Zach is right. We do not directly react to what happens to us, only to how our mind feels about what happens to us. In a way, we are already in a simulation because we derive pleasure or misery through how our mind interprets our experience. Which is why, as you know, the same exact sensory input can create very different emotional states depending on the sensory input that preceded it. Same stuff, different interpretation by our minds.
Don’t assume that those who choose the experience machine are hedonists. Maybe their lives suck and they just want something more than despair to look forward to. Also, I have been through the experience of being in a relationship and realizing years later the person never loved me. Not even sure they liked me. It not only changes the experience but it changes everything afterwards. Sometimes, ignorance is bliss.
I wonder if the allure of "living in reality" is analogous to that of dualism. If there's nothing "special" about the human experience and everything reduces to brain states, then maybe there's nothing "special" about "reality" and everything reduces to experiences (as Zach seems to argue). In other words, perhaps to people who would enter the Experience Machine, arguments for not entering sound similar to arguments for dualism.
I don't think this scenario is that far away. We can already read and control minds, so I guess the machine is just around the corner. My fear is that sadist will put me in it and set it in maximum torture mode. Hell. I don't think this is far into the furture either. Imagine the joy sadistic people can wreak on others if they get their hands on a machine like that. When we can control the pain centers, and not just physical but emotional pain.
Zach's friend here 👋 He has tormented me with these arguments for many moons.
He's right that the imagination exercise is flawed. You cannot appropriately evaluate the thought experiment because you cannot imagine yourself in an unreal situation <where what we are evaluating is precisely the fact that it is unreal> without knowing that it is unreal. The whole point is that you don't know it is unreal--this makes it impossible to imagine.
This does nothing to support Zach's actual view, namely: "If something is worse for me, it must feel worse for me. It can’t be worse if it makes no psychological difference." All it shows is that the experience machine is a poor thought experiment for evaluating the principle.
Here is the argument I've levied to Zach before. It's quite simple, really.
1. It's good to get what I want. (perhaps not *all* desires, but all things being equal, the vast majority of them)
2. I want plenty of things which are not accomplished by my experiencing them. I hope, for instance, that Zach has a good week. I hope that my family complete and publish my masterwork after my death.
3. Therefore, It is good for me that things to happen though I do not experiencing them, since they accomplish my desires. It is better, for instance, that my life's work gets accomplished, rather than ruined, after I die.
Zach might respond to this, as Hobbes did, that I want the pleasure of the experience of knowing that he had a good week. But if that were what I was really after, I would seek out evidence to confirm that he had a good week. If that were true, I would ask him how he is doing more often.
The M&M question is interesting, but I would prefer the simulation M&Ms, though I would not enter the experience machine. The pleasure of M&Ms has nothing to do, for me, with the conviction that it is real. The pleasures whose reality matter, I think, are those that are founded on shared experiences. For instance, for many, sex is one of those things. When A's pleasure is founded on a certain representation of how the world is, say, that through her actions B is feeling pleasure, then, if that representation is mistaken, the pleasure is misfounded, and I am deluded. If you only care about experiences like Zach, the delusion doesn't matter. To me, like Paul it matters immensely. Deluded pleasure can be worse than pain.
Hopefully this offers readers some resources to resist Zach's sophistry.
Yea, so—I agree with (1) that it's good for you to get what you want, but this is instrumental. Getting what you want often makes your life (that is, your experiences) better. It can also make your life worse, though, and it can—trivially—make no difference at all.
Suppose you had a very strong preference that the number of grains of sand on the planet be even, rather than odd. That's either satisfied or it isn't, but I don't think the quality of your life hinges on it *at all*.
Which leads me to (2). I don't doubt that you do want these things (which is very sweet of you) but I think those desires are rational insofar as they can influence your experiences and irrational otherwise. On your view, your answer to "how is your life going?" will always be "I have no idea, because I'm missing so much critical data about whether or not my desires are fulfilled. Are my loved ones actually happy? Do they actually love me? How many grains of sand are there?" etc.
And even death does not set you free! The quality of your life is still in flux, your desires still being satisfied or unsatisfied based on the way you are remembered, the lives of those you lead behind, the progress of philosophy, and etc. etc.
So it is with tenderness and mercy that I reject (3), setting you free from such a totalitarian model of 'wellbeing'.
It's true that some desires don't seem matter, and it's a challenge for desire-satisfaction theories of well-being (or any theory of well-being that involves desire-satisfaction as one measure, i.e., any plausible one) to distinguish between relevant and irrelevant desires for assessing well-being. A plausible suggestion, for instance, is that the desires that involve me are the only ones that count. Aristotle classically involves the fate of your families and friends as relevant also. Regardless, the desires that involve me extend beyond the desires that involve my experiences. I want, for example, my masterwork to be published, regardless of whether I am there to see it come to fruition.
