Survey Results: Suffering Vs. Oblivion

[Content warning: suffering, oblivion]

Every so often, someone on Reddit realizes that about half of people wipe themselves with toilet paper sitting down, and the other half do it standing up. This discovery is followed by horror on both sides that other people do it differently.

I occasionally have the same feeling when I talk about ethics. Every so often I run up against a base-level clash of intuitions with somebody else, where they disagree on a preference I would have expected to be self-evident. This is pretty bad, since some forms of consequentialism are a lot more elegant if we imagine that all apparent moral disagreements are just people failing to think through their own preferences clearly enough and everyone really agrees about morality deep down.

A lot of these clashes of intuitions have to do with the idea of suffering versus oblivion. In order to explore this further, I asked people on Twitter and Tumblr to answer some survey questions like the following:

1. Would you rather:

Option A: Live the rest of your life working 16 hour days, seven days a week, as a McDonalds cashier. You will have no time off except the time you need to eat, sleep, and use the restroom. Your entire life will be spent doing McDonalds cashier related tasks. When you are no longer able to perform your tasks, you will die painlessly.

Option B: Die painlessly right now.

2. Would you rather:

Option A: Live a long but unhappy life. You live to be 120. You spend most of your time unhappy, but you are not actively suicidal.

Option B: Live a short but happy life. You die ten years from now, in 2026, after being hit by a car. Until then, you do fulfilling work, have happy relationships, and meet with success in most projects.

3. Would you prefer:

Option A: A world with 100 trillion trillion sentient beings, all of whom are miserable, but not quite so miserable that they wish they were never born.

Option B: A world with 1 million sentient beings, all of whom are happy and consider their world a utopia.

4. What percent certainty of going to Heaven would you need before you would prefer a world with both Heaven and Hell to a world where death ends inevitably in oblivion?

Before checking the results, I had five hypotheses.

First, people would be split in their answers to these questions, with strong feelings on both sides.

Second, answers to all questions would correlate along a general factor of oblivion-preference versus suffering-preference. That is, people who would prefer oblivion to working at McDonalds would also be more likely to prefer a short life of happiness to a long life of unhappiness, et cetera.

Third, this factor would predict whether somebody endorsed a form of population ethics which promotes creating new people (question 3 is sort of just asking this already, but I also included a more direct question along those lines).

Fourth, this factor would predict some real-world consequences like whether people believed in a right to euthanasia and whether they were signed up for cryonics.

Fifth, happier people would be more likely to prefer suffering over oblivion, because they view life as generally excellent and so oblivion represents more of a sacrifice for them.

1090 people took the survey (social media is a wonderful thing). The survey specifically asked that you not think about other people or the question’s effects on them and just choose whatever made you selfishly happier. Several people complained to me that the concept of “generally unhappy, but does not wish they were never born” didn’t make sense to them, which I guess is data in and of itself. The results were:

The first hypothesis was confirmed:

On the fourth question, people answered everything, from demands of certain salvation (n = 320), to being okay with certain damnation (n = 69), and everything in between.

The second hypothesis was weakly confirmed:

Orange results are significant at p < 0.05, red results at p < 0.01. We did not adjust for multiple comparisons.

There are significant correlations between most of the questions, but they are not very strong. When I limited the analysis to the people who felt most strongly about their answers to the questions, the correlations went up a bit:

The third hypothesis was not supported.

I asked people a pretty direct question about the ethics of creating new people:

5. Creating new sentient beings is:

Option A: Exactly comparable to improving the lives of existing beings. Creating a life that experiences 100 utils is exactly as good as improving existing lives 100 utils.

Option B: Generally good, but less good than improving the lives of existing beings by the same amount

Option C: Morally neutral

Option D: Bad

Option E: I am not a consequentialist or otherwise don’t want to answer this question

The results were:

The only correlation with any other question was with the one hundred trillion trillion sentients question, which is basically asking the same thing in different words. The correlation with all the other questions was not significant and in fact very close to zero.

The fourth and fifth hypotheses were weakly supported.

I gathered a general score of people’s pro-oblivion or pro-suffering bias based on their answers to the first three questions and how strongly they felt about them. It predicted the following:

The specific questions here were:

Whether you believe everybody has a right to commit suicide if they want, including people who are not terminally ill.

Whether you believe people who are terminally ill should have a right to suicide, ie traditional euthanasia.

Whether you are interested in signing up for cryonics.

And whether you consider yourself a happy person.

All were correlated with preference for oblivion over suffering in the expected direction, but not very strongly.

So it looks like people have very different opinions about when to choose death versus suffering, but that these opinions are inconsistent and only weakly driven by broad cross-situation intuitions.

You can see the survey here, but please don’t take it since I’m done getting data. You can download anonymized results here (.xlsx, .csv)

This entry was posted in Uncategorized and tagged , , . Bookmark the permalink.

510 Responses to Survey Results: Suffering Vs. Oblivion

  1. Alex Godofsky says:

    When I took this I found myself spending a lot of time looking at the second problem in particular and thinking “what counts as ‘happy’ or ‘unhappy’?” I didn’t feel like it was a very good question because my answer ended up being entirely predicated on what I decided probably counted as “unhappy”.

    I would have also appreciated an option on the third question analogous to on the fourth (“I refuse to answer this”); I left it blank.

    • Vladimir says:

      I agree with this. The second question wasn’t as clear in the tradeoff as the mcdonalds question, so I ended up choosing the longer life despite choosing “die immediately” for #1. Also for the “creating happy sentient beings” question, all I kept thinking of is Eliezer’s argument that it might not be important to make more unique people if you live in a Big Universe, so that tilted me slightly in favour valuing existing people.

    • Mammon says:

      +1. IMHO a lot of these fell in the “not a good question” category, because depending on how I interpreted the question I could go either way.

      For example, the heaven/hell question depends on your personal conception of heaven and hell.

      • JayT says:

        I felt the same way about the heaven/hell one. There are some depictions of hell that I think seem like they would be fairly entertaining and I would only need like a 60/40 split. Then there are the really bad depictions where I would need basically a 100% chance of going to heaven before I agreed to it.

        • Devilbunny says:

          I’m an atheist and thus not vulnerable to definitions of hell that amount to “permanent separation from God”, but isn’t the usual understanding something like “eternal physical and/or mental torment”? Like, being tortured for eternity? Spending a few trillion years being burned, then a few trillion being frozen, and then a few trillion of both at the same time, just as a warmup to the whole eternity thing, where they really start to make it rough?

    • Anon says:

      Fourthed. Reading that question I was thinking, “am I even happy right now?” There are plenty of things I’m hoping to do with my life, but ‘be maximally happy all the time’ is definitely not at the top of the list.

      Is my 120-year-old unhappy life comparable to the scenario from the first question (unending drudgery with no hope of respite)? Or is it just a life where struggle and disappointment are the norm and maybe happiness and success will peak through the clouds every once in a while (AKA normal real life)?

      • sconn says:

        I’d sign up for the unhappy long life if I could also accomplish things. You know, if I cured cancer or wrote the great American novel or something. But not if it included also not getting to do the things I want.

        The big thing that I could feel swaying me emotionally is the thought of ten more happy years. I like the thought of ten more happy years. But at the end of those years …. I would be more likely to want to switch answers. I’m not ready to die NOW. I’m never ready to to die NOW. But I’m already accustomed to the idea that I will only get a certain number of years, and reading the question I find I’m kind of insensitive to the number.

