A cognitive algorithm for “free will.”
So here’s your homework problem: What kind of cognitive algorithm, as felt from the inside, would generate the observed debate about “free will”? - Dissolving the Question
Edit: In the linked post by Eliezer, he shifts the debate around free will and determinism. Instead of asking which is true, or trying to resolve or dissolve the philosophical question, he asks a different but related question about human psychology. Why does it feel like it makes sense to have this philosophical debate in the first place? What is it about our psychology that makes positions vis-a-vis free will and determinism seem so compelling? It is this question that I have attempted to answer here.
Put a glass of water in front of yourself. Now, make a decision: take a drink, or do not take a drink. Or at least imagine doing so.
When you’ve made your decision and acted on it, get very clear on which option you chose. Perhaps you should write it down, or at least make a mental note. Do not proceed until you’ve come to this point.
Now, you have a memory, one in which you seem to have made a decision of your own free will. Why does it feel that way? And why might somebody else argue that, no, in fact, you were fated to make that decision?
Let’s consider the one or two main choices you made for this exercise. The first choice was whether or not to comply with the instructions. If you did, the second choice was whether or not to take a drink. Of course, there were other choices: how exactly to carry this out (in your imagination or in real life? with water, or coffee?). But we’ll ignore them for now.
The process of making these major decisions took place slowly enough that you have some felt memory of the psychological operations you experienced in the course of making them. You read the instructions. Then, you may have split the future into two imagined alternatives. Some mysterious further operations that you’d have a very hard time putting into words took place. Your subconscious mind was playing a role too, but that’s beyond your comprehension. But you clearly remember taking the physical actions to bring about one world state or the other.
Some aspects of this process of thought and action took place slowly and consciously enough to leave an accessible memory. Others were too fast, or two unconscious, to leave a trace.
The words of the instructions, the sense of splitting the world into two options (or not doing so), and the physical action (or lack thereof) all happened slowly enough to create a memory.
When you access that memory and reconstruct for yourself your past experience, you imagine yourself conceiving of two options—to comply with or to bypass the instructions, to drink or not to drink—and a sensation of selecting, then acting upon one of the options.
What you do not remember, because it happened too quickly or unconsciously, was the process by which that selection occurred. Or the operations underneath that. There is a physical sensation of volition that some part of your brain produces, which is perhaps a purely somatosensory feeling. Moving your arm (or not) when no external force is yanking or constraining your body physically feels distinct from being acted upon by an external force. That somatosensory sensation is the feeling of volition that you find associated with this decision in your memory.
The reason why it feels that you have made this decision of your own “free will” is that you cannot detect the overwhelming evidence against it. The mysterious, rapid, unconscious core brain activity that structures your conscious qualitative experience is unavailable to you. You have learned by imitation to describe certain somatosensory sensations in terms like will and volition. The way the instructions were written demands that you describe this reconstruction of memory in terms of “choice.”
The counterargument of determinism arises because we can observe ourselves sometimes experiencing a lack of the familiar association between the choice-making memory and the somatosensation of volition. At times, you remember making a conscious choice, but did not physically feel like you made a decision. At other times, you feel a sensation like volition, but do not remember making a conscious choice. These disturbing disconnects provide some of the missing evidence that an unconscious process is driving your psychological operations, your motor impulses, your somatosensations, and the formation of memory.
Furthermore, you realize that your ability to give an account of precisely which psychological operations led to a decision rapidly bottom out. For any given decision, after saying, “and why did you think that?” a finite (and probably small) number of times, you lose the ability to answer. Yet you infer that there must be a reason why. And you realize that reason is unavailable to conscious inspection, yet must have originated in your brain.
To really solve this homework question, you’d probably also need to explain the emotional valence of the debate, why people speculate on their own psychology in general, and why it is part of our culture. But I think this is a decent account of the mental mechanisms that produce an intuitive notion of free will, and also some evidence for an argument against it.
I haven’t thought it through very carefully, but my hunch is that you’re not quite getting the crux of the issue.