You wish restrict the 'rational' set of desires to ones about my experience. On what grounds, other than simply assuming the principle you are supposed to be arguing for? The evidence on my side is that people pursue and care about things beyond their experience, and I take what matters *to* them as a reliable guide of what matters *for* them. I've no idea why you argue that an unrestricted desire-satisfaction theory is 'totalitarian'--please note that you are the one excluding from rationality a vast array of human desires.
Finally, your point about having no idea whether my life is good is deeply misguided. It is perfectly compatible with justified belief to infer on the basis of inconclusive evidence. I can think my life is going well based on the evidence I have, knowing that I might be mistaken about that. Saying that I am entirely ignorant because I might be wrong is just a manifestation of general radical skepticism. As for whether death might set you free, the vast majority of theorists I've encountered on the subject find nothing strange in the idea that your life might go better or worse after you die. The very first treatment of this question by Aristotle argues decisively in favour.
I've said this a great many times Zach, but it just seems to me that *you* care exclusively about your experiences. That is no grounds for arguing that everyone else *should* care only about their experiences. I've never heard you offer a single argument in favour of that normative claim. And the empirical claim that people in fact only care about their experiences is false, as I've argued re: my caring about your week.
I think self-reporting without having actually gone through the experience of the Experience Machine is potentially unreliable. If I imagine the experience of being in it as completely seamless and reliable I don’t see how one could turn it down. This is our one life. Maximizing the happiness, stimulation, peace and contentment (or whatever positive feelings one considers most valuable) just clearly is the obvious choice. If a feeling of inauthenticity is detracting from one’s experience in the EM the technicians will tweak the settings and eliminate that issue.
Great post. I think the reverse set-up is an important twist on the original thought experiment.
I am stunned that anyone would even consider plugging in. Maybe I have trust issues but I would worry about the machine breaking down or someone replugging me into a pain machine just for fun. I don’t think I’m paranoid but allowing anyone or anything to control my experience of life is just unthinkable.
There is, of course, an extreme scenario where your current life is unbearably painful but even then there is always the option to terminate it. And who is to say that your afterlife won’t be pure pleasure?
I’m on team Zach. We can’t know that the reality we are in right now is *really* reality.
Lately I’ve been going to sleep with a nicotine pouch in my mouth and it gives me the most vivid dreams. They feel very real in part because they’re not especially bizarre. They’re all plausible situations from real life.
It wouldn’t shock me to find that the dream is me typing this comment and I will wake up shortly.
It is true- I will never know if I am in or out of the machine or if a machine exists. Given a choice,though,so I theoretically might know-I cling to the idea of individual minds functioning within mostly random material circumstance and occasionally creating new ideas .
I like the purity of the M&M thought experiment, but when I answered the question I simulated the probable annoyance of booting up the experience machine and weighed it against the marginal value of a single M&M. Maybe it would help to stipulate a really smooth interface, like the ability to pop yourself into and out of Experience Machine Headspace with a password-like thought pattern, making the switching cost trivial enough to be overshadowed by a colorful little packet of sugar.
Nov 20, 2023·edited Nov 20, 2023Liked by Paul Bloom
Is your son’s position that doing something you shouldn’t doesn’t make you any worse off? I remember arguing with Dan Hausman about something like this. I didn’t understand it then either.
Great discussion of the experience machine. I had never heard those variants. On your terms, would coming out of an experience machine mean learning that all your relationships, all the people you have ever known about, don’t actually exist? That would be devastating, and undermine my inclination to go back. A shared experience machine is a stronger competitor with a real world. Not sure it wins… but it’s much more real. Would a hedonist see any difference between a solipsistic and a shared experience machine?
1. Good point. Learning that your life has been an illusion must be horrible, and so the hedonist would have to count that as a negative of the EM. (A better example, then, would involve you never leaving the machine.)
2. No, the hedonist sees no difference at all between the 2 machines, since the experience is identical.
That’s the hedonists’ reductio. There seems to me to be a big difference. In co-op mode I can take actions with and for other people. My actions have real consequences. In solo mode there’s just hedonic value.
This accords with good cultural practices. It is better to play co-op than solo. You’re interacting with friends. You’re not just wasting your time.
It’s a version of Shweder’s pajama, the orignal moral dumbfounding example.
Rick’s daughter challenged his proscription against wearing pajamas all day. Sitting around playing video games all day is not as good as sitting around playing video games with friends all day. You may be equally better off playing solo, but it’s not as good.