    • HHELLD says:

      I would say all the questions are not anyhow precise, resulting in wide interpretations, resulting in upper-capped correlation.

    • Lysenko says:

      My conception of the world is pretty much summed up by the opening to the prose version of the Hitch-Hiker’s Guide (you know the one):

      This planet has – or rather had – a problem, which was this: most of the people living on it were unhappy for pretty much of the time.

      Mild to moderate unhappiness is pretty commonplace, and people bear up under it with varying degrees of success, much the same way many people bear up under chronic aches and pains with varying degrees of success.

      For my part, I haven’t dipped into real depression in some time now and hopefully I won’t anytime soon, but I’ve also come to terms with the fact that a certain degree of unhappiness is probably just…there. Again, rather like long-lasting minor to moderate pains: as long as you can successfully divert your brain from them: Interesting intellectual stimulation, easy sex, drug/alcohol/fats and sugars abuse (but I repeat myself), guided meditation, friends and family, whatever. People choose different coping mechanisms that are to varying degrees helpful or self-destructive, and when the unhappiness gets too severe some percentage have those mechanisms fail and decide that death is preferable.

      So why wouldn’t I choose a long, generally unhappy life? As far as I can tell, so has the overwhelming majority of the human race for the overwhelming majority of history.

      Those blessed by the right combination of circumstance, intelligence and cognitive skills fitted to those circumstances, and brain chemistry to go through life with ‘happy and content’ as a default state strike me as more alien and distant to me than the richest of multi-billionaires. Hell, half of THEM are pretty damn unhappy, apparently.

      • Forge the Sky says:

        I’ve come to think that people are divided, more than anything, by two metrics: baseline anxiousness and baseline happiness.

        The sorts of things people do and the sorts of people they become are mostly due to that.

        But maybe I’m biased by my experiences. I used to be high-ish anxiety and average happiness, now I’m low-ish anxiety and high happiness. The world just seems really cool and it’s awesome being me in it. Even though I’m not particularly living a golden life; it’s really just perspective.

        So I guess you’d find me pretty weird, based on what you say. People often do. But people also tend to like being around me. Happy passion is magnetic.

        Also, it works wonders for a straight man’s love life 😉

        • Koldos the Shepherd says:

          Care to elaborate what shifted your baselines?

          • Forge the Sky says:

            Sure, I’ll try my best. I’m a bit short on time at the moment but check back tomorrow.

          • Forge the Sky says:

            I have a strong interest in helping other people become happier and more dynamic. The difference it makes is remarkable. But the process is long, difficult, and kinda hard to describe, so unfortunately there’s no one-size-fits-all solution. If only it were as simple as ‘take this supplement and exercise.’

            That said, I think that being in good health is a great start and being in ill health will make this a much more uphill battle, so get that sorted if you can.

            This is gonna be a bit philosophical for a bit, but I think it’ll come around to making sense.

            The bulk of the work is in habituation, mindset, and (for lack of a better term) ego discipline. Basically, I started with this – a determination to accept the truth, whatever it might be, however little I may like it. In spite of that, I had quite a few illusions about reality and the world – things that you use to create an identity, basically. For example, I was VERY idealistic about romantic relationships – find the ONE, stick with her forever, make it work no matter what by heroic force of will kinda stuff.

            So when I was confronted with a romantic relationship that WAS falling apart in spite of all that, I had to rip all that up and start over. It was like a part of me – a central ‘I’ – looked at the naive, idealistic boy I was, and said – you’re good. You’re nice. I like you. But you’re not true, so you have to go. You’re gonna fight like hell to stay, but I’m going to win.

            And as I got to work I was surprised by what I found – my head was just FULL of ghosts like this. Ones that told me that I needed to use big words to impress people, cause it’s really important to seem smart. Ones that made me feel guilty for wanting things. Ones that told me I shouldn’t talk to strangers because that was scary. Ones that did all sorts of things like that. And when you tried to remove them – by which I mean, ignore the impulses and desires they create and instead do what you genuinely want to in the moment (rather than what you’re conditioned to do or ‘ought’ to do) – I found out that it hurt like hell. I thought that was strange. The model I used to explain it (psychodynamic handwaving it might be, but it worked to model the situation) is that all these things were defending my ego – because there was, fundamentally, a sense that I wasn’t enough/worthy/good if I didn’t listen to all the ghosts and do as they say. And so I was being twisted around by them without even knowing it, living in imaginary headspaces, having imaginary preferences, doing things based upon imaginary systems of risk and reward that correlated poorly with ACTUAL risk and reward….little wonder I wasn’t more than basically content. I didn’t even know what I ACTUALLY liked to do or be.

            So I began steadily trying to keep myself aware of what my impulses were in the moment. And when I found myself recoiling from that impulse, I’d chase down why and I’d find fear, or anger, or pain hiding under the surface – and I’d do my best to amplify that negative emotion, feel it fully, so I didn’t have to hide it any more. Get it out. And then I’d just DO.

            A few of these things still pop up from time to time, but now it’s mostly easy and infrequent. And when you strip away all the old ghosts and stories you told yourself about yourself, you can kinda just choose what attitude you’re gonna have at your core. I just kept a central attitude through all this that I’m awesome just for existing and I’m gonna win against obstacles, and that just became my default. There’s nothing egotistical about it – I don’t think I’m better than anyone and I don’t need to prove anything to anyone, I’m just…enough.

            See, the thing now is – I’m not trying to tell stories about my actions and preferences and past and so on. I just AM what I am, and when I tell stories about who I am or why I’m doing what I’m doing I do it mostly for fun or utility. Just for fun I basically started telling myself an ‘identity story’ that I’m just godlike in intellect and charisma, and people love it. Because they can tell I’m not taking it too seriously. But it does make me project confidence and charisma, and when some super-cute girl starts pushing back and trying to topple me, and then we go back and forth until she succeeds and then we both just laugh about it cause I don’t care, it’s just a story, well….

            Things tend to go well 😉

            The other thing I did (this is gonna be a long post sorry) is I forced myself to do a lot of things. Just go on a random road trip, see the place and talk to the people, spur of the moment. Go to clubs and dance. Make out with a girl I don’t know. Go to a bar and talk to the dudes in the athletic jerseys. Play an instrument. Train for an athletic competition. Make a bunch of videos for YouTube. Whatever. If it kinda scares you so much the better.

            What you find if you pay attention is this – you will actually end up liking some things you couldn’t have predicted. And some stuff you thought you liked was, when you’re honest with yourself, just something you were doing to get praise/self-esteem, or that you were using as a buffer to NOT do other things. A reading hobby can actually be a pretend-I’m-smarter-than-everyone hobby, or a I’m-scared-to-socialize hobby.

            Eventually you rewire yourself. And in the process you got in lots of circumstances that make you less afraid of new things and more competent in general. The anxiety can be just another ghost.

            One final thing – I have struggled with biologically-created anxiety as well. This doesn’t respond the same way precisely; I’m better at managing it when it happens but I can’t just will it away. But I’m lucky enough to have found what triggers it for me and can generally avoid it (I mentioned it in a previous thread; basically my immune system reacts strongly to barley and one of the symptoms of my accidentally eating it is anxiety. Biology is weird).

            And who knows, this could all just be false attribution to something that wasn’t actually causative. It made a huge difference for me, in the end. Not much can bother you once the roots of your existence have been turned anyways.