For one thing, there’s a little compatibilism joke I like:
Hahaha. The moral, I think, is that there are two stories:
“the chess engine moved the knight because that was the result of a deterministic process involving a bunch of interacting transistors on a chip”
“the chess engine moved the knight because it evaluated 400,000 move sequences and found that moving the knight was best”
These sound like different stories, but they are in fact the same story told at two different levels of abstraction.
So that’s one hint: we can get tripped up by failing to appreciate that stories at multiple levels of abstraction can all be “true” simultaneously.
Then another related hint is a thing I wrote here about the hard problem of consciousness:
So I think it’s a slighty wrong starting point to think along the lines of “the processes in our brains to which we have conscious access are a subset of all the processes in our brains”. Instead I think we should be saying, “when we introspect, we’re not directly seeing any of the processes in our brains, instead the things we perceive are certain states of certain models that abstract over some of the processes in our brains”—just as “a dog barking” is an abstract model that represents some complicated statistical predictions about incoming photons and sound waves and so on.
I don’t really have a conclusion here, and I don’t mean for this to be a strong disagreement with what you wrote, just maybe a tweak in the framing or something. :-)
If you read the original post I quoted and linked at the top, Eliezer is interested in explaining why the idea of a debate between “free will” and “determinism” arises in the first place, on the level of cognitive processes. I can see now that I should have explained this in more detail!
So even if compatibilism is true, why, on a cognitive gears-level mechanisms level, does it feel to many people like it makes sense to argue about whether or not we have free will?
Yes I understood that. I think my comment is a relevant thing to keep in mind when thinking about that question.
Like, we can only perceive our introspective world through the lens of abstract models of that world that our brains build. So we should be thinking: What do those models look like?
There’s a thing that Dan Dennett calls (I think) “sophisticated naive physics”, where you take “intuitive physics” as an object of study, determine its ontology, rules, etc. We understand that the ontology and relationships and affordances of “intuitive physics” may be quite different from anything in actual physics, but we can still study it. By the same token, I’m proposing that the question of “why do people have the intuitions they have about free will” should be studied through a lens of “sophisticated naive introspection”, i.e. determine the ontology and relationships and affordances of the abstract model space that we use when we introspect, while accepting that those things may be quite different from anything actually happening in the brain. Whereas your OP seems to take a different perspective, namely that introspection provides an accurate view of a subset of the things in the brain.
I do like your framing of “sophisticated naive introspection,” and I think it makes sense as an object of study. That seems like a nice generalization of the program of study that Eliezer is calling for in his post.
However, it seems to me that your distinction between an “abstract model of the world that our brains build” vs. an “accurate view,” or the idea that there’s a categorical difference between introspection and what we might call “actual psychology,” seems wrong to me.
Everything we use to study the brain, including naive introspection, is an abstract model of the world. Psychology refines, and sometimes directly contradicts the findings of naive introspection; the same is true of the relationship between naive and scientific physics. However, in both cases, intuition and reality line up pretty nicely, at least for some subset of situations. For that limited set of use-cases, the difference between naive and scientific approaches to the discipline isn’t worth worrying about in practical terms.
One of the unstated assumptions of my argument here is that the abstract psychological model afforded by naive introspection is, in this case, adequate to at least make some headway on this problem. It wouldn’t be nearly adequate to resolve it—I haven’t even really scratched the surface. But that wasn’t my goal here. I just wanted to give the best quick account I could of why “free will” and “determinism” seem to make psychological sense, based on my own intuitions and thoughts and given a short investment of time.
Fair enough :)
That is, unavailable to my conscious mind.
What you say is similar to Sam Harris’s arguments. Both have two principle problems: that the argument depends on an implicit definition of free will, that it is a conscious choice; and an implicit claim about selfhood, that the conscious mind is the self,and the unconscious mind is not.
The compatibilist definition of free will is concerned with whether you are able to act on your desires … not on whether your desires are conscious or unconscious.
