I believe that Eliezer’s analysis of “free will” answers your question. Free will (he says) is neither outside of an otherwise lawful universe, nor incompatible with a lawful universe, nor merely compatible with a lawful universe, but requires a lawful universe. He dubs this position “requiredism”.
I find this not merely convincing, but obviously right. What do you think?
This is somewhat dated in the sense that LW-style decision theory later converged on treating agents-that-make-decisions as abstract algorithms rather than as their instances embedded in the world, see discussion of “algorithm” axis of classifying decision theories in this post.
With TAG, I don’t see what their decision theory has to do with the matter. Whatever their decision theory, it is impotent to achieve anything unless their physical instances embedded in the world are able to physically act in the world to achieve their aims, which is the thing that incompatibilists deny.
The point is about the frame of Yudkowsky’s explanation, where “you are physics” instead of “you are an algorithm”. The latter seems convergently more useful for decision theory of embedded agents, which can be predicted by other agents or can have multiple copies. So this doesn’t concern some prior meaning of “free will”, it motivates caring about a notion of free will that has to do with abstract computations of agent’s decisions rather than agent instances embedded in the world.
You are an algorithm embedded in physics. You are not any of the other people executing this algorithm, you are this one. Conducting yourself according to these decision theories still causes the physical actions only of this one, and is only acausally connected to the others of which these theories speak. Deciding as if deciding for all is different from causally deciding for all.
You are an algorithm embedded in physics. You are not any of the other people executing this algorithm, you are this one.
There is an algorithm and the person executing the algorithm, different entities. Being the algorithm, you are not the person executing it. The algorithm is channeled by the person instantiating it concretely (in full detail) as well as other people who might be channeling approximations to it, for example only getting to know that the algorithm’s computation satisfies some specification instead of knowing everything that goes on in its computation.
Conducting yourself according to these decision theories still causes the physical actions only of this one
The use of “you are the algorithm” frame is noticing that other instances and predictors of the same algorithm have the same claim to consequences of its behaviors, there is no preferred instance. The actions of the other instances and of the predictors, if they take place in the world, are equally “physical” as those of the putative “primary” instance.
only acausally connected to the others of which these theories speak
As an algorithm, you are acausally connected to all instances inclusing the “primary” instance in the same sense, by their reasoning about you-the-algorithm.
Deciding as if deciding for all is different from causally deciding for all.
I don’t know what “causally deciding” means for algorithms. Deciding as if deciding for all is actually an interesting detail, it’s possible to consider its variants where you are only deciding for some, and that stipulation creates different decision problems depending on the collection of instances that are to be controlled by a given decision (a subset of all instances that could be controlled). This can be used to set up coalitions of interventions that the algorithm coordinates. The algorithm instances that are left out of such decision problems are left with no guidance, which is analogous to them failing to compute the specification (predict/compute/prove an action-relevant property of algorithm’s behavior), a normal occurrence. It also illustrates that the instances should be ready to pick up the slack when the algorithm becomes unobservable.
You still haven’t said why it motivates that. Even if you are not talking about a prior definition of free will, why does your novel definition have to do with algorithms?
Why would decision theory determine the nature of free will? I would have though it was the other way round: free will has implications for what decisions are.
Requiredism holds that determinism is an advantage to fee will because the connection between a decision and the resulting action is deterministic. Randomness, or at least, too much randomness in the wrong place, would prevent me from acting reliably on my decisions Of course, determinism also removes the elbow room, the ability to have decided differently, that is of such concern to libertarians. Determinism is only an overall advantage to free will if elbow room is unimportant or impossible, so Requiredism needs compatibilism as a starting point.
I don’t think it’s obvious that libertarian style elbow room, or CHDO, is unimportant or impossible, so I don’t think requiredism is obvious,.
It’s also pretty clear that 100% determinism doesn’t imply 100% reliability. If you reach to pick up a cup, and knock it over, that could be a determined event .. and of course, such errors are fairly common.Such considerations leads to an argument against reliablism: reliable action only needs a good enough level of determinism, because we don’t put decisions into practice with 100% success, and a good enough level of determinism is compatible with a small degree of indeterminism which could allow indeterministic free will.
