosely save for the bottom-up Pansychics.

But programmatic steps don't work with real decisions. Let's say someone is asked, "Ice Cream or Cake?". They think about it for a bit, unsure of what to pick, then decide on Cake.

If it's simply a comparison between Ice Cream preference and Cake preference, why did they deliberate? How exactly does one add force vectors, or some other weighting, to qualitative preferences in order to compare them? Rather once a decision is made it just gets assumed that there was some deterministic explanation -> "You wanted cake more so..."

On the flip side saying it comes down to randomness in some call to an RNG...again why is there deliberation then? And randomness is always just the surrender of explanation, as it forgoes any intelligibility in the world. The fact quantum indeterminism can be modeled stochastically, even used in an RNG, just further shows there's a wide country of possible behaviors between Inexorable Fate and Hyper Chaos.

So my contention is that there's as much if not more mystery in "random" and "deterministic" decisions as there would be when consciousness selects among possibilities.

Now I don't even think it is worth arguing "Do humans have free will?" until a prior (weaker?) question, that of coherence, is answered. So in this thread I think it's better to just try and discuss if any being, including God in Theism or the Ur-Mind in Idealism, can have free will.

And it seems to me that if causation can be grounded in a Ground of Being that is itself conscious you also explain "determinism" and "randomness", so there is at least one possible world where one free being exists.

[Laird]

But programmatic steps don't work with real decisions. Let's say someone is asked, "Ice Cream or Cake?". They think about it for a bit, unsure of what to pick, then decide on Cake.

If it's simply a comparison between Ice Cream preference and Cake preference, why did they deliberate? How exactly does one add force vectors, or some other weighting, to qualitative preferences in order to compare them? Rather once a decision is made it just gets assumed that there was some deterministic explanation -> "You wanted cake more so..."

On the flip side saying it comes down to randomness in some call to an RNG...again why is there deliberation then? And randomness is always just the surrender of explanation, as it forgoes any intelligibility in the world. The fact quantum indeterminism can be modeled stochastically, even used in an RNG, just further shows there's a wide country of possible behaviors between Inexorable Fate and Hyper Chaos.

Hmm. No offence (we're in agreement on most things in this thread), but I'm not sure that that's responsive to the strongest possible framing of the position of our opponent(s).

You present a simple alternative between a deterministic preference comparison and a call to an RNG, but Paul's actually suggesting the possibility of a mixture of determinism and randomness via a complex algorithm running on the neural networks of our brains.

my contention is that there's as much if not more mystery in "random" and "deterministic" decisions as there would be when consciousness selects among possibilities.

I agree with your contention, even if not due to that which preceded it.

Now I don't even think it is worth arguing "Do humans have free will?" until a prior (weaker?) question, that of coherence, is answered. So in this thread I think it's better to just try and discuss if any being, including God in Theism or the Ur-Mind in Idealism, can have free will.

Fair enough. I think the simple argument that I made in post #314 proves this possibility. Do you agree?

And it seems to me that if causation can be grounded in a Ground of Being that is itself conscious you also explain "determinism" and "randomness", so there is at least one possible world where one free being exists.

Mmm. I think I see what you're saying, but the argument could probably do with having its premises and conclusion enumerated for extra clarity.

Hmm. No offence (we're in agreement on most things in this thread), but I'm not sure that that's responsive to the strongest possible framing of the position of our opponent(s).

You present a simple alternative between a deterministic preference comparison and a call to an RNG, but Paul's actually suggesting the possibility of a mixture of determinism and randomness via a complex algorithm running on the neural networks of our brains.

Hmmm....does that change anything? It seems a mixture leaves as little explanation as the individual pieces?

Fair enough. I think the simple argument that I made in post #314 proves this possibility. Do you agree?

I agree. If someone has an actual affirmative case *against* your simple argument, they can tell us why everything is deterministic or why everything is random or...

Mmm. I think I see what you're saying, but the argument could probably do with having its premises and conclusion enumerated for extra clarity.

Oh I agree one would need to show a little more work...something like Aquinas' 5th Way or the Nondualist idea of change as Awareness/Void/Form.

[Laird]

Hmmm....does that change anything? It seems a mixture leaves as little explanation as the individual pieces?

I don't think the possibility of mixture is especially relevant, since true randomness is not as far as I know required (in general) for (otherwise deterministic) neural networks to be fully functional, but the complexity I do think is very relevant.

I agree.

Oh I agree one would need to show a little more work...something like Aquinas' 5th Way or the Nondualist idea of change as Awareness/Void/Form.

Fair enough, although I was thinking more along the lines of making it clear and explicit exactly how "it is (logically) possible that there exists a conscious Ground of Being which explains 'determinism' and 'randomness'" leads to the conclusion that "free will is (logically) possible"... but perhaps that is what you meant anyway by that response, and I simply misunderstood it.