The libertarian definition of free will is concerned with whether your decisions are free of external determimation, or unpredictable in principle… not on whether they are conscious or unconscious. If the unconscious processes that lead to decisions, and about which you know nothing, are indeterministic, then the case for libertarian free will is strengthened, not weakened.
You don’t observe them, so you don’t observe them as deterministic. And in fact, you don’t know that they are determinustic , because that would require empirical investigation which neither you nor Yudkowsky have performed. An assumption of determinism has been smuggled in by a choice of language, the use of the word “algorithm”.
The issue of conscious control over decision making is interesting and important as well, but it is at best one third of the whole issue.
One of the disconnects between Yudkowsky’s original post and several of the comments here is that his post was concerned not primarily with making a philosophical case for one position or another, but instead for shifting the debate from philosophy to cognitive psychology. Why, on an anatomical and information processing level, does it feel like a debate about free will vs. determinism makes sense to have in the first place?
Even if we can dissolve or resolve the philosophical question, this psychological question remains in full force.
Since several people seem to have missed this point (unless it’s ME who’s missing a point!), I clearly should have reiterated his question more clearly at the top of this post!
There’s potentially a cognitive explanation for every other debate including coke versus pepsi. But that doesn’t explain away anything—it’s just a finer level of description of the same thing.
I don’t think Yudkowsky saw himself as making a psychological point that leaves the philosophical question unanswered, I think he himself as answering the one and only real question, and most of his readers see things the same way.
I have sort of unlearnt how to think of free will in a nondeterministic sense. As such, I tripped over the part where you said there were “arguments against free will.” Like, yes of course the sensation of volition is produced by a deterministic, predictable process; how else could it be about the deciding process? Aboutness only exists in causal systems.
A more interesting question may be what the sensation is for? What part of our high-level cognition depends on noticing that we are making a decision?
This post will make more sense if you read the linked post by Eliezer at the top. In it, he asks you to identify a “cognitive algorithm” by which the very idea that there is a debate between “free will” and “determinism” feels like it makes psychological sense.
There was a comment here, but I completely wiped it because it was too confused.
Sorry, but I can no longer participate in the free-will debate. Apparently I have unlearnt how to think in that particular broken way. Anything that has to do with indeterminism relating to choice is no longer legible to me.
A better question is, if you write a complex enough decision making algorithm, would there necessarily be a part of it that would naturally map into the free will quale?
A fundamental difficulty in thinking logically about free will is that it involves thinking logically.
Logic, by its very nature, has embedded as its most essential hidden premise a deterministic structure. This makes all reasoning chains, no matter what their subject (including this one), to be deterministic. In other words, a deterministic structure is imposed upon the elements that will be logically analyzed so that they can be logically analyzed.
This leads one, if they ignore this structure is present as the very first link in chain, then proceeds to analyze the entire chain minus this hidden first premise in an attempt to determine what can be abstracted out from it, to incur into an involuntary ‘begging the question’ and to conclude all elements present in the chain, and all their mutual relations, are strictly deterministic. And, by extensions, that free will doesn’t exist in reality, when the most we can actually say is that free will doesn’t exist as a deduced link within deterministically structured logical reasoning chains.
Notice that this doesn’t preclude free will from being part of deterministically structured logical reasoning chains, it only says where free will cannot be present. It can still be present as an irreducible axiomatic premise, an “assuming free will exists...” used to reach further deductions. But that’s it. Any attempt at moving it from the position of an axiom down into the chain proper will invariably fail because the chain itself doesn’t admit of it.
I’m not sure how logical deduction is related in any way at all to physical (or even psychological) determinism. It is normal, and extremely common, to reason logically about non-deterministic systems.
Even logic itself is non-deterministic, when viewed as a sequential system. From “A and (if A then B)” you can derive “B”, as usual. You can also derive “A and B”, or “(A or not-B) and (if not-B then not-A)” or any of an infinite number of other sentences. Nothing says which you will logically derive, it just says what you can and what you can’t.