If ‘lawful’ just means ‘not completely random’ then I agree. But I’ve never been convinced that there’s no conceivable third option beside ‘random’ and ‘deterministic’. Setting aside whether there’s a non-negligible chance that it’s true, do you think the idea that consciousness plays some mysterious-to-us causal role—one that isn’t accurately captured by our concepts of randomness or determinism—is definitely incoherent?
Consciousness does play a mysterious-to-us-today causal role. It is mysterious, in that no-one has yet explained how there can possibly be such a thing as subjective experience, yet there it is. Perhaps someone might explain it in the future, but no-one has done so today. And it must be causal, not epiphenomenal, because the doctrine of epiphenomenalism just adds another layer of mysteriousness on top of that one, explaining the obscure by the more obscure. Epiphenomenalism is no more coherent an idea than p-zombies.
Randomness vs. determinism is a red herring. The universe has to be lawful, for us to be able to direct it into desired configurations. Randomness, such as some current theories of quantum mechanics say are physically unreducible to determinism, is an obstacle to doing that, but has no more significance than that. That goes for chaos as well, which some put forward as a “third alternative” to randomness and determinism. But none of these matter for this view of what “free will” is.
I recommend that people do click through to the article by Eliezer that I linked before, if they haven’t already. It’s not very long, and any précis I could write would just be a repetition of it. Epiphenomenalism, btw, is described by the first diagram in that article.
And it must be causal, not epiphenomenal, because the doctrine of epiphenomenalism just adds another layer of mysteriousness on top of that one, explaining the obscure by the more obscure.
I don’t follow this. Adding another layer of mysteriousness might not make for a satisfying explanation, but why must it be false? (I also think the p-zombie is a perfectly coherent idea, for whatever that’s worth.)
When I say “must” I’m rounding to zero probabilities so negligible that they should not even come to my attention. Epiphenomenalism has consciousness be a real thing (that is what it is a theory of) but which has only a one-way connection to the rest of the universe, like a redundant gear in a clock that is not part of the train that drives the hands. Nowhere else do we see such a thing; in fact, by definition, we could not. The hypothesis is doing no work.
Nowhere else do we see such a thing; in fact, by definition, we could not.
I think the second clause implies that our not seeing it anywhere else provides no evidence. (Just for the obvious Bayesian reason.)
The hypothesis is doing no work.
I’m not sure why it has to. The ‘consciousness is real’ part isn’t a hypothesis; it’s the one thing we can safely take as an axiom. And the ‘consciousness doesn’t affect anything else’ part is as reasonable a candidate for the null hypothesis as any other, as far as I can tell. Where does your prior against redundant gears come from?
What would legitimately draw the hypothesis to our attention? One of the things that we have experience of is being able to act in the world. Epiphenomenalism says that we do not act in the world, we are merely passengers without the power to so much as twitch our little fingers. This is so plainly absurd that only a philosopher could take it seriously, but as Cicero remarked more than two thousand years ago, no statement is too absurd for some philosophers to make.
What would legitimately draw the hypothesis to our attention?
The fact that subjective experience exists and we haven’t been able to figure out any causal role that it plays, other than that which seems to be explicable by ordinary physics (and with reference only to its ordinary physical correlates).
Epiphenomenalism says that we do not act in the world, we are merely passengers without the power to so much as twitch our little fingers. This is so plainly absurd that only a philosopher could take it seriously
I’ve been trying to articulate why I find it hard to reconcile this with your endorsing Eliezer’s requiredism, and this is the best I can do:
I don’t think I see a meaningful difference between epiphenomenalism (i.e. brain causes qualia, qualia don’t cause anything) and a non-eliminative materialism that says ‘qualia and brain are not separate things; there’s just matter, and sometimes matter has feelings, but the matter that has feelings still obeys ordinary physical laws’. In both cases, qualia are real but physics is causally closed and there’s no room for libertarian free will.
If the quoted passage referred to that kind of materialism rather than to epiphenomenalism, it would be an argument for libertarianism. And I know that’s not what you intended, but I don’t fully understand what you do mean by it, given that it must not conflict with requiredism (which is basically ‘compatibilism but more so’).
In both cases, qualia are real but physics is causally closed and there’s no room for libertarian free will.
Libertarian FW isn’t ruled out by the causal closure of the physical, it’s ruled out by determinism (physical or not). Causal closure would rule out something like interactionist dualism, but that’s fairly orthogonal to LFW...it could even be deterministic.
I don’t see a difference between that argument and saying that a jumbo jet doesn’t cause anything, only its atoms do.