            I don’t know how replicable all this is. But I know a few more people like me, and a common thread seems to be reaching some sort of breaking point where all your ego defenses have failed. But maybe such times are actually common, and most just don’t react in the proper way to grow from it, and just make new ego investments instead.

            Forge on.

          • Koldos the Shepherd says:

            Thank you for the extensive reply. I found it very interesting that the same “determination to accept the truth, whatever it might be, however little I may like it” can lead some people to improve their baseline happiness while leading others to descend into limitless despair because all meaning is imaginary etc.

          • Forge the Sky says:

            Dunno if you’re still keeping track, but

            The way I think of it is, when you look at things as they are and perceive that there’s no intrinsic meaning to things, that’s nihilism. And it has two possible responses. You can despair. Or you can realize how awesome that is, because it means you can give your own meaning to everything.

            Nihilism can make little gods of men, if they only have the wherewithal to accept that degree of freedom and responsibility.

    • neonwattagelimit says:

      Another agreement here about question #2. Although maybe it was deliberately vague? For #1, the answer was clear: I’d rather die painlessly right now than do nothing but be a McDonald’s cashier for the rest of my life.

      But “spend most of my time unhappy” until 120? How unhappy am I? Am I still able to do things? Are there still things I take pleasure in? Do I experience occasional moments of happiness or contentment? If the answer to all these questions is “yes” – and the wording of the question makes me suspect it is, though I’m not sure – then I’d pick living to 120 over dying in ten years.

      I agree with anon that ‘be maximally happy all the time’ isn’t a great goal. I spend most of my life in a mood best described as “vaguely content neutrality.” While I am prone to occasional bouts of moderate – though still functional – depression, most of the time I am neither particularly happy nor particularly unhappy. This is fine, far as I’m concerned. I spend more time thinking about what I do and how I behave – whether I’m being a good person, living up to my ethical standards, doings that are productive and/or interesting and/or fun – than whether I actively feel happy (so long as I’m not depressed, that is).

      • J Mann says:

        I went the other way – I’d rather be a McDonald’s cashier, because at least I could talk to the customers for short periods, and think about stuff. I’d take the chance at being able to make it a worthwhile experience.

        • sconn says:

          Yeah, if I wasn’t actually depressed, just busy flipping burgers, I could totally make my own happiness out of that. You can be happy almost anywhere if your psyche will let you. (The opposite is also true.)

        • N8Greene says:

          Agreed. My first impulse was to just die, but after some thought I decided I would prefer working at McDonald’s. At any time I could stop working and be euthanized, so why not give it a try? I would still be able to: talk to (and maybe help) co workers, get to know some funny regulars, listen to music while cleaning up, masturbate (or even have real sex), ride my bike to work, eat food I like.

          Granted, the schedule is grueling and my body might not last long, but who knows? Read the stoics and try to make your own attitude.

    • Eve Matteo says:

      The survey did specifically ask for your own selfish opinions, and not how it affects other people. Therefore, what counts as “happy” would be what counts as happy to you, and what counts as “unhappy” would be what counts as unhappy to you.

      My issue with the question is, the survey is clearly geared towards Scott’s readership (one question using Utils shows me that he didn’t really expect it to be taken by anyone outside this community), yet he used the unprecise wording “mostly”. Mostly could be anything over 50%. At 60% unhappiness, I’d still take the longer life, but at 95% unhappiness, I’d take the early death. I’m not quite sure where the split is for me. I’m thinking finding out where that split is might have gotten some more interesting results.

  2. dinofs says:

    What kinds of people took the survey? Because if you’re looking for people’s moral intuitions, asking mostly the kinds of people who would know what “utils” are would likely get you a skewed response. I wonder how much the answers of people who are not familiar with ethical philosophy (or at least not familiar with consequentalism, or rationalism) would differ from those of people who’ve spent more time thinking about the kinds of questions you’re asking.

    • satanistgoblin says:

      What kinds of people took the survey?

      Twitter followers and subredditors of slatestarcodex.

    • Samedi says:

      I didn’t know what a util was so I had to Google it. Apparently it is an arbitrary unit of measure, so not really a unit of measure at all. I don’t understand how such concept is consistent with the rationalist or scientific program. Can someone enlighten me as to the usefulness of a unit of measure which is not empirically verifiable? It strikes me as highly problematic.

      Scott often writes about the flaws in social science methodology. I would think that a failure to conform to the standards of empirical science is among them. How can social scientists do good science if their terms are not empirically verifiable?

      • Jaskologist says:

        It’s not arbitrary so much as hypothetical. We know that, say, eating pizza is more good than eating poop, so in theory we should be able to measure the amount of good for each. Once we can quantify it, we should then be able to perform calculations on it, and usher in the utilitarian utopia.

        The fact that we haven’t come up with a way to actually measure utils is but one of many problems for utilitarians.

        • Samedi says:

          I see your point. I suppose my question is, what is actually being measured? I read “more good than” as “prefers”. So, any measurement would have to measure the strength of one person’s preference for one thing over another. That should be, in theory at least, measurable perhaps with some kind of neurological imaging. But at the end you still only have the data point that one person prefers one thing over another, or even many people prefer one thing over another. That doesn’t seem very useful.

          If we discard legacy metaphysical terms like “good” and “bad” then we can think of ethical preferences in a practical way. If someone holds a preference I find objectionable I can attempt to persuade them otherwise. (Here we are in the realm of rhetoric rather than science.) I can also reward preferences (especially when expressed by behavior) I like or punish preferences I don’t like. And this, I would argue, is exactly what people do.

  3. Alraune says:

    Why was there no option for 5. of believing that creating new sentient beings is superior to equivalently increasing the happiness of existing ones?

    • Scott Alexander says:

      Does anyone really believe this?

      • I do. I had to take option E, and I was disappointed that you didn’t even offer me my real choice.

        • Scott Alexander says:

          Well, I guess my survey about how other people have ethical preferences I wouldn’t expect failed because someone had an ethical preference I didn’t expect. How could I have possibly seen this coming?

          • Gbdub says:

            Consider the “hundred trillion sentients” question – how do 11% of people choose the hundred trillion option, unless at least somebody feels like “more lives is a good in itself”?

            I’m honestly surprised you included the 100 trillion sentients question if you thought no one assigned utility to new life.

            I’m also perhaps irrationally annoyed by “asymmetrical” survey questions even if I think no one will select one option – usually you get a range of options from worse -> neutral -> better, but you left off the whole “better” end!

          • wintermute92 says:

            This does seem like another endorsement of the “full coverage wherever possible” principle for survey design. Non-continuous opinions can be hard this way (e.g. “I am not a consequentialist”), but as a general rule it’s good to deal in every part of a continuum (i.e. ask all of “better, the same, worse”).

            I’m not 100% sure it would have caught this case, but it certainly seems relevant.

        • Decius says:

          Implying that a world tiled with sentient beings would be optimal?

          • Gbdub says:

            Why must utility be linear with population? If there are only 100,000 people, adding another is a huge relative increase in utility (otherwise people may go extinct, and then no one is left to produce utility). If there are already a trillion people, adding an extra is less meaningful. A stable maximum is certainly possible.

        • Mike H says:

          Thought the same thing, but I’m not a utilitarian or consequentialist, though both types of reasoning find their way into my own.