I don't think the possibility of mixture is especially relevant, since true randomness is not as far as I know required (in general) for (otherwise deterministic) neural networks to be fully functional, but the complexity I do think is very relevant.

Can you explain, I really don't see it. Especially since neural networks rarely have algorithmic explanations last I checked.

By which I mean people know they can get the right answer a good percentage of the time but not understand how it works.

Fair enough, although I was thinking more along the lines of making it clear and explicit exactly how "it is (logically) possible that there exists a conscious Ground of Being which explains 'determinism' and 'randomness'" leads to the conclusion that "free will is (logically) possible"... but perhaps that is what you meant anyway by that response, and I simply misunderstood it.

Ah I think if you can show - or rather strongly argue for the possibility - causation requires a grounding in an Intellect/Awareness that suffices to show one possible world where it rationally makes sense there's a being with free will.

But your simple proof might be better, as it cuts through and just requests what is the "gotcha" waiting behind the "how" questioning.

[Laird]

Can you explain, I really don't see it.

Sure. My contention is that the more complex a (deterministic) process such as a neural network is, the more appropriate it is to characterise its evaluations as "deliberation", rather than a (mere) "comparison of preferences".

Especially since neural networks rarely have algorithmic explanations last I checked.

A fair point, but does lack of insight into a process necessarily entail that that process is not deliberative?

[Laird]

Heh. Not sure I really explained anything so much as simply asserted it more clearly, but perhaps my response progresses the exchange in some small way.

Sure. My contention is that the more complex a (deterministic) process such as a neural network is, the more appropriate it is to characterise its evaluations as "deliberation", rather than a (mere) "comparison of preferences".


A fair point, but does lack of insight into a process necessarily entail that that process is not deliberative?

Isn't this suggesting the Hard Problem has a Physicalist if not Computationalist solution?



Maybe so, perhaps deliberation is just waiting for a computationalist solution to finish its load time, but my point was that there is no easily available explanation that should make one think a deterministic "how" is less mysterious than a free will "how".



In fact, IMO, they would both ultimately have to be grounded in some metaphysical picture of causality. And of course "determinism" un-grounded i[s] just randomness of a special kind.

[Laird]

Deliberation feels like trying to grasp the solution, and is distinct from someone saying, "Well I'll sleep on it but for now not think about it."

Also...Isn't this suggesting the Hard Problem has a Physicalist if not Computationalist solution?

Ah, yep... I'm ignoring the hard problem here and simply allowing for the sake of (fairness to Paul's) argument that the neural network of the brain is conscious. Like you though, I don't believe that the hard problem does have a physicalist or computationalist solution.

my point was that there is no easily available explanation that should make one think a deterministic "how" is less mysterious than a free will "how".

And it's a point I've already affirmed.

Ah, yep... I'm ignoring the hard problem here and simply allowing for the sake of (fairness to Paul's) argument that the neural network of the brain is conscious. Like you though, I don't believe that the hard problem does have a physicalist or computationalist solution.

I think a challenge here is there's the process of training a neural network, and just having a single new instance being classified post training.

So you have a set of probabilistic weights in a kind of "curve fitting" situation...but isn't this just a mixture of probability and "determinism"?

If "decision" means the effect following from a set of prior causes [that include qualitative feels of mental state] (liking cake, liking ice cream, wanting to be healthy, needing a pick-me-up, having a sugar addiction...) then the same issues pertaining to cause-effect relations would apply here I think.

Save the problem is the weights are supposedly based on qualitative feels, not quantitative factors. One can claim the former correspond in some way to the latter, but that would imply a known solution to the Hard Problem. And there's still a missing "how" regardless.

Just to clarify it seems to me we have three options [with causation in general] ->

A set of causes leads to one out of a set of possibilities for no reason at all: So this would be randomness, though it gets harder to believe this is just-so randomness if the outcomes have a stochastic distribution that we can model. [Part of] why Thomas Nail argues even non-conscious matter moves in a way that's neither deterministic nor random.

A set of causes leads to one and only one outcome: This is the deterministic option, but note there's nothing actually distinguishing this from the first option save the probability is all concentrated in a singular option we assume to occur 100% of the time. So randomness of a special kind.

A conscious agent can pick the effect, to a limited degree, from a set of causes: The question of "Why this outcome, rather than another outcome?" is in our own experience answered by conscious agents making choices. Just the start of an argument, but it's the only option that at least grounds causality in the sense of a reason why of all the effects that can happen only one does. Our limits are due to restrictions of Form, whereas a hypothetical God-as-Ground's limits only be that which is logical [because the Ground is Unconditioned].