Formal logic, mathematics, informal deductive reasoning, algorithmics etc. are all interchangeable for the effects of my point, and usually also mutually translatable. Using any of them to model reality always yields a deterministic chain even when probabilistic paths are involved, because on can always think of these as branching in a manner similar to MWI: starting from such and such probabilities (or likelihoods, if the question is about one’s knowledge of the world rather than about the world itself) we end up with a causal tree, each of whose branches, when looked backwards, forms a logic causal chain.
That’s why free will cannot be modeled in terms of probabilities or likelihoods. Inserting a RNG in a logical chain only makes it more complex, it doesn’t make it less deterministic, and again causes free will proper to disappear, as it’s then reduced to mere randomness.
If they’re all interchangeable for the effects of your point, then I’m even more unsure what your point is than when I started.
You seem to be mixing up “discussing non-determinism” with “doing non-deterministic things”, which seems like the essence of a map/territory confusion.
That’s a contradiction in terms
Which? Logical or causal?
In any case, the point of causal determinism is that there is only on possible outcome to a state, ie. only one path going forwards.
Huh?that’s not generally acknowledged.
If you mean an RNG as opposed to a pseudo RNG, yes it does make it less deterministic...by definition.
That is not universally acknowledged.
Not really. The sentence you split forms a single reasoning. The first part is the claim, the second is the justification for the claim. You can read them in reverse if you prefer, which would gives it a more syllogistic form.
Both, since causal determinism is logically modelled. More specifically, causal determinism is a subset and a consequence of logical determinism, which is inherent to all forms of logical reasoning, including this one.
That’s precisely what MWI and similar notions disagree with. But yes, if we assume a single world, then the consequence is one of the alternatives, and none of the others.
True. I’m arguing against the generally acknowledge view. My position is based on traditional non-physically-reducible qualia-based concepts of free will as present in, e.g., Aquinas and Aristotle.
Evidently, if one assumes all qualia is physically-reducible, then free will as such doesn’t exist and is a mere subjective interpretation of deterministic and/or randomly-determined processes, but that’s precisely the same I’ve said, except that coming from the other direction.
You mean:-
But that doesn’t make sense. The point of MWI is that the branching structure as whole is deterministic, not that the branches are individually.
That makes no sense. Maps aren’t territories, even though territories are modelled with maps. Modelling isn’t ontological identity.
That’s not obvious at all. It’s not obvious that being reducuble to physics is the same as being reducuble to deterministic physics, it’s not obvious that indeterministic physics can’t support free will, and it’s not obvious that you need a quale of free will to have free will. (Just as you can live and die without knowing you have a spleen).
That depends on how you consider probabilities. One usual take, when it comes to concrete events, is that the probability of something that actually happened is 1.0, since it actually happened. Therefore, when you look at a sequence of causes and events backwards, that is, as history, this after-the-fact sequence is always strictly deterministic even if every single one of its links had a less-than-1.0 probability of happening before it actually happened in that specific way.
Well, if you prefer that terminology, I can restate it this way: maps that only provide deterministic and/or probabilistic (which I understand as a superset of deterministic) nodes cannot deal with neither-deterministic-nor-probabilistic features of the territories they’re trying to make.
To provide an example: a map that only provides RF frequencies says nothing of colors unless it also maps the connection of RF frequencies, to colors, via visual cortexes and all the associated biological organs, and provides primitives for the qualia of those colors.
Sorry, I wasn’t clear. “Physical reducibility” is a technical expression that refers to the philosophical assumption that the whole of concrete object, that is, both its quantitative properties as well as its qualitative properties, arises exclusively from its quantitative properties, in other words, that the a concrete object is “nothing but” the physical object.
I’m not sure what you mean by “indeterministic physics”. Do you mean QM?
I’m not sure I understand this point either. Are you referring to philosophical zombies?
These a few problems with that. One is that you just figured out how the universe works without examining the the universe. Another is that it you can’t get MWI out if it...unless you regard it as a statement only about subjective probability.
