What do you mean with the term “scientifically” in that sentence? If I put identity into Google Scholar I’m fairly sure I fill find a bunch of papers in respectable scientific journals that use the term.
I mean that if you have two carbon atoms floating around in the universe, and the next instance you swap their locations but keep everything else the same, there is no scientific way in which you could say that anything has changed.
Combine this with humans being just a collection of atoms, and you have no meaningful way to say that an identical copy of you is “not really you”. Also, ‘continuity of consciousness’ is just a specific sensation that this specific clump of atoms has at each point in time, except for all the times when it does not exist because the clump is ‘sleeping’. So Quirrel’s objection seems to have no merit (could be I’m missing something though).
“Obviously” is a fairly strong word. It makes some sense to label the negation of any emotion a emotionless state. Unfriendly AI doesn’t hate humans but is indifferent.
Yes, there is an insight to be had there, I will acknowledge that much.
However, to say that the opposite of a friendly AI is a paper clip maximiser is stupid. The opposite of an AI which wants to help you is very obviously an AI which wants to hurt you. Which is why the whole “AK version 2 riddle” just doesn’t work. The Patronus goes from “not thinking about death” (version 1) to “Valuing life over death” (version 2). The killing curse goes from “valuing death over life” (version 1) to “not caring about life” (version 2). You can visualise the whole thing as a line measuring just the one integer, namely “life-death preference”:
Value death over life (-1) ---- don’t think about it either way (0) ----- Value life over death (+1)
The patronus gets a boost by moving from 0 to +1. The killing curse gets a boost by moving from −1 to 0. That makes no sense. Why would the killing curse, which is powered by the exact opposite of the patronus, receive a boost in power by moving in the same direct as the Patronus which values life over death?
Only fake wisdom can get ridiculous results like this.
The patronus gets a boost by moving from 0 to +1. The killing curse gets a boost by moving from −1 to 0. That makes no sense. Why would the killing curse, which is powered by the exact opposite of the patronus, receive a boost in power by moving in the same direct as the Patronus which values life over death?
I parsed it as follows: the Killing Curse isn’t powered by death in the same way that the Patronus draws power from life, but it does require the caster not to value the life of an opponent. Hatred enables this, but it’s limited: it has to be intense, sustained hatred, and probably only hatred of a certain kind, since it takes some doing for neurologically typical humans to hate someone enough to literally want them dead. Indifference to life works just as well and lacks the limitations, but that’s probably an option generally available only to, shall we say, a certain unusual personality type.
Ideology might interact with this in interesting ways, though. I don’t know whether Death Eaters would count as being motivated by hate or indifference by the standards of the spell; my model of J.K. Rowling says “hate”, while my model of Eliezer says “indifference”.
Yes, that ideology is precisely what bothers me. Eliezer has a bone to pick with death so he declares death to be the ultimate enemy. Dementors now represent death instead of depression, patronus now uses life magic, and a spell that is based on hate is now based on emptiness. It’s all twisted to make it fit the theme, and it feels forced. Especially when there’s a riddle and the answer is ‘Eliezer’s password’.
I don’t know if MoR influenced the movies, but Deathly Hallows 1 or 2 showed an image of Death looking like the movie’s image of Dementors. It seems to me like a natural inference.
Isn’t that because the only static element of a dementor’s appearance is its black, concealing cloak, and that overlaps neatly with the Grim Reaper portrayal of death?
You say that like Rowling had no choice but to use this well-known image for Dementors. Also, they’re supposed to look somewhat like corpses underneath.
What are you trying to argue in the great-grandparent? What am I supposed to take from the black cloaks, aside from the fact that it makes Dementors look like Death? I can imagine that perhaps Rowling chose this appearance because it allowed a frightening reveal later on. But that reveal uses the words “rotting”, “death” and “deathly”. On our first sight of a Dementor she also compares it to something “dead” and “decayed”. She did this because fear of death seems near as universal as you can get. Dementors’ most feared ability, destruction of the soul, has the same explanation.
The parallels that MoR!Harry sees are real, and they exist because death is (widely held to be) bad.
“don’t think about it either way” does not necessarily mean indifference, it means reverting to default behaviour.
Humans are (mostly) pro-social animals with empathy and would not crush another human who just happens to be in their way—in that they differ from a falling rock. In fact, that’s the point of hate, it overrides the built-in safeguards to allow for harmful action. According to this view, to genuinely not give a damn about someone’s life is a step further. Obviously.