        • Kyle Strand says:

          Serious question: could you please try to explain your reasoning and/or intuition for this? Like Scott, I’m somewhat confused as to how this belief could even arise.

          • Let’s say you like birds. Would you like a world that contains one billion thrushes, or a world that contains one billion total birds split between thrushes, nightingales, cuckoos, magpies, herons, hawks, toucans, parakeets, storks, albatrosses, and blue-footed boobies?

            Let me be less indirect.

            Happiness is not fungible between humans, in that no two humans ever experience exactly the same happiness. Jill!happiness and John!happiness may belong to the same genus, but they are different species. They may be similar enough that you can aggregate them, but that aggregation is lossy because it necessarily ignores the differences between Jill and John’s experience. Therefore, given the choice between giving Jack and Jill both an extra 25 utils, or creating Jack with 50 utils of his own, I choose the latter. In either case we’ve increased the total utils, but in the second case we’ve created a brand-new kind of happiness, namely Jack!happiness. This is a good thing.

          • Aaron Woodside says:

            I’m not an advocate of the view necessarily, but the first justification that came to mind goes something like this:

            As far as I can tell people experience “utils” in a range, nobody is approaching infinity. Say your population of sentient beings is normally distributed around some mean of experienced utility, lets use 100. You have the option of either creating a new being with 50 utils or taking an existing one and bumping it from 100 to 150. In the former case you’ve created a new entity with lots of upside potential. In the latter case you’ve pushed an existing entity to a possibly unstable extreme. By creating new beings you improve the theoretical maximum utility in addition to improving the average.

          • denis says:

            “in addition to improving the average.”

            I think you mean the sum, because the average is decreasing.

        • hlynkacg says:

          Because the alternative is extinction, not just of one’s self but of one’s species.

          • LaochCailiuil says:

            This is interesting to me, if somewhat odd, the idea that species just existing is more important than species or population of sentients existing that actually want to be around. Is this what you mean?

          • Anonymous says:

            Learn to overcome the crass demands of flesh and bone, for they warp the matrix through which we perceive the world. Extend your awareness outward, beyond the self of body, to embrace the self of group and the self of humanity. The goals of the group and the greater race are transcendent, and to embrace them is to achieve enlightenment.

          • LaochCailiuil says:

            Anonymous I understand the sentiment, in other words the group and it’s survival should be more important to me than myself? That’s a difficult one to swallow I must admit.

          • MugaSofer says:

            Firstly, humans do usually want to be alive, so even the most hardcore bite-all-the-bullets preference utilitarian should not want the human race to go extinct.

            Secondly, yeah, I at least am sad about e.g. the extinction of random animal species – just like I would be sad if a library full of unique books was burned. This would apply to humanity even if there are other sophonts of equivalent or greater worth in existence, and all other concerns about humanity’s extinction had been cancelled out.

            Thirdly, of course, there are questions of human values getting to spread across the universe, all that potential snuffed out etc.

            Any value matters more to me than myself in sufficient amounts, so that isn’t an issue. Do you … not?

          • LaochCailiuil says:

            I’ve neurosis(I think that’s the term) around my existence and my feelings about society at large and it’s existence. I seriously don’t want to come off trolly, however overtime it feels like I’ve lost good feelings about humanity existing in and of itself. I’m not saying this is correct or good or etc. it’s just how I feel. The weird thing is how it’s tied up with my identity and how I’m not sure I’d want to go back to the good feeling because I’d feel like I’d lose my identity. I think this comes off of a lot of social isolation and depression.

          • Gbdub says:

            Isn’t the simple utilitarian answer that, if we are extinct, no one is left to produce more utils?

          • Koldos the Shepherd says:

            Any definition of utility worth its salt should be transhumanist, so I don’t think we need to special-case human survival.

            We got intelligence once, it can probably happen again. We got life once, it can probably happen again.

      • Yes. I don’t but for example I’m fairly sure Lucidian does. (

      • J says:

        My solution to the problem of population ethics is to view humanity as a single organism. So my choices about creating new people have to do with whether they increase or decrease humanity’s overall chances of healthy survival.

        Thus, if the choice is about spreading a billion new lives across the galaxy so that we’re no longer dependent on earth for survival, that’s better than allocating their utils + some extra utils to existing people stuck on earth.

        • Koldos the Shepherd says:

          You are fighting the hypothetical. Would your preference change if the million was distributed across the universe in a manner that made them optimally safe?

          • Jiro says:

            Fighting the hypothetical here is valid because the questions have to be interpreted as they are written.

          • J says:

            Scott was surprised that people would think that having fewer people people is inherently good. My point is that either more or fewer can be good, but it all depends on what it means for overall civilization health (and NOT merely on individual happiness). It’s how I handle the famous “repugnant conclusion” that requires assigning an inherent value to higher population. So in my value system I can’t answer the survey question without more context about what it means for the civilization.

        • Vamair says:

          I’d rather see it as a whole of humanity over time. That is, you get a million happy people, but that’s it. No more sentients after that. Or the history continues until you get 100 trillion trillion, and then humanity dies off.

        • chaosmage says:

          I use the same solution, and chose the 1 million utopia because I figured that scenario left more resources for a space program.

      • hlynkacg says:

        Does anyone really believe this?


        In fact, I find it strange that the question even has to be asked.

      • Tom says:


      • Zur says:

        Wasn’t this the point of Asimov’s robot novels?
        The spacers were long lived and had great lives but the earthpeople had a healthier society because since they were having more babies their society was more dynamic.

        • FullMeta_Rationalist says:

          I’ve seen the same meme in High Fantasy novels. I.e. the elves live longer, but humans live more meaningful lives.

          As a counterpoint. I believe the High Elves of the West were implied to have derived their wisdom and equanimaty from their longevity. Their silly songs about the simple things in life reflected a zen-like understanding of the eternal recurrence that only comes with experience.

          • Guy says:

            Kind of off topic, but a friend of mine has an excellent headcanon explaining the lack of elven engineering (as opposed to magic): elves live so long that they have no need to accomplish anything with any sort of speed. If they decide to stand there and yell at a mountain to get it to move, they are perfectly able to keep standing there and yelling until erosion has done the job more thoroughly than any dynamite.

          • Jiro says:

            That reasoning not only requires that the elf doesn’t need X fast, it also requires that the elf doesn’t need to have any of the second or higher order consequences of X happen fast, not even the consequence of “being in a state where X hasn’t happened yet for a long time”. Unless the elf has a terminal goal of moving mountains, this seems implausible.

            Even if the mountain just blocks his view of Venus, why in the world would the elf not care that his view is blocked for a million years, regardless of how long he lives?

          • FullMeta_Rationalist says:

            If elves are slow, then what does that make treants?

          • Jiro says:

            Assuming you mean ents (treants are the Dungeons and Dragons version), they have the same problem. If it really took them days to figure out what to do, what do they do if the forest is burning down?

          • FullMeta_Rationalist says:

            I just remembered the story of a real world elf. And his name was Cliff Young. Adiabatic Processes are the best kind of processes.

      • Jaskologist says:

        I also wanted that option on that question (didn’t manage to finish the survey, fwiw).

        Two people with happy lives are better than one with a life of double happiness.

        (But then, I don’t believe that everything is reducible to a single util measure.)

        • Happybara says:

          That’s interesting. I agree that two happy people are better than one doubly-happy person, but I’d also rather create one happy person than two miserable people.