But I think we have had this discussion elsewhere.
The unstated part of the argument being that free will must be neither-deterministic nor probabilistic?
I know what “reductionism” means. The problen is that you haven’t explained why reducing the quale of free will disposes of fee will, since you haven’t explained why free will “is” the quale of free will, or why free will (the ability as opposed to the quale) can’t be physically explained.
That’s the best known example.
Not really. You can conceivably have free will while having no qualia , or while having a bunch of qualia, but not that one.
I’m not sure I understood these two points. Can you elaborate?
Actually, the state part. It’s my original comment. Although maybe I wasn’t as clear as I thought I was about it.
This isn’t quite the same reductionism as understood in physics, it has to do with Whitehead’s discussion of the problem of bifurcationism in nature (see the next block for details). In this context even a Jupiter-sized Culture-style AI Mind orders of orders of magnitude more complex than a human brain still counts as “physical reduction” in regards to “objective corporeality” if one assumes its computations capable of qualia-perception.
Free will is always perceived as qualia. You perceive it in yourself and in others, similarly to how you perceive any other qualia.
Any attempt at reducing it to the physical aspects of a being describes at most the physical processes that occur in/with/to the object in correlation with that qualia. Therefore, two philosophical options arise:
a) One may assume the qualia thus perceived is as fundamental as the measurable properties of the corporeal object, thus irreducible to those measurable properties, and that the corporeal object is thus a composite of both measurable properties and qualia properties.
In this scenario the set of the measurable properties of a corporeal object can be abstracted from it forming a pseudo-entity, the “physical object”, which is the object studied via mensuration, that is, via mathematical (and by extension logical) procedures and all they provide, among which statistical and probabilistic methods. Any conclusion arrived through them is then understood to describe the “physical object”, which, being only part of the full corporeal object, makes any such conclusion partial by definition, as they never cover the entirety of all properties of the corporeal object, in particular never covering its qualitative properties, as all they ever cover are its quantitative properties.
b) Or one may assume the qualia thus perceived is a consequence of those measurable properties, reducible to them, and therefore the corporeal object is those measurable properties, that is, that the corporeal object and the physical object are one and the same.
The burden of proof for case “a” is much lighter than that of case “b”. In fact, case “a” is the null hypothesis, as it corresponds to our direct perception of the world. Case “b”, in contrast, goes against that perception, and therefore is the one that needs to provide proof of its assertions. In particular, in the case of free will, it’d need to identify all the measurables related to what’s perceived as free will, then show with absolute rigor that they produce the perceived qualia of free will in something formerly devoid of it, and then, somehow, make that generated qualia perceptible as qualia to qualia-perceivers.
To use a classic analogy, even something much more simple, such as showing that the qualia “color red” is the electromagnetic range from 400 to 484 THz cannot be done yet. Note that this isn’t the same as showing that the qualia “color red” is associated with and carried by that EM range. For instance, if I close my eyes and think about an apple, I can access that qualia without a 400~484 THz EM wave hitting my eyes. As such, my affirmation that the qualia “color red” is distinct from the EM wave is straightforward and needs no further proof, while any affirmation involving the assertion that the qualia “color red” is reducible to, first, the measurable physical property “400~484 THz EM wave”, second, to the measurable physical properties of neurons in a brain, are the ones that need thorough proof.
While any such proof—for colors, as the entry level “easy” case, then for the much more difficult stuff such as free will—doesn’t appear, opting for “a” or for “b” will remain an arbitrary preference, as philosophical arguments for one and for the other cancel out.
From the summary of the bifurcation problem I provided above I think it’s more clear what I mean as indeterministic. From an “a” point of view QM is still entirely about physical objects, saying much about their measurable properties but nothing actually about their qualia. Hence, all it says is that some aspects of corporeal objects are fuzzy, the range of that fuzziness however being strictly determined and that, if MWI is correct, even this fuzziness is more apparent than real, since what it really is saying is not that such physically measurable aspects are fuzzy, but rather that the physical object branches very deterministically into so many ways.
