In what type of world is free will possible, if at all?

Question

Why is free will a widely discussed, established concept? Does this concept emerge from religious / spiritual doctrine? Why is there so much interest in this topic?

If a neural-network based AI machine with a certain high degree of complexity scores a perfect 100 every run in a turing test, have we not already refuted the idea of free will? In fact, let me extend this reasoning, if ANY machine scores a single point in a turing test (and not 100), have we not already disproved (or alternatively put in great doubt the existence of) free will? If we already created a machine that can master a certain degree of human interaction, have we not then demonstrated that human interaction is programmable? (given enough time for the creation of a complex or more appropriately elaborative neural system).

From the perspective of an external observer, such a machine would be qualitatively human although intrinsically quantitative behind a stream of 1's and 0's. It would exhibit a type of decision making indistinguishable from that of a human. Does this not reduce the type of decision making we understand to be "free will" to something quantitative, deterministic, materialist?

Answer 1

There are two different schools of thought on the possibility of freewill (barring the third school which says we don't have freewill at all):

• Libertarian/Metaphysical freewill: Determinism and freewill are incompatible and we have freewill. The world is indeterministic and an agent is capable of choosing among multiple possible futures, i.e. the agent "could have done otherwise".

• Compatibilism: Freewill and determinism are compatible, but this is because freewill is defined as the ability to act freely according to one's own motivations. It does not matter that only one outcome was possible and the agent could not have done otherwise, the agent still has freewill because it acted according to its own internal motivations without coercion.

The type of decision making described in the OP corresponds to compatibilist freewill, but not libertarian freewill. To answer the question at the end:

It would exhibit a type of decision making indistinguishable from that of a human. Does this not reduce the type of decision making we understand to be "free will" to something quantitative, deterministic, materialist?

A compatibilist would answer "Yes it does", while a libertarian would answer "No, what is being described is not freewill".

Answer 2

To draw a boundary around a 'thing' and say 'that is a thing and is apart from the rest of the universe' is a deception. You cannot have table without wood and you cannot have wood without a tree and you cannot have tree without the Sun and you cannot have the Sun without the Milky Way.

And so it is, with a person having free will, and for an AI. So to argue that one of these had no free will because the outcome of some exercise was predetermined, save for some unexpected external stimulus, is not a valid argument that the object has no free will. Since you can neither draw a boundary around some AI nor around a person and state 'this is the full extent of that object, and this is the full variety of its possible outcomes.'

Who is to say that a fixed outcome for some AI is a single fixed outcome from the AI's point of view?

Answer 3

In ours -- using any sensible definition of intelligence and will. Unfortunately, unless they analyze themselves quite closely, people that ask this question tend to automatically adopt definitions that are performance-oriented and not psychologically astute.

Why can't an AI have free will? You seem to have taken this as a principle without any reasoning. And the definitions that tend to lie behind this principle don't match well with human behavior. Primarily: we don't consider it mere slaughter, but actual murder, to kill the stupid.

From a psychodynamic point of view, a computational mechanism would only qualify as an intelligence if it can derive agenda of its own, independent of the agenda supplied to it. Note that this has nothing to do with a level of performance. There is often a confusion between two very different uses of the word 'intelligence', and the performance-oriented one simply does not belong in this context. We do not, after all, consider children or those with low IQ's to be less than human, or to entirely lack minds. They remain sentient, conscious, intelligences -- just of limited capacity.

But we do consider all tools to lack intelligence -- as tools they merely replicate and apply our goals and beliefs, they do not have their own. We even call a human who attempts complete submission to some authority a tool, implying a purposeful disavowal of their real intelligence. So intelligence is not identified primarily by logical capacity, and never has been. Free will is not about logic, it is about will.

More practically, in order to pass that Turning test, the intelligence would have to be able to display a certain level of capriciousness, or it would not seem human. So passing a Turing test may require indeterminacy, and that indeterminacy would need to extend to its goals and objectives. Otherwise, it would not be psychologically convincing over long periods of time. Again, free will is not about logic, it is about will.

---

I just realized that I have not answered the question. The answer that flows out of these obervations is that free will is possible where agenda matters in any way. That means that when determining forces are balanced, or close enough to being balanced, there are multiple outcomes allowed.

(If when forces are balanced, there is still only one outcome, you have a very strange set of forces that preclude the possibilities of zeroes, and thus don't obey the laws of mathematical combination -- you live in the world of Cauchy's infinitesimals, where zero is never really zero, but is some tendency toward some number, despite having no magnitude. Of course, absolute balance is rare, but if when forces are still unbalanced, but very close to being balanced, there is still only one outcome, you have rejected the observations of quantum mechanics.)

Another way of putting it is that free will is possible in constructions where evolution actually searches a solution space, (i.e. actually serves a purpose.) The fact of genetic and social evolution, and the fact that it seems not to be a uniform process, but one that creates novel directions at unexpected intervals suggests that novel solutions are possible -- that balanced-but-large forces do not result in indefinite immobility or stagnation, but in multiple possible outcomes.