          So for me, there’s apparently some “happiness cutoff,” above which you can start adding people. I couldn’t quite tell you where that cutoff is though, and figuring it out seems pretty squishy.

          • sconn says:

            It’s a practical question for me, being a parent. Having one more kid is a lot of work and may leave less attention for the kids I have. On the other hand, that one kid would obviously have infinitely more happiness than if we didn’t have him. I definitely think in terms of a cut-off — will we all still be pretty okay with one more kid? Then let’s have another. If it pushes us from “generally happy” to “slightly miserable”, then let’s not.

            Of course it’s difficult to tell in advance how much harder one more kid will be, but a good rule of thumb is that if you’re overwhelmed and unhappy NOW, creating another person is a bad call. If you notice a lot of extra happiness floating around, why not share that with a new person?

        • Skivverus says:

          Alternatively: there’s an effort-to-utility conversion implicit in activity, and it has diminishing returns on increasing both quality and quantity of life: if you’re further along the curve on quality than quantity, you spend your next effort on quantity, and vice versa.

          On this perspective, the trillion^2 miserable lives versus the thousand^2 happy ones depends entirely on the agency of the lives in question to improve or reproduce their own situation after the hypothetical expires.

      • Alex says:

        I assume familarity with the parenthood paradox. In case not, first link that came up:

        So in terms of revealed preferences of parents we are confronted I think with three possible explainations:

        a) Parents are stupid and ignorant of said research. They are making a gut choice. Given the number of parents in the total population we should not be surprised that stupid and ignorant people hold inconsistent opinions.

        b) Parents are rational but assume that their child’s lifetime happiness will more than offset their suffering and altruistically make this choice. My impression is that this is what parents tell themselves and others.

        c) Parents indeed do indeed value creating sentinent live over happiness. Otherwise they would not reproduce. Something something evolution.

        • Anonymous says:

          I don’t see why it can’t be all three, in various proportions, held by different people.

        • Stuart Armstrong says:

          >a) Parents are stupid and ignorant of said research.

          As a parent, I can confirm variants of this. Most parents (including me) had an incredibly poor knowledge of what parenting involves (simple lesson: use the outside view, rather than projecting yourself+child).

          But for many parents, it’s about preferences (eg looking for meaning, doing your duty, taking on a new role) rather than hedonism.

        • Two McMillion says:

          There’s a fourth option: we don’t actually know what happiness is or how to measure it.

          • Anonymous says:


            The best I’ve been able to come up with regarding happiness is that it’s a reward for living well. No progress of any kind regarding quantification, though.

      • Adam says:

        I don’t think these people are taking the hypothetical the same way you would. The answers given are basically they’d prefer two people with identical levels of utility to one with double the utility, which of course means in the first world their utility is also increased because they prefer that world, in which case they’re just saying they prefer a world with more total utility.

        Either that or they’re just valuing diversification as a risk management strategy, effectively saying it is better to own two different stocks valued at $5 each than one valued at $10, all else being equal, as your risk of ruin is less in the former case. A larger population is less likely to be wiped out.

        Neither of these is the same thing as saying it’s an ultimate good to just tile the universe with sentient beings without regard to what extent they actually enjoy their own existence. Even in the second case, at a certain point the further risk decrease from one additional person eventually goes to zero.

        • Jiro says:

          Even in the second case, at a certain point the further risk decrease from one additional person eventually goes to zero.

          That’s actually a major flaw in the questions. Asking people whether they would prefer X to Y doesn’t necessarily mean that they subscribe to a theory which leads them to preferring unlimited amounts of X to Y. Even if the question doesn’t put a number on it, people will answer as though they were asked about a finite quantity of X. If you want to know whether they would prefer X to an unlimited level, you have to explicitly say “would you prefer an unlimited level of X”. And if you want to ask if they would prefer an unlimited level of X even if it affects other considerations, you must explicitly specify that too.

          Even the question about directly creating new sentient beings would normally be read by most people to mean “creating some finite number of sentient beings” and does not imply there is no point of diminishing returns past which they would no longer want to create any.

          • Adam says:

            Not only that, but people are pretty clearly imagining having their own children, and it is hard to extricate rational reasons to prefer more persons to fewer from involuntary post hoc justification of the basic compulsion felt by all biological creatures that voluntary reproduce to create copies of themselves.

            No one is imagining a hypothetical world in which they flip a switch and, several hundred billion years later, long after humans have gone extinct, 100 trillion trillion gas creatures from 2061 come into existence magically that we could never communicate with or even recognize as alive, that can’t even communicate with each other but nonetheless are capable of experiencing positive utility. They’re imagining all of the pride and good feeling that comes from family, culture, and shared history.

      • eyeballfrog says:

        The answer to “Does anyone really believe X?” is “yes” independent of X.

      • jes5199 says:

        do it for the children!

  4. Rob says:

    “to being okay with certain damnation (n = 69)”

    Eeeeer…. surely they are just winding you up?

    • jeorgun says:

      I ended up answering 50%, but a more honest answer would’ve been “either 0 or 100, depending on the hell in question”. If it’s something like the hell in Hell is the Absence of God or The Great Divorce, or if it’s finite and you go to heaven after atoning for your sins or whatever, I’d absolutely prefer that over oblivion, so the question of heaven is moot; whereas if it’s “spend all eternity in literally the worst conditions possible”, I’m not gonna take that chance no matter how good heaven is.

      • Garrett says:

        Doesn’t that assume that the hell-bound are randomly selected as opposed to deserving? A whole stack of theology is about determining who is going to hell (or, what one needs to do/be/think/say in order to not go there).
        You can be fine with 99% of people going to hell if you are in the 1%.
        Or you can look at the Jehovah’s Witnesses who believe in hell and that they almost certainly will end up there because heaven is finite and so there probably won’t be room for them.

        • jeorgun says:

          In that scenario, I interpreted the question as asking about the probability that I’d manage to successfully live a virtuous enough life to end up in heaven.

      • wintermute92 says:

        It’s worth noting that some religions (even major Judeo-Christian ones) use a Hell or Hell-analogue that’s simply not that bad. Judaism tends to deal in “nothingness” or “not heaven”, Mormonism deals in “atonement” (with a side of “nothingness”) and so on.

        While I assume many people inferred Catholic-style “infinite bliss, infinite suffering”, it’s possible to get a wildly different standard while still answering in line with mainstream Western religion.

      • Eric Rall says:

        And then there’s Archbishop Blackadder’s conception of Hell:

        Hell isn’t as bad as it’s
        cracked up to be. […] No; you see, the thing about Heaven, is
        that Heaven is for people who like the sort of things that go
        on in Heaven, like, uh, well, singing, talking to God, watering pot
        plants. […] Whereas Hell, on the other hand, is for people who like the other
        sorts of things: adultery, pillage, torture — those areas.

    • Scott Alexander says:

      I’m not sure how many people misunderstood the question. I noticed that the expected correlations got a little stronger when I deleted everything below 50, which is consistent with some of those people giving numbers the wrong way round.

      • Theo Jones says:

        On the most recent open thread, there was this poster

        My initial thought, about halfway through the survey, was so glad I’m Catholic, it saves ever so much time and energy worrying about made up ethical debates.

        (Years back, I had someone respond to an argument that I’d made with, “I guess you’re not a utilitarian, then,” in the same tone as you’d say, oh, I guess you’re the sort of person who picks their nose in public. I am so glad to have found this blog, it has improved my opinion of utilitarians ever so much.)