The thing about built-in default behaviour given by evolution is that it will not trigger in some cases.
“Unreliable elements were subjected to an alternative justice process”—subjected by who? What does an “alternative justice process” do? With enough static noun phrases, you can keep anything unpleasant from actually happening.
or HPMoR Ch.48
Your brain imagines a single bird struggling in an oil pond, and that image creates some amount of emotion that determines your willingness to pay. But no one can visualize even two thousand of anything, so the quantity just gets thrown straight out the window.
or HPMoR Ch.87
Because the way people are built, Hermione, the way people are built to feel inside [...] is that they hurt when they see their friends hurting. Someone inside their circle of concern, a member of their own tribe. That feeling has an off-switch, an off-switch labeled ‘enemy’ or ‘foreigner’ or sometimes just ‘stranger’. That’s how people are, if they don’t learn otherwise.
My point with that is, it’s completely in line with what Eliezer usually talks about, so you know it’s a perspective he holds, not just rationalization.
For completeness’ sake,
Not like certain people living in certain countries, who were, it was said, as human as anyone else; who were said to be sapient beings, worth more than any mere unicorn. But who nonetheless wouldn’t be allowed to live in Muggle Britain. On that score, at least, no Muggle had the right to look a wizard in the eye. Magical Britain might discriminate against Muggleborns, but at least it allowed them inside so they could be spat upon in person.
still feels off. Oh, wait, I know! Maybe Harry is being Stupid here. Or Eliezer is being a Bad Writer. Again.
1) Saying you can’t tell after the fact whether something occured is not the same as saying it never occured. The fact that we can’t experimentally determine if two carbon atoms have distinct identity is not, repeat not the same as saying that they don’t have separate identity. Maybe they do. You just can’t tell.
2) That has nothing to do with continuity of consciousness. Assume the existence of a perfect matter replicator. What do you expect to happen when you make a copy of yourself? Do you expect to suddenly find yourself inside of the copy? Let’s say that regardless of what you expect at that point, you end up in your same body as before, the old one not the new one. What do you expect to experience then, if you killed yourself? This has nothing, nothing to do with statements about quantum identity and equivalence of configuration spaces. It is about separating the concept of a representation of me, from an instance of that representation which is me. I expect to experience only what the instance of the representation which is currently typing this words will experience as it evolves into the future. If an exact copy of me was made at any time, that’d be pretty awesome. It’d be like having a truly identical twin. But it wouldn’t me me, and if this instance died, I wouldn’t expect to live on experiencing what the copy of me experiences.
3) Sleeping is a total non-sequiter. Do you expect that your brain is 100% shut off and disarticulated into individual neurons when you are in a sleeping state? No? That’s right—just because you don’t have memories, doesn’t mean you didn’t exist while sleep. You just didn’t form memories at the time.
1) As far as I understand it, atoms don’t have a specific ‘location’, there are only probabilities for where that atom might be at any given time. Given that it is silly to speak of individual atoms. Even if I misunderstood that part, it is still the case that two entities which have no discernible difference in principle are the same, as a matter of simple logic.
2) Asking “which body do you wake up in” is a wrong question. It is meaningless because there is no testable difference depending on your answer, it is not falsifiable even in principle. The simple fact is that if you copy Sophronius, you then have 2 Sophronius waking up later, each experiencing the sensation of being the original. Asking whose sensation is “real” is meaningless.
3) It is not a non-sequitur. Sleep interrupts your continuity of self. Therefore, if your existence depends on uninterrupted continuity of self, sleep would mean you die every night.
I notice that you keep using concepts like “you”, “I” and “self” in your defence of a unique identity. I suggest you try removing those concepts or any other that presupposes unique identity. If you cannot do that then you are simply begging the question.
1) Saying you can’t tell after the fact whether something occured is not the same as saying it never occured. The fact that we can’t experimentally determine if two carbon atoms have distinct identity is not, repeat not the same as saying that they don’t have separate identity. Maybe they do. You just can’t tell.
The linked article by Elizer Yudkowsky is straight up wrong for the following reasons:
(1) Eliezer’s understanding of the physics here is bunk. I’m actually a trained physist. He is not. But bonus points to you if you reject this argument because you shouldn’t accept my authority any more than you should accept his. I assume you read Griffiths’ Quantum Mechanics or a similar introductory book and came to your own conclusions?