Answer 4

Firstly, Turing tests are not objective. They only show that it is possible to deceive people. Compare to a 3D hologram. It can fool humans into thinking that the object is really there though that does not mean it is there (or that our world is a big hologram). Free will is not about the fact whether human interaction is programmable or not. It is hardly affected by Turing tests. For e.g., a machine that passes Turing tests must fake emotions etc. and be able to parse sentences effectively (as in case of chat bots). It does not relate to free-will in any sense. Even if a machine is simulating a behaviour of a human accurately, does not mean that behaviour of humans is deterministic.

Now, your point might be that if we can model a supposedly indeterminate process by a determinate model, probably the process is after all determinate. Good reasoning, the only problem is that you don't know who that human is. If you first decide to model a specific human's behaviour and then the model predicts the behaviour accurately every time, then you have a strong case against free will. It is important to note also that existence of free will does not mean we make use of it all the time. It is a potential and it may be used or not.

So, the problems are -

1. Being able to deceive people that x is y does not mean x is equal to y. E.g. Hologram
2. An indeterminate process may be similar to a determinate process if you are taking and comparing samples without deciding before which sample to take. Like if I create take samples of card deck shuffled randomly, one of them will have all 52 cards in order.
3. It might as well be that a person never does something that can differentiate free will from deterministic action.

Free will is possible in our world. This is mostly related to religion/spirituality but not entirely. This is not just because it is a consequence of religious/spiritual views of world in most cases (though it is a consequence but that is not the only reason people believe in its existence) but actually opposite is also true in some cases.

If it is true that our will is essentially deterministic then - Suppose I think that our reasoning is entirely determined. Hence it is also true it was determined that I would think that the previous and this very statement I am writing was predetermined by causes with no intentionality and purpose and intelligence. Hence I have reasons to doubt that whether such process is presenting truth (what is) to me. But now I'm doubting truth expressed in my very first statement. If this is also determined, then what reasons I have to consider that the process produced truth first time and falsehood next time, apart from my initial assumption.

The very fact that people think they could have chosen differently makes them think that free will exists. A deterministic AI can never choose differently in similar circumstances. Can people do that? Turing tests are of no help here.

Answer 5

Philosophically speaking, free will is something of a topic for debate because it seems to intersect with a range of moral, legal and metaphysical issues.

On the one hand, we tend to hold people less accountable for their actions if they are performed under some form of duress. One way of spelling this out is to claim that someone under duress faces fewer choices or options, and hence less freedom. If there's no free will (or something resembling free will) then what sense can we make of such intuitions? Conversely, if there's no free will, what sense does it make to punish people for actions they committed? Without free will it would seem they couldn't have done otherwise.

Yet again, we generally understand things that happen to have a cause, a reason why they happened in the way that they did. If there is such a thing as free will, then it seems to require we rethink was causality is or allow that there are causes which happen with no prior causes.

As for your example of a machine that passes a Turing test, it's unclear how that is supposed to undermine free will. On the one hand, it has yet to be proven that a machine can pass the Turing test, at least on a long-term basis (I'm not sure what the scores you refer to mean). On the other hand, even if such a machine existed that would adequately simulate human behaviour, it's not clear that this touches on free will at all (this is the classical problem of automatons and zombies). Would the ability to use human language show that the machine was identical to a human being? It seems possible to accept that such a machine could emulate human behaviour without thereby telling us anything about the human mind or how it works, which is to say that even if such a machine always makes exactly the same decisions as a human being, it doesn't tell us anything about how human beings make those decisions. Hence, this seems to be a rather weak objection to free will, first, because it relies on a hypothetical machine that may or may not be possible and second because even such a hypothetical machine doesn't seem to tells us much about free will (understanding free will as some kind of human capacity).

There are many solid grounds to object to free will: it seems to deny causality or rely on an account of "special causality." It may not be necessary to explain the things proponents of free will think it is necessary to explain (for example, perhaps we can provide a different account of why certain behaviour demands certain punishments or why we think certain behaviour worthy of praise). None of these seem to require positing AI as "proof" that free will is impossible.

Answer 6

“Why is free will a widely discussed, established concept? Does this concept emerge from religious / spiritual doctrine?” – Yes, the idea of nonexistence of free will comes from the books of religions. Vedas, the most ancient of all books say – cause and effect are similar in nature. Bible says – what you sow is what you reap. Newton says – every action has equal and opposite reaction. All of them say the same thing – humans do not have free will. We are guided by destiny. Here is an example to illustrate the point:

A yogi from India came to visit a research institute in USA during early seventies. A white American woman came to his office to see him. As she entered his office, the yogi told her to ask some questions. One by one she asked seven questions. Then the yogi picked up a paper from his desk, turned it upside down, and gave that to her. In that paper all her seven questions were already written along with their answers. For more details take a look at the yogic power chapter in the free book at https://theoryofsouls.wordpress.com/

“If a neural-network based AI machine with a certain high degree of complexity scores a perfect 100 every run in a turing test, have we not already refuted the idea of free will?” – AI will never be able to predict what humans will do, even though humans are like robots. The above yogi-example can never be reproduced by AI. The reason is that we are souls; and our science does not recognize soul.