        On reflection, though, I think that maybe religious types also spend a great deal of time pondering ethics, it’s just different questions.

        When I got to the Hell/Heaven vs Oblivion question, it was very easy, because in my concept, God is in the first universe, and not in the second, and even if I screw up badly enough to take the descending ‘vator, I would have lived in a world where there was a God.

        Other people who put that answer may have had similar reasons.

        For my part, I put 50% on that one.

      • Pickle says:

        I did not understand n “percent certainty of going to heaven” to imply 100-n% certainty of going to hell. I understood it as “your belief that you will go to heaven is n% correlated with the outcome as compared to population statistics,” i.e. at 0% you go to heaven at population odds.

        • Elephant says:

          I’m not a survey respondent, but this is similar to how I interpreted it when reading the question above. I took 0% to mean “life ending in oblivion” with no chance of going to heaven or hell, and a finite percentage to mean that degree of certainty (x) of life resulting in going to heaven or hell, presumably based on behavior, and 1-x chance of ending in oblivion. The question then seemed to be about how big x would have to be that you’d care about it.

        • Broggly says:

          I had a Calvinist interpretation, that the question was asking what proportion of the population was in the Book of Life.

      • wfro says:

        I answered 0% knowing exactly what it intended and assuming that the remaining percentage chance was hell-bound, for what it’s worth.

      • Yro says:

        Personally I answered 0%, as what scares me about death is the end of sentience. If hell meant losing my mind, then I would want a higher percentage. If it meant being practically immortal but having to push a stone up a hill for eternity, I could perhaps live with it.

        My answers were inconsistent as you look at them because I do not expect others to share this preference.

        I am an atheist btw.

      • keranih says:


        It is very possible that removing outliers that don’t fit with your hypothesis was not the best choice.

        Rationally speaking, I mean.

      • Sivaas says:

        My assumption was that the figure was the degree of certainty that I would go to heaven, so 0% would be completely unsure whether I would go to Heaven or Hell, 100% would be something like a divine intervention telling me no matter what I’d go to Heaven, and I think I was assuming some divine declaration of probability for lower numbers.

        I think I’d still take 100% probability Hell over oblivion, but it’s a much closer call than the question I thought I was being asked, although having a guarantee of some religious higher power might be enough value on its own to make it a better world.

      • Orphan Wilde says:

        I’d choose oblivion over a lifetime of monotony, but an eternity of pain (or possibly even monotony – if hell is an eternity of boredom I might have to rethink that question) over oblivion.

        There are reasons, mainly coming down to “Pain is interesting in a way that monotony isn’t, and also eternity is a nice compensation package in any case”. I think I may have put 50% down for the “hell” question, though.

        • Adam says:

          I suspect that you have never experienced truly crippling pain, to the point you cannot move at all, to the point that your brain regularly shuts down and you pass out, to the point you would gladly kill yourself if you could but you can’t even move yourself to the window ledge ten feet away. It is not interesting or preferable to monotony in any way, not remotely.

          Edit: Heck, I’m just thinking of my own experiences, but honestly, I’m not even sure I’ve had the worst of it. My wife is congenitally immune to lidocaine, which she unpleasantly found out while receiving a root canal, an experience she doesn’t even remember now but I just can’t imagine how that experience, repeated endlessly, can possibly be preferable to anything, let alone monotony.

          • Orphan Wilde says:

            Imagine being wireheaded so that you experience pleasure so intensely you would do anything to make it stop.

          • Adam says:

            What the heck does that mean? You may as well ask me to imagine a rock so heavy it weighs nothing.

          • Immortal Lurker says:

            @Orphan Wilde

            What I think you are saying here is that you have some mental function you can run which allows you to gain positive utilons from pain and suffering. You focus on how interesting the experience is, and suddenly you don’t mind it any more.

            Even allowing that ability to be effective for unending torture, I think that is fighting the hypothetical to some extant. There is the pithy answer, which is as soon as you arrive in hell, the demons snip whatever neurons allowed you to derive utilons from torture, and proceed to torture you. But even that is fighting the hypothetical.

            I think that “Hell” in that sentence should be replaced with “The worst thing that could possibly happen to you, and you are allowed to violate any law of physics, biology, and neurology to make it worse”. There are possible minds in mindspace where the worst thing is oblivion, or monotony. I am very confident that I am not one of them. I am confident that nearly every evolved mind is not one of them, enough so that I assume anyone claiming that the worst possible thing is monotony or oblivion is mistaken about their preferences.

          • Orphan Wilde says:

            @Immortal Lurker:

            Effectively, yes, pleasure and pain, as qualia, are bloody well near identical. An identical sensation can be pleasant in one context and unpleasant in another.

            Cutting the neurons, or whatever, would just be a mental press on “Negative context”. And anything at all would be unpleasant with a sufficiently negative mental context. The mind control, in that context, is just a press down on a mental “Negative Utility” button. Your utility goes down; your utility multiplier on every activity becomes negative.

            With that kind of mind control being applied, you could be exactly as miserable eating a tart as being tortured; physical punishment is unnecessary and pointless and subtracts from our understanding by imposing biological biases. This implies a kind of Hell far, far worse than the one Scott describes for its subjective occupants, one that would look pleasant from the outside.

            So I’d be inconsistent if I preferred a happy but short life to a long but unhappy one, but preferred oblivion to hell. That’s not a good understanding of what “unhappy” is. [ETA: Preferred a long unhappy life over a happy but short one, but then also preferred oblivion to hell.]

          • onyomi says:

            “oblivion, or monotony.”

            Not saying you did, but let’s be careful not to conflate oblivion and monotony. I’d much prefer oblivion to an eternity of monotony. In fact, one of the first ways I recall imagining hell as a child was simply being stuck in some kind of tiny, featureless room for all eternity with no way of achieving oblivion or escape. I’m not sure whether being stabbed in an immortal face by demons for all eternity would be worse; I think it might be better. Anything which offers the chance of interaction with other beings, even beings one hates, seems preferable to eternal loneliness.

            Contrasted to that, oblivion is not boring. It’s just nothing. Were you bored for the billions of years which preceded your birth?

      • wintermute92 says:

        I actively dropped the question because my answer was going to be intensely hell-dependent. There are major Western religions preaching versions of Hell that I would take at a 0% chance of salvation, simply as better than nonexistence.

        “Infinite suffering vs. infinite bliss” would get either 100% or 51% (infinities are tricky, not sure), “God’s grace vs. lake of fire” would invite a little more risk, and the various “outside God’s grace” or “place where you atone” hells look way nicer than nothingness.

      • Ninmesara says:

        You should have been more specific about the hell in question… Adding “Unsong’s Hell” (plus a requirement to actually read the relevant chapter, of course) to the statement of the question would have gone a long way to solve the ambiguity. I don’t believe the ambiguity with respect to Heaven is important, but given the number of people who think of Hell as “absence of God”, “nothingness”, or similar, it would have been a good idea.

        My idea of Hell is something like Unsong’s Hell instead of the milquetoast alternatives, but many people seem to think differently.

        Other than this, it’s a great survey, and the other questions seem unambiguous.