(2) Specifically the experimental result Eliezer quotes has to do with how we calculate probabilities for quantum mechanical events. There are an infinitely many ways one could calculate probabilities—math describes the universe, it doesn’t constrain it. But if you do so naively, you end up with one answer if you treat “P1 at L1, P2 at L2” as a different state than “P1 at L2, P2 at L1″ than if you treat them as the same state. Experimental results show that the latter probabilities are correct. One interpretation is that P1 and P2 are the same particle, so the state is “P at L1, P at L2”. That’s one interpretation. Another perfectly valid interpretation is that “Particle of type
at L1, Particle of type
at L2″ is the actual state—that is to say that the particles keep their identity but identity doesn’t factor into the probabalistic calculus. That’s why the term used by phsyisits is distinguishable rather than identity. These particles are indistinguishable, but that does not mean they are identical. That would be an unwaranted inference.
(3) All of that is a moot point, because it doesn’t match up at all with what we are talking about: the continuity of self as it relates to human minds. Calculating probabilities about particles in boxes tells us nothing about whether I would expect to wake up in a computer after a destructive upload, or how that relates to a personal desire to cheat death. I don’t care about the particles making up my mind: I care about sustaining the never stopping information processing system which gives rise to my subjective experiance. It does not obviously follow that if my mind state were perfectly saved before I was shot in the head, and then at some distant point in the future a brain configured exactly like mine was created, that I would subjectively experience living on in the future. Not anymore than it makes sense to say that my recently deceased aunt lives on in my mother, her identical twin.
I assume you read Griffiths’ Quantum Mechanics or a similar introductory book and came to your own conclusions?
FWIW, I have a master’s degree in physics and I’m working to get a PhD (though in a subfield not closely related to the basics of QM; I’d trust say Scott Aaronson over myself even though he’s not a physicist).
Another perfectly valid interpretation is that “Particle of type
at L1, Particle of type
at L2″ is the actual state—that is to say that the particles keep their identity but identity doesn’t factor into the probabalistic calculus.
FWIW, I have a master’s degree in physics and I’m working to get a PhD.
Awesome. Please forgive my undeserved snark.
What do you mean by identity?
Honestly I’m not sure. I only envoke the concept of identity in response to nonsense arguments appearing on LessWrong. Normally when I say ‘identity’ i mean the concept of ‘self’ which is the whatever-it-is which experiences my perceptions, thoughts, inner monologues, etc, or whatever it is that gives rise to the experience of me. How this relates to distinguishability of particles in quantum mechanics, I don’t know.. which is kinda the point. When calculating probabilities, you treat two states as the same if they are indistinguishable … how this gets warped into explaining what I’d expect to experience while undergoing a destructive upload is beyond me.
Also, ‘continuity of consciousness’ is just a specific sensation that this specific clump of atoms has at each point in time
Or not. Memories are genuinely lost, if someone makes a Horcrux and then dies some years later. Moreover, according to the Defense Professor in snake form, the maker’s personality could also change due to influence from the (two) victim(s). The result need not act like the maker at time of casting would act if placed in a new environment.
I mean that if you have two carbon atoms floating around in the universe, and the next instance you swap their locations but keep everything else the same, there is no scientific way in which you could say that anything has changed.
Combine this with humans being just a collection of atoms, and you have no meaningful way to say that an identical copy of you is “not really you”. Also, ‘continuity of consciousness’ is just a specific sensation that this specific clump of atoms has at each point in time, except for all the times when it does not exist because the clump is ‘sleeping’. So Quirrel’s objection seems to have no merit (could be I’m missing something though).
Yes, there is an insight to be had there, I will acknowledge that much.
However, to say that the opposite of a friendly AI is a paper clip maximiser is stupid. The opposite of an AI which wants to help you is very obviously an AI which wants to hurt you. Which is why the whole “AK version 2 riddle” just doesn’t work. The Patronus goes from “not thinking about death” (version 1) to “Valuing life over death” (version 2). The killing curse goes from “valuing death over life” (version 1) to “not caring about life” (version 2). You can visualise the whole thing as a line measuring just the one integer, namely “life-death preference”:
Value death over life (-1) ---- don’t think about it either way (0) ----- Value life over death (+1)
The patronus gets a boost by moving from 0 to +1. The killing curse gets a boost by moving from −1 to 0. That makes no sense. Why would the killing curse, which is powered by the exact opposite of the patronus, receive a boost in power by moving in the same direct as the Patronus which values life over death?