I’m usually smart enough to avoid interjecting myself into arguments between much smarter people, but it seems to me your son’s argument is a bit circular- he claims experience is all that matters while simultaneously discounting the experience of choosing not to plug into the machine.
This is all very interesting but I don't understand why not signing on for the Experience Machine necessarily, automatically makes one not a hedonist. I feel like I'm pretty solidly a hedonist yet am not tempted by that experiment.
Well, perhaps it's relevant that my idea of pleasure is a lot like Novick's--unlike you I do not find it weird that he cites "writing a great novel, or making a friend, or reading an interesting book" as exemplarily pleasurable experiences. I'm also given to gluttony, lust and sloth (to name a few of my favorite deadly sins), but reading and writing books and friendly human interaction are right up there. I can't really imagine programming a more pleasurable life than I can find by myself. That may be a failure of imagination, but it's what I've got to work with.
well, if your life is at _maximum_ pleasure -- no boredom, physical pain, etc. -- then the Experience Machine really would be wasted on you. (But, as a hedonist, you're definitely going to want it when you get older and your body starts to hurt.)
For me, it’s seems arbitrary to assume any “real” experiences at all, in the sense of (a) representing ground zero in an otherwise potentially endless regress, and (b) being imbued with some “extra” quality (authenticity) to be valued in itself.
The only reason I can see for preferring one “reality” (source of sensory input and output) over another is how the choice impacts other experiencers. I would like my choice to be compatible with as many positive experiences as possible, for as many experiencers as possible. I suspect one reason many people distrust the EM is that it is perceived as an egoistic and/or lonely choice.
But as Chalmers argues in “Reality+” it could, at least in principle, be the more altruistic and sociable choice.
As to the poll at the end, my intuition is that I would want to know if my existence made a positive difference to others, or if the simulation just made me experience a “world” in which I did.
Zach is right. We do not directly react to what happens to us, only to how our mind feels about what happens to us. In a way, we are already in a simulation because we derive pleasure or misery through how our mind interprets our experience. Which is why, as you know, the same exact sensory input can create very different emotional states depending on the sensory input that preceded it. Same stuff, different interpretation by our minds.
Well-described here https://bigthink.com/perception-box/can-you-trust-your-own-brain/
2023: Who knew real life was not worth an M&M!
1823: “Who knew this life was not worth a potato.” (Lord Byron, Don Juan: Canto VII iv. 67)
Don’t assume that those who choose the experience machine are hedonists. Maybe their lives suck and they just want something more than despair to look forward to. Also, I have been through the experience of being in a relationship and realizing years later the person never loved me. Not even sure they liked me. It not only changes the experience but it changes everything afterwards. Sometimes, ignorance is bliss.
I wonder if the allure of "living in reality" is analogous to that of dualism. If there's nothing "special" about the human experience and everything reduces to brain states, then maybe there's nothing "special" about "reality" and everything reduces to experiences (as Zach seems to argue). In other words, perhaps to people who would enter the Experience Machine, arguments for not entering sound similar to arguments for dualism.
I don't think this scenario is that far away. We can already read and control minds, so I guess the machine is just around the corner. My fear is that sadist will put me in it and set it in maximum torture mode. Hell. I don't think this is far into the furture either. Imagine the joy sadistic people can wreak on others if they get their hands on a machine like that. When we can control the pain centers, and not just physical but emotional pain.
The idea of an Experience Machine hell was nicely explored by Ian Banks in "Surface Detail":
https://en.wikipedia.org/wiki/Surface_Detail
Zach's friend here 👋 He has tormented me with these arguments for many moons.
He's right that the imagination exercise is flawed. You cannot appropriately evaluate the thought experiment because you cannot imagine yourself in an unreal situation <where what we are evaluating is precisely the fact that it is unreal> without knowing that it is unreal. The whole point is that you don't know it is unreal--this makes it impossible to imagine.
This does nothing to support Zach's actual view, namely: "If something is worse for me, it must feel worse for me. It can’t be worse if it makes no psychological difference." All it shows is that the experience machine is a poor thought experiment for evaluating the principle.
Here is the argument I've levied to Zach before. It's quite simple, really.
1. It's good to get what I want. (perhaps not *all* desires, but all things being equal, the vast majority of them)
2. I want plenty of things which are not accomplished by my experiencing them. I hope, for instance, that Zach has a good week. I hope that my family complete and publish my masterwork after my death.
3. Therefore, It is good for me that things to happen though I do not experiencing them, since they accomplish my desires. It is better, for instance, that my life's work gets accomplished, rather than ruined, after I die.