Whether such “fuzziness within a determined range in a single world” or such “deterministic branching in many worlds” works as carriers for, or in correlation to, qualia properties of the full corporeal object, including but not limited to the free will qualia perceived by qualia-perceivers, is an entirely different problem, and there’s no easy, straight jump from one domain to the other. I suppose there may be, but no matter how much physically measurable randomness properties one identifies and determines, there’s still no self-evident link between this property of the physical object and the “free will” qualia of the full corporeal object.
From the exposed, you may have determinations in the form of single values or that of value ranges with inherent randomness while having no qualia, but stating these physical determinations imply having the “free will” qualia is a logical jump.
Taking from the “color red” example again, you may have an extremely energetic 400~484 THz EM wave, and yet no “color red” qualia at all for the simple lack of any qualia-perceiver in its path, or for the lack of any qualia-perceiver who however lacks the ability to extract a “color red” qualia from that carrier, or because the EM wave was absorbed by a black body etc.
Hence, while physically measurable randomness may be a “free will” qualia carrier, the lack of qualia perception would still result in the “free will” qualia carried by it to be lost. Conversely, a qualia-perceiver may have free will even in the absence of the typical physical carrier of “free will” qualia, as in the analogous case of a mind capable of imagining the “color red” qualia despite the absence of it usual “400~484 THz EM wave” carrier.
The argument that probabilities of past events are always 1 doesn’t prove anything objective, anything about the universe, unless you can show that probability always has to be interpreted objectively. As it happens there is also a subjective explanation for the rule.
The probabilities or squared measures of the branches in MWI have to aim to 1. So you cant have more than one branch of probability 1.
Ok, but why should I care? The question I care about is whether free will per se exists.
I’ve always seen it as, the sensation of free will comes from the ability to predict one’s own actions before they occur, in such a way as those predictions themselves seem to have causal efficacy to make themselves come true. Some predictions have that feeling—that they themselves are what’s making the thing being predicted happen (that the part of the brain doing the predicting is outputting that straight to the part handling the behavior, and this hand-off of information is perceived by other regions) - and others don’t.
The former always feel like volition, even when they are “long range” (plans for the future, for instance), but particularly when they are “short range” (like moving one’s arm, which has a very volition-ish feeling). It’s a bit like imperative, or better yet, performative statements (“I now pronounce you man and wife”) in language, which directly cause the thing they are describing.
When you feel as if you did not freely will something, it’s always either because you did not predict it, or because your prediction did not include the further prediction that it, itself, was what was causing the event, due to some kind of dissociation between the part of the brain doing the predicting, and the part acting. (When I dream at night, for instance, I rarely have any sense of free will, because part of the nature of dreaming is that predictive mechanisms are relaxed and various portions of the brain are heavily dissociated from one another.)
I thought about that too! The reason I didn’t include it as a contributor to the “free will” sensation was that for me, being able to predict my next action seems like it cuts against free will. If I’m sitting here predicting that I’ll take a drink of my coffee, and haven’t taken a drink yet, and that prediction proves correct, it feels like in the intervening time between making the prediction and taking the drink, I did not have free will to choose whether or not to take the drink! The predictive part of my brain is in control, not me.
By contrast, if I sit here considering whether or not to take a drink of my coffee, and don’t know whether I’ll choose to do so or not (before some time limit expires), that feels like free will. If I can’t predict my own actions, yet physically feel in control of my body when I do act, then that is what I seem to experience as free will.
But I wouldn’t be surprised if people’s “free will” sensation arises in different contexts for different people.
That’s why I included the causal bit. There’s a different sensation between “I can predict it (because something besides me will make it happen)” and “I can predict it (because I will make it happen)”. The latter feels like a plan, or an intention, with an energy of “the prediction itself is what’s doing the work” whereas the former feels like an expectation. And the capacity to make plans or intentions and see them through, is part of what people mean when they talk about free will, it seems like.