Only fake wisdom can get ridiculous results like this.
I parsed it as follows: the Killing Curse isn’t powered by death in the same way that the Patronus draws power from life, but it does require the caster not to value the life of an opponent. Hatred enables this, but it’s limited: it has to be intense, sustained hatred, and probably only hatred of a certain kind, since it takes some doing for neurologically typical humans to hate someone enough to literally want them dead. Indifference to life works just as well and lacks the limitations, but that’s probably an option generally available only to, shall we say, a certain unusual personality type.
Ideology might interact with this in interesting ways, though. I don’t know whether Death Eaters would count as being motivated by hate or indifference by the standards of the spell; my model of J.K. Rowling says “hate”, while my model of Eliezer says “indifference”.
Yes, that ideology is precisely what bothers me. Eliezer has a bone to pick with death so he declares death to be the ultimate enemy. Dementors now represent death instead of depression, patronus now uses life magic, and a spell that is based on hate is now based on emptiness. It’s all twisted to make it fit the theme, and it feels forced. Especially when there’s a riddle and the answer is ‘Eliezer’s password’.
I don’t know if MoR influenced the movies, but Deathly Hallows 1 or 2 showed an image of Death looking like the movie’s image of Dementors. It seems to me like a natural inference.
Isn’t that because the only static element of a dementor’s appearance is its black, concealing cloak, and that overlaps neatly with the Grim Reaper portrayal of death?
You say that like Rowling had no choice but to use this well-known image for Dementors. Also, they’re supposed to look somewhat like corpses underneath.
I increasingly feel like I’ve lost track of what you’re trying to argue here. Would you mind recapitulating it for me?
What are you trying to argue in the great-grandparent? What am I supposed to take from the black cloaks, aside from the fact that it makes Dementors look like Death? I can imagine that perhaps Rowling chose this appearance because it allowed a frightening reveal later on. But that reveal uses the words “rotting”, “death” and “deathly”. On our first sight of a Dementor she also compares it to something “dead” and “decayed”. She did this because fear of death seems near as universal as you can get. Dementors’ most feared ability, destruction of the soul, has the same explanation.
The parallels that MoR!Harry sees are real, and they exist because death is (widely held to be) bad.
“don’t think about it either way” does not necessarily mean indifference, it means reverting to default behaviour.
Humans are (mostly) pro-social animals with empathy and would not crush another human who just happens to be in their way—in that they differ from a falling rock. In fact, that’s the point of hate, it overrides the built-in safeguards to allow for harmful action. According to this view, to genuinely not give a damn about someone’s life is a step further. Obviously.
The thing about built-in default behaviour given by evolution is that it will not trigger in some cases.
Rationality and the English Language
or HPMoR Ch.48
or HPMoR Ch.87
My point with that is, it’s completely in line with what Eliezer usually talks about, so you know it’s a perspective he holds, not just rationalization.
For completeness’ sake,
still feels off. Oh, wait, I know! Maybe Harry is being Stupid here. Or Eliezer is being a Bad Writer. Again.
Yes you are missing a few things.
1) Saying you can’t tell after the fact whether something occured is not the same as saying it never occured. The fact that we can’t experimentally determine if two carbon atoms have distinct identity is not, repeat not the same as saying that they don’t have separate identity. Maybe they do. You just can’t tell.
2) That has nothing to do with continuity of consciousness. Assume the existence of a perfect matter replicator. What do you expect to happen when you make a copy of yourself? Do you expect to suddenly find yourself inside of the copy? Let’s say that regardless of what you expect at that point, you end up in your same body as before, the old one not the new one. What do you expect to experience then, if you killed yourself? This has nothing, nothing to do with statements about quantum identity and equivalence of configuration spaces. It is about separating the concept of a representation of me, from an instance of that representation which is me. I expect to experience only what the instance of the representation which is currently typing this words will experience as it evolves into the future. If an exact copy of me was made at any time, that’d be pretty awesome. It’d be like having a truly identical twin. But it wouldn’t me me, and if this instance died, I wouldn’t expect to live on experiencing what the copy of me experiences.