Happy to leave this here if you’ve had enough, but if you do want a response I’ll need more than that to go on. I’ve been struggling to understand how your position fits together, and that doesn’t really help. (I’m not even sure exactly what you’re referring to as ‘that argument’. Admittedly I am tired; I’ll take a break now.)
I believe that Eliezer’s analysis of “free will” answers your question. Free will (he says) is neither outside of an otherwise lawful universe, nor incompatible with a lawful universe, nor merely compatible with a lawful universe, but requires a lawful universe. He dubs this position “requiredism”.
I find this not merely convincing, but obviously right. What do you think?
This is somewhat dated in the sense that LW-style decision theory later converged on treating agents-that-make-decisions as abstract algorithms rather than as their instances embedded in the world, see discussion of “algorithm” axis of classifying decision theories in this post.
With TAG, I don’t see what their decision theory has to do with the matter. Whatever their decision theory, it is impotent to achieve anything unless their physical instances embedded in the world are able to physically act in the world to achieve their aims, which is the thing that incompatibilists deny.
The point is about the frame of Yudkowsky’s explanation, where “you are physics” instead of “you are an algorithm”. The latter seems convergently more useful for decision theory of embedded agents, which can be predicted by other agents or can have multiple copies. So this doesn’t concern some prior meaning of “free will”, it motivates caring about a notion of free will that has to do with abstract computations of agent’s decisions rather than agent instances embedded in the world.
You are an algorithm embedded in physics. You are not any of the other people executing this algorithm, you are this one. Conducting yourself according to these decision theories still causes the physical actions only of this one, and is only acausally connected to the others of which these theories speak. Deciding as if deciding for all is different from causally deciding for all.
There is an algorithm and the person executing the algorithm, different entities. Being the algorithm, you are not the person executing it. The algorithm is channeled by the person instantiating it concretely (in full detail) as well as other people who might be channeling approximations to it, for example only getting to know that the algorithm’s computation satisfies some specification instead of knowing everything that goes on in its computation.
The use of “you are the algorithm” frame is noticing that other instances and predictors of the same algorithm have the same claim to consequences of its behaviors, there is no preferred instance. The actions of the other instances and of the predictors, if they take place in the world, are equally “physical” as those of the putative “primary” instance.
As an algorithm, you are acausally connected to all instances inclusing the “primary” instance in the same sense, by their reasoning about you-the-algorithm.
I don’t know what “causally deciding” means for algorithms. Deciding as if deciding for all is actually an interesting detail, it’s possible to consider its variants where you are only deciding for some, and that stipulation creates different decision problems depending on the collection of instances that are to be controlled by a given decision (a subset of all instances that could be controlled). This can be used to set up coalitions of interventions that the algorithm coordinates. The algorithm instances that are left out of such decision problems are left with no guidance, which is analogous to them failing to compute the specification (predict/compute/prove an action-relevant property of algorithm’s behavior), a normal occurrence. It also illustrates that the instances should be ready to pick up the slack when the algorithm becomes unobservable.
You still haven’t said why it motivates that. Even if you are not talking about a prior definition of free will, why does your novel definition have to do with algorithms?
Why would decision theory determine the nature of free will? I would have though it was the other way round: free will has implications for what decisions are.
Requiredism holds that determinism is an advantage to fee will because the connection between a decision and the resulting action is deterministic. Randomness, or at least, too much randomness in the wrong place, would prevent me from acting reliably on my decisions Of course, determinism also removes the elbow room, the ability to have decided differently, that is of such concern to libertarians. Determinism is only an overall advantage to free will if elbow room is unimportant or impossible, so Requiredism needs compatibilism as a starting point.
I don’t think it’s obvious that libertarian style elbow room, or CHDO, is unimportant or impossible, so I don’t think requiredism is obvious,.
It’s also pretty clear that 100% determinism doesn’t imply 100% reliability. If you reach to pick up a cup, and knock it over, that could be a determined event .. and of course, such errors are fairly common.Such considerations leads to an argument against reliablism: reliable action only needs a good enough level of determinism, because we don’t put decisions into practice with 100% success, and a good enough level of determinism is compatible with a small degree of indeterminism which could allow indeterministic free will.