    • Evan Þ says:

      Not really. I can imagine someone who prefers continued existence even under torture to oblivion. And Scott’s question doesn’t even specify that Hell involves torture; C. S. Lewis’s The Great Divorce postulates a Hell mostly comprised of boredom and maybe psychological torment, while Mammon on Reddit comments that he imagines Hell as “a vacation club for people who need some amount of coercion to get along.” In either of those circumstances, I could readily imagine someone being okay with P(Hell)=1.

      Disclaimer: Being myself religious, and sure (P > .99) that Heaven exists and that I’ll go there, I found it very difficult to even put myself in the right frame of mind to consider going to Heaven as a randomly-picked variable.

      • Koldos the Shepherd says:

        Can you elaborate on your religion? I am very surprised that there would be a religion that would allow you have 99% confidence of ending up in heaven. Is that because pretty much everyone does, because you are in some way very special or some other mechanism? Are there really no bad life choices open to you that give more than 1% probability of not ending up in heaven?

        • Two McMillion says:

          I can’t speak for Evan, but I also have a >99% certainty I’ll go to heaven, and I’m reasonably sure Evan has the same religion as me. Your question about if there’s any act I can do that would lead me to not go to heaven is coming at the issue backwards. It’s the fact that I will go to heaven that keeps me from doing bad things. God saves by grace through faith, and it’s being saved that causes you to take good actions, not the other way around.

          • Adam says:

            That gives you P(heaven) > 0.99 conditioned on you actually being saved, but why should the probability that you’re correct in believing yourself to be saved be 100%? Surely, some people have believed this but were later proven wrong. It’s not completely impossible you’re such a person.

          • Two McMillion says:

            Well, I’m a Christian, and the Bible talks about exactly this in a couple places. One place is 2 Peter 1:

            His divine power has granted to us all things that pertain to life and godliness, through the knowledge of him who called us to his own glory and excellence, by which he has granted to us his precious and very great promises, so that through them you may become partakers of the divine nature, having escaped from the corruption that is in the world because of sinful desire. For this very reason, make every effort to supplement your faith with virtue, and virtue with knowledge, and knowledge with self-control, and self-control with steadfastness, and steadfastness with godliness, and godliness with brotherly affection, and brotherly affection with love. For if these qualities are yours and are increasing, they keep you from being ineffective or unfruitful in the knowledge of our Lord Jesus Christ. For whoever lacks these qualities is so nearsighted that he is blind, having forgotten that he was cleansed from his former sins. Therefore, brothers, be all the more diligent to confirm your calling and election, for if you practice these qualities you will never fall. For in this way there will be richly provided for you an entrance into the eternal kingdom of our Lord and Savior Jesus Christ.

            There’s a lot to talk about here, but I want to point out that it’s an example of what I said before: doing good things isn’t what sends you to heaven; being on your way to heaven is what makes you do good things. That’s why it starts with “he has granted to us his precious and very great promises” before going on to “for this reason make every effort” and so on.

            So when you ask the question, “How do I know that I am saved?” the first part of the answer is, “What does my life look like?” Does it have a pattern where I am increasing in virtue and knowledge and so on, or am I staying in the same place or moving backwards? When we read the rest of the Bible, we also realize that many of these qualities are as much internal as external. When I become more virtuous, am I also secretly becoming more prideful about it? Depending on your answer, there may be cause to examine yourself and see where you really stand.

            When the Bible talks about “real faith”, it always refers to action: what acts have you taken because of what you believe? How has it made a difference in your life? If there’s not one, maybe you should look into that.

            And then there’s also an ineffable quality that being saved has that being unsaved lacks. The Bible talks about how we are indwelt by God’s spirit. Slowly you begin to see changes in yourself. Slowly you become more aware of spiritual truths. Slowly you begin to see a spiritual dimension to events that doesn’t diminish any part of the physical reality. Never bright lights or fire from heaven, but a definite push towards certain things and away from others. A growing hatred of wrong and a love of what is right; a growing ability to see good and evil in starker contrast, a growing vision of both how the world is broken and what the world redeemed will be like. Something inside you has changed, and you have a sense of both yourself broken and yourself set right again. You read the Bible or hear it preached and suddenly words that never made sense to you before fit together perfectly, and you see how obviously it was true all along. You live life but not for life; you learn how to smile in suffering, you learn how to say “I was wrong” and “I’m sorry”; you learn how to make peace and forgive. Each individual step always seems perfectly natural. It’s only when you look back that you see God was with you all along.

            So if you ask me how I know I’m saved- well, I know it as I know most anything else I’ve known for sure. I’ve been there. I’ve walked through it and stood beside it and laughed and cried and screamed over it. And maybe that’s not a knowledge I can communicate to someone else. Maybe the path I’ve walked isn’t one I can share. But I look into myself and when I see light it is light shining in from outside, and the only choice worth making is kneel before the One from whom that light shines.

          • Jiro says:

            So if you ask me how I know I’m saved- well, I know it as I know most anything else I’ve known for sure.

            How do you know anything else for sure?

            Knowing something always includes at least the possibility that you’ve made mistakes or rationalizations. Are you claiming that religion is something about which it is impossible for you to have made mistakes?

          • Skivverus says:

            Going to borrow a phrase from a different Scott A. – the way I read this is that Two believes that his religious beliefs are sanity-complete: sure, he could be rationalizing them or outright wrong, but he could also be a brain in a vat in a universe where 1 + 1 = 7; his priors on further exploration of the topic resulting in a change of opinion (or of action) are indistinguishable in practice from zero (or whatever his minimum value is for priors).

            Whether or not this is actually the case, well, that depends on which definition of “knowledge” you’re using.

          • Adam says:

            What Jiro said. I don’t know a single thing with 100% certainty. I mean, I’m infinitesimally close on “other people exist” and “the sun will rise tomorrow,” but metaphysical propositions on which all of humanity has had widespread disagreement throughout its entire history? I don’t think you’re calibrating properly.

          • Two McMillion says:

            Knowing something always includes at least the possibility that you’ve made mistakes or rationalizations. Are you claiming that religion is something about which it is impossible for you to have made mistakes?

            Well, of course making mistakes is also possible. But my past experience in correcting mistakes has been less like exchanging geocentricism for heliocentricism and more like exchanging Newton for Einstein, if you understand my meaning. Most changes of opinion aren’t about overthrowing existing beliefs but about subsuming existing beliefs. In other words, there’s a discernible process by which beliefs level up, in which they gain not only greater explanatory power, but also a kind of aesthetic beauty. To change my beliefs, I need to be shown that the change in beliefs has this “level up” quality, and while Eilezer’s version of atheism is indeed very powerful and very aesthetically pleasing, it is not quite up to the level of my current set of beliefs. I think I can explain Eliezer, but I do not believe that he can explain me. “Sanity-complete” is a good way of putting it, if possibly an undersell.

            And of course there is the part that comes from living it all, which isn’t easily communicable.

          • cypher says:

            If doing good from within is a sign of being saved, what does that say of athiests and followers of other religions that do good things – or even great (both in the large sense and the positive sense) things?

          • FacelessCraven says:

            @Cypher – “If doing good from within is a sign of being saved, what does that say of athiests and followers of other religions that do good things – or even great (both in the large sense and the positive sense) things?”

            …Nothing? Doing good out out thankfulness and a desire to more closely emulate God is one reason to do good. there are others as well. The good done isn’t an entry on a ledger.

            @Adam – “What Jiro said. I don’t know a single thing with 100% certainty.”