Zach might respond to this, as Hobbes did, that I want the pleasure of the experience of knowing that he had a good week. But if that were what I was really after, I would seek out evidence to confirm that he had a good week. If that were true, I would ask him how he is doing more often.
The M&M question is interesting, but I would prefer the simulation M&Ms, though I would not enter the experience machine. The pleasure of M&Ms has nothing to do, for me, with the conviction that it is real. The pleasures whose reality matter, I think, are those that are founded on shared experiences. For instance, for many, sex is one of those things. When A's pleasure is founded on a certain representation of how the world is, say, that through her actions B is feeling pleasure, then, if that representation is mistaken, the pleasure is misfounded, and I am deluded. If you only care about experiences like Zach, the delusion doesn't matter. To me, like Paul it matters immensely. Deluded pleasure can be worse than pain.
Hopefully this offers readers some resources to resist Zach's sophistry.
Yea, so—I agree with (1) that it's good for you to get what you want, but this is instrumental. Getting what you want often makes your life (that is, your experiences) better. It can also make your life worse, though, and it can—trivially—make no difference at all.
Suppose you had a very strong preference that the number of grains of sand on the planet be even, rather than odd. That's either satisfied or it isn't, but I don't think the quality of your life hinges on it *at all*.
Which leads me to (2). I don't doubt that you do want these things (which is very sweet of you) but I think those desires are rational insofar as they can influence your experiences and irrational otherwise. On your view, your answer to "how is your life going?" will always be "I have no idea, because I'm missing so much critical data about whether or not my desires are fulfilled. Are my loved ones actually happy? Do they actually love me? How many grains of sand are there?" etc.
And even death does not set you free! The quality of your life is still in flux, your desires still being satisfied or unsatisfied based on the way you are remembered, the lives of those you lead behind, the progress of philosophy, and etc. etc.
So it is with tenderness and mercy that I reject (3), setting you free from such a totalitarian model of 'wellbeing'.
It's true that some desires don't seem matter, and it's a challenge for desire-satisfaction theories of well-being (or any theory of well-being that involves desire-satisfaction as one measure, i.e., any plausible one) to distinguish between relevant and irrelevant desires for assessing well-being. A plausible suggestion, for instance, is that the desires that involve me are the only ones that count. Aristotle classically involves the fate of your families and friends as relevant also. Regardless, the desires that involve me extend beyond the desires that involve my experiences. I want, for example, my masterwork to be published, regardless of whether I am there to see it come to fruition.
You wish restrict the 'rational' set of desires to ones about my experience. On what grounds, other than simply assuming the principle you are supposed to be arguing for? The evidence on my side is that people pursue and care about things beyond their experience, and I take what matters *to* them as a reliable guide of what matters *for* them. I've no idea why you argue that an unrestricted desire-satisfaction theory is 'totalitarian'--please note that you are the one excluding from rationality a vast array of human desires.
Finally, your point about having no idea whether my life is good is deeply misguided. It is perfectly compatible with justified belief to infer on the basis of inconclusive evidence. I can think my life is going well based on the evidence I have, knowing that I might be mistaken about that. Saying that I am entirely ignorant because I might be wrong is just a manifestation of general radical skepticism. As for whether death might set you free, the vast majority of theorists I've encountered on the subject find nothing strange in the idea that your life might go better or worse after you die. The very first treatment of this question by Aristotle argues decisively in favour.
I've said this a great many times Zach, but it just seems to me that *you* care exclusively about your experiences. That is no grounds for arguing that everyone else *should* care only about their experiences. I've never heard you offer a single argument in favour of that normative claim. And the empirical claim that people in fact only care about their experiences is false, as I've argued re: my caring about your week.
I think self-reporting without having actually gone through the experience of the Experience Machine is potentially unreliable. If I imagine the experience of being in it as completely seamless and reliable I don’t see how one could turn it down. This is our one life. Maximizing the happiness, stimulation, peace and contentment (or whatever positive feelings one considers most valuable) just clearly is the obvious choice. If a feeling of inauthenticity is detracting from one’s experience in the EM the technicians will tweak the settings and eliminate that issue.
Great post. I think the reverse set-up is an important twist on the original thought experiment.
I am stunned that anyone would even consider plugging in. Maybe I have trust issues but I would worry about the machine breaking down or someone replugging me into a pain machine just for fun. I don’t think I’m paranoid but allowing anyone or anything to control my experience of life is just unthinkable.
There is, of course, an extreme scenario where your current life is unbearably painful but even then there is always the option to terminate it. And who is to say that your afterlife won’t be pure pleasure?