If ‘lawful’ just means ‘not completely random’ then I agree. But I’ve never been convinced that there’s no conceivable third option beside ‘random’ and ‘deterministic’. Setting aside whether there’s a non-negligible chance that it’s true, do you think the idea that consciousness plays some mysterious-to-us causal role—one that isn’t accurately captured by our concepts of randomness or determinism—is definitely incoherent?
Consciousness does play a mysterious-to-us-today causal role. It is mysterious, in that no-one has yet explained how there can possibly be such a thing as subjective experience, yet there it is. Perhaps someone might explain it in the future, but no-one has done so today. And it must be causal, not epiphenomenal, because the doctrine of epiphenomenalism just adds another layer of mysteriousness on top of that one, explaining the obscure by the more obscure. Epiphenomenalism is no more coherent an idea than p-zombies.
Randomness vs. determinism is a red herring. The universe has to be lawful, for us to be able to direct it into desired configurations. Randomness, such as some current theories of quantum mechanics say are physically unreducible to determinism, is an obstacle to doing that, but has no more significance than that. That goes for chaos as well, which some put forward as a “third alternative” to randomness and determinism. But none of these matter for this view of what “free will” is.
I recommend that people do click through to the article by Eliezer that I linked before, if they haven’t already. It’s not very long, and any précis I could write would just be a repetition of it. Epiphenomenalism, btw, is described by the first diagram in that article.
I don’t follow this. Adding another layer of mysteriousness might not make for a satisfying explanation, but why must it be false? (I also think the p-zombie is a perfectly coherent idea, for whatever that’s worth.)
When I say “must” I’m rounding to zero probabilities so negligible that they should not even come to my attention. Epiphenomenalism has consciousness be a real thing (that is what it is a theory of) but which has only a one-way connection to the rest of the universe, like a redundant gear in a clock that is not part of the train that drives the hands. Nowhere else do we see such a thing; in fact, by definition, we could not. The hypothesis is doing no work.
And I see p-zombies as another incoherent idea.
I think the second clause implies that our not seeing it anywhere else provides no evidence. (Just for the obvious Bayesian reason.)
I’m not sure why it has to. The ‘consciousness is real’ part isn’t a hypothesis; it’s the one thing we can safely take as an axiom. And the ‘consciousness doesn’t affect anything else’ part is as reasonable a candidate for the null hypothesis as any other, as far as I can tell. Where does your prior against redundant gears come from?
What would legitimately draw the hypothesis to our attention? One of the things that we have experience of is being able to act in the world. Epiphenomenalism says that we do not act in the world, we are merely passengers without the power to so much as twitch our little fingers. This is so plainly absurd that only a philosopher could take it seriously, but as Cicero remarked more than two thousand years ago, no statement is too absurd for some philosophers to make.
The fact that subjective experience exists and we haven’t been able to figure out any causal role that it plays, other than that which seems to be explicable by ordinary physics (and with reference only to its ordinary physical correlates).
We have also not figured out how the physical brain does the things that we do.
Agreed, and that’s part of why I see mysterious libertarian free will as not having been ruled out.
I’ve been trying to articulate why I find it hard to reconcile this with your endorsing Eliezer’s requiredism, and this is the best I can do:
I don’t think I see a meaningful difference between epiphenomenalism (i.e. brain causes qualia, qualia don’t cause anything) and a non-eliminative materialism that says ‘qualia and brain are not separate things; there’s just matter, and sometimes matter has feelings, but the matter that has feelings still obeys ordinary physical laws’. In both cases, qualia are real but physics is causally closed and there’s no room for libertarian free will.
If the quoted passage referred to that kind of materialism rather than to epiphenomenalism, it would be an argument for libertarianism. And I know that’s not what you intended, but I don’t fully understand what you do mean by it, given that it must not conflict with requiredism (which is basically ‘compatibilism but more so’).
Libertarian FW isn’t ruled out by the causal closure of the physical, it’s ruled out by determinism (physical or not). Causal closure would rule out something like interactionist dualism, but that’s fairly orthogonal to LFW...it could even be deterministic.
I don’t see a difference between that argument and saying that a jumbo jet doesn’t cause anything, only its atoms do.
Happy to leave this here if you’ve had enough, but if you do want a response I’ll need more than that to go on. I’ve been struggling to understand how your position fits together, and that doesn’t really help. (I’m not even sure exactly what you’re referring to as ‘that argument’. Admittedly I am tired; I’ll take a break now.)