3) Sleeping is a total non-sequiter. Do you expect that your brain is 100% shut off and disarticulated into individual neurons when you are in a sleeping state? No? That’s right—just because you don’t have memories, doesn’t mean you didn’t exist while sleep. You just didn’t form memories at the time.
1) As far as I understand it, atoms don’t have a specific ‘location’, there are only probabilities for where that atom might be at any given time. Given that it is silly to speak of individual atoms. Even if I misunderstood that part, it is still the case that two entities which have no discernible difference in principle are the same, as a matter of simple logic.
2) Asking “which body do you wake up in” is a wrong question. It is meaningless because there is no testable difference depending on your answer, it is not falsifiable even in principle. The simple fact is that if you copy Sophronius, you then have 2 Sophronius waking up later, each experiencing the sensation of being the original. Asking whose sensation is “real” is meaningless.
3) It is not a non-sequitur. Sleep interrupts your continuity of self. Therefore, if your existence depends on uninterrupted continuity of self, sleep would mean you die every night.
I notice that you keep using concepts like “you”, “I” and “self” in your defence of a unique identity. I suggest you try removing those concepts or any other that presupposes unique identity. If you cannot do that then you are simply begging the question.
Well...
The linked article by Elizer Yudkowsky is straight up wrong for the following reasons:
(1) Eliezer’s understanding of the physics here is bunk. I’m actually a trained physist. He is not. But bonus points to you if you reject this argument because you shouldn’t accept my authority any more than you should accept his. I assume you read Griffiths’ Quantum Mechanics or a similar introductory book and came to your own conclusions?
(2) Specifically the experimental result Eliezer quotes has to do with how we calculate probabilities for quantum mechanical events. There are an infinitely many ways one could calculate probabilities—math describes the universe, it doesn’t constrain it. But if you do so naively, you end up with one answer if you treat “P1 at L1, P2 at L2” as a different state than “P1 at L2, P2 at L1″ than if you treat them as the same state. Experimental results show that the latter probabilities are correct. One interpretation is that P1 and P2 are the same particle, so the state is “P at L1, P at L2”. That’s one interpretation. Another perfectly valid interpretation is that “Particle of type
at L1, Particle of type
at L2″ is the actual state—that is to say that the particles keep their identity but identity doesn’t factor into the probabalistic calculus. That’s why the term used by phsyisits is distinguishable rather than identity. These particles are indistinguishable, but that does not mean they are identical. That would be an unwaranted inference.
(3) All of that is a moot point, because it doesn’t match up at all with what we are talking about: the continuity of self as it relates to human minds. Calculating probabilities about particles in boxes tells us nothing about whether I would expect to wake up in a computer after a destructive upload, or how that relates to a personal desire to cheat death. I don’t care about the particles making up my mind: I care about sustaining the never stopping information processing system which gives rise to my subjective experiance. It does not obviously follow that if my mind state were perfectly saved before I was shot in the head, and then at some distant point in the future a brain configured exactly like mine was created, that I would subjectively experience living on in the future. Not anymore than it makes sense to say that my recently deceased aunt lives on in my mother, her identical twin.
FWIW, I have a master’s degree in physics and I’m working to get a PhD (though in a subfield not closely related to the basics of QM; I’d trust say Scott Aaronson over myself even though he’s not a physicist).
What do you mean by identity?
Awesome. Please forgive my undeserved snark.
Honestly I’m not sure. I only envoke the concept of identity in response to nonsense arguments appearing on LessWrong. Normally when I say ‘identity’ i mean the concept of ‘self’ which is the whatever-it-is which experiences my perceptions, thoughts, inner monologues, etc, or whatever it is that gives rise to the experience of me. How this relates to distinguishability of particles in quantum mechanics, I don’t know.. which is kinda the point. When calculating probabilities, you treat two states as the same if they are indistinguishable … how this gets warped into explaining what I’d expect to experience while undergoing a destructive upload is beyond me.
Or not. Memories are genuinely lost, if someone makes a Horcrux and then dies some years later. Moreover, according to the Defense Professor in snake form, the maker’s personality could also change due to influence from the (two) victim(s). The result need not act like the maker at time of casting would act if placed in a new environment.
See also major’s point.