I’m on team Zach. We can’t know that the reality we are in right now is *really* reality.
Lately I’ve been going to sleep with a nicotine pouch in my mouth and it gives me the most vivid dreams. They feel very real in part because they’re not especially bizarre. They’re all plausible situations from real life.
It wouldn’t shock me to find that the dream is me typing this comment and I will wake up shortly.
I wouldn't mind being a hedonist but I like a combination of random material circumstance bouncing around this thing I name me.
but would you be satisfied with the _experience_ of random material circumstance bouncing around this thing you name you?
No,the thing I name me most of all enjoys the few moments when it feels really and truly in control.
but you'll get the same enjoyment if you're in the machine right? (in fact, you can't be sure right now that you're not in the machine.)
It is true- I will never know if I am in or out of the machine or if a machine exists. Given a choice,though,so I theoretically might know-I cling to the idea of individual minds functioning within mostly random material circumstance and occasionally creating new ideas .
I like the purity of the M&M thought experiment, but when I answered the question I simulated the probable annoyance of booting up the experience machine and weighed it against the marginal value of a single M&M. Maybe it would help to stipulate a really smooth interface, like the ability to pop yourself into and out of Experience Machine Headspace with a password-like thought pattern, making the switching cost trivial enough to be overshadowed by a colorful little packet of sugar.
good point -- the experiment doesn't factor in the inconvenience of plugging in
Is your son’s position that doing something you shouldn’t doesn’t make you any worse off? I remember arguing with Dan Hausman about something like this. I didn’t understand it then either.
Great discussion of the experience machine. I had never heard those variants. On your terms, would coming out of an experience machine mean learning that all your relationships, all the people you have ever known about, don’t actually exist? That would be devastating, and undermine my inclination to go back. A shared experience machine is a stronger competitor with a real world. Not sure it wins… but it’s much more real. Would a hedonist see any difference between a solipsistic and a shared experience machine?
1. Good point. Learning that your life has been an illusion must be horrible, and so the hedonist would have to count that as a negative of the EM. (A better example, then, would involve you never leaving the machine.)
2. No, the hedonist sees no difference at all between the 2 machines, since the experience is identical.
That’s the hedonists’ reductio. There seems to me to be a big difference. In co-op mode I can take actions with and for other people. My actions have real consequences. In solo mode there’s just hedonic value.
This accords with good cultural practices. It is better to play co-op than solo. You’re interacting with friends. You’re not just wasting your time.
It’s a version of Shweder’s pajama, the orignal moral dumbfounding example.
Rick’s daughter challenged his proscription against wearing pajamas all day. Sitting around playing video games all day is not as good as sitting around playing video games with friends all day. You may be equally better off playing solo, but it’s not as good.
I’m usually smart enough to avoid interjecting myself into arguments between much smarter people, but it seems to me your son’s argument is a bit circular- he claims experience is all that matters while simultaneously discounting the experience of choosing not to plug into the machine.
This is all very interesting but I don't understand why not signing on for the Experience Machine necessarily, automatically makes one not a hedonist. I feel like I'm pretty solidly a hedonist yet am not tempted by that experiment.
Hmmm. If you're a hedonist, why aren't you tempted by a life that has a lot more pleasure than you have now?
Well, perhaps it's relevant that my idea of pleasure is a lot like Novick's--unlike you I do not find it weird that he cites "writing a great novel, or making a friend, or reading an interesting book" as exemplarily pleasurable experiences. I'm also given to gluttony, lust and sloth (to name a few of my favorite deadly sins), but reading and writing books and friendly human interaction are right up there. I can't really imagine programming a more pleasurable life than I can find by myself. That may be a failure of imagination, but it's what I've got to work with.
well, if your life is at _maximum_ pleasure -- no boredom, physical pain, etc. -- then the Experience Machine really would be wasted on you. (But, as a hedonist, you're definitely going to want it when you get older and your body starts to hurt.)
We'll see.
For me, it’s seems arbitrary to assume any “real” experiences at all, in the sense of (a) representing ground zero in an otherwise potentially endless regress, and (b) being imbued with some “extra” quality (authenticity) to be valued in itself.
The only reason I can see for preferring one “reality” (source of sensory input and output) over another is how the choice impacts other experiencers. I would like my choice to be compatible with as many positive experiences as possible, for as many experiencers as possible. I suspect one reason many people distrust the EM is that it is perceived as an egoistic and/or lonely choice.
But as Chalmers argues in “Reality+” it could, at least in principle, be the more altruistic and sociable choice.
As to the poll at the end, my intuition is that I would want to know if my existence made a positive difference to others, or if the simulation just made me experience a “world” in which I did.