            Sure. But one of the specific instructions, as I understand it, is to not freak out about going to hell. Like, worrying about going to hell if you do it wrong is itself doing it wrong. You are guaranteed not to be doing it right, and you can and should make every effort to do it less wrong, but there are no points, there is no pass/fail threshold. It’s a relationship, not a test. Hence the stuff about Grace and Faith and so on; if God promises you that you’ll be fine if you trust in him, either you trust him and everything’s fine, or you don’t believe him, and then it’s easier just not to believe IN him, and then what’s the problem?

          • Adam says:

            Dude, the entire point is everything you said is one possibility. You stated it as if it is the absolute truth. Of course if it’s the absolute truth, then you’re good to go, no danger. Assuming that to begin with, and then using that assumption to justify your belief that you’re going to heaven, is question-begging.

      • Winter Shaker says:

        That sounds to me like quite some confidence. I agree with Koldos in being curious as to how you get so high a probability.

        I mean, my own estimate of the probability of Heaven as traditionally described actually existing is about as close to zero as yours is to 1, but if I thought that there was a Heaven with your level of confidence, my probability estimate of ‘and I will go there’ would be much lower. Roughly, if I thought that it was at all likely that we lived in a universe with something like one of the Abrahamic gods behind it, I would be terrified of going to Hell for choosing the wrong religion, because if I was a mainstream [Christian / Muslim], the existence of a billion-plus [Muslims / Christians] would be proof that billions of people can believe something catastrophically wrong with just as much intensity as others can believe something true, with the decision largeley predetermined for you by your upbringing, before you even get into ‘actually, neither of the above; some other existing religion, or even one that doesn’t exist yet, is true’ territory (or ‘one specific denomination or subgroup is true’).

        So I’m curious as to how you get from the high confidence that there is a heaven, to the comparably high confidence that we don’t live in a universe that has a hell as well.

        • Civilis says:

          Disclaimer: Although Catholic, I am not the Pope, and therefore cannot speak for official Catholic doctrine. What I write below is an attempt to explain a concept I’m a little fuzzy on myself in modern terms. Further, the ideas of Heaven and Hell are so distorted by being wrapped in parables, metaphors, and Dante’s later allegory that we can’t know what God has in mind.

          My understanding is that Catholic theology posits a third option, Purgatory, which can be likened to Temporary Hell Lite or Club Fed Hell (as opposed to Maximum Security Regular Hell). If you’re bad but not irredeemable, you end up spending a lot of time (which is subjective, considering we’ve passed out of time entirely at this point) coming to understand what you did wrong, then you get moved on to Heaven. So it’s easy to think that most people will end up in Purgatory if you think people as a whole are generally good.

          Catholic theology got tripped up at some point by people asking about the extreme edge cases, people who were incredibly incredibly good people in life who may have lived in times or places where they never could have even heard of Christianity, or people that were not Christian but were innocent enough to be incapable of being evil, such as the really really young. There seems to be general agreement that whatever happens in these cases, it isn’t bad. Dante included a ‘good-but-not-Christian’ section in Hell in his Inferno in the form of Limbo.

          • Adam says:

            That causes me to wonder how it is possible, given infinite time, for a person to be completely beyond any future hope of redemption. Surely some extremely long but finite period of suffering is sufficient punishment for any finite crime.

          • Anonymous says:

            Pretty sure some of those people are members of the Invisible Church.

          • Jaskologist says:

            Punishment is only half the issue. Heaven is a place where everybody always picks cooperate over defect. You’re asking how long they need to be punished for defecting, when really we want to know what it will take for them to go and defect no more.

          • Anonymous says:

            And Presence charms requiring more Willpower points than you have to resist is not the answer.

          • Adam says:

            Jask, you just said heaven is a place where everyone always chooses cooperate over defect. 1) No person’s willful choices are ever guaranteed to be stable for an infinite period of time, so you can never guarantee this for any person, even if they managed to never defect while spending several decades alive on earth, and 2) if heaven is like this because everyone always cooperates without having to choose, then you can guarantee a person will always cooperate by simply putting them in heaven.

          • Civilis says:

            That causes me to wonder how it is possible, given infinite time, for a person to be completely beyond any future hope of redemption. Surely some extremely long but finite period of suffering is sufficient punishment for any finite crime.

            I was once in a fantasy PBEM game where you play a Greek-style deity in charge of a group of followers, and the game left me with a theoretical question that occasionally, on those dank, sweaty summer nights, keeps me awake.

            To translate it into understandable terms, imagine a proposal for justice reform: we build a secure and isolated prison in the middle of nowhere, and when we say we execute someone, we just fake their death and put them there. Criminals still fear the death penalty, but there’s no actual risk of killing an innocent person. If we find we’ve accidentally put away an innocent person there for a capital crime, we give them a ton of money and put them in a witness protection program. Since we’re human, we obviously can’t make this work perfectly, but if we could, would it be better than the system as it is now?

          • Adam says:

            To me, yes, of course that would be better. To someone who thinks a proper purpose of the criminal justice system is to actually dole out punishment in proportion to offense and just to aid other lines of effort in reducing the treat and occurrence of crime, the answer might be no, it’s better to actually kill people even if some of them are bound to be innocent.

          • Civilis says:

            What percentage of people need to receive the maximum punishment for a crime for that punishment to be effective?

            In this case, we have a crime, and we know the range of punishments the judge is allowed to give out, but we don’t know what sentence previous defendants received. In order to reduce crime, it’s in the judges best interest to make sure people know what the maximum sentence is, regardless of how often that sentence is given out.

          • Adam says:

            I’ll freely I’m no criminologist, but I was under the impression that the best empirical evidence suggests that sentence severity has almost no deterrent effect anyway. The perceived likelihood of getting caught and punished at all does, though.

          • Civilis says:

            The classical, allegorical definition of God, heaven and hell, postulates a 100% chance you will get caught. To human minds, when dealing with the infinite afterlife, the punishment as far as we can understand it effectively amounts to either ‘slap on the wrist and time served’ or ‘death penalty’. If you want to reduce crime, I think this is one point where you’d play up the greater potential sentence.

          • Adam says:

            Well, the key thing is believing there is a 100% chance of getting caught. So between non-believers and indulgences, that wasn’t necessarily the case. God could have smartly invested effort into presenting convincing evidence of the truth of his word and not allowing it to be presented through corruptible priests.

          • Mary says:

            “My understanding is that Catholic theology posits a third option, Purgatory, ”

            Purgatory is not a third option. It’s the ante-chamber to Heaven. If you made it to Purgatory, you’re saved. You’re just getting clean before you make your entrance.

          • Mary says:

            “No person’s willful choices are ever guaranteed to be stable for an infinite period of time, so you can never guarantee this for any person, even if they managed to never defect while spending several decades alive on earth, ”

            No person ON EARTH.

            The Afterlife is a different mode of existence, possibly not even involving being related to time in the same way we are now.

          • Civilis says:

            Well, the key thing is believing there is a 100% chance of getting caught. So between non-believers and indulgences, that wasn’t necessarily the case. God could have smartly invested effort into presenting convincing evidence of the truth of his word and not allowing it to be presented through corruptible priests.

            Without getting too much into the early history of the church, which, admittedly, like all human things is necessarily imperfect, this isn’t how Catholics see it.

            Nobody is perfect. We all make mistakes. To a Catholic, we are all sinners, because we all have moments of pride, lust, greed, etc. Being good just means you handle that better. It’s like someone that escaped from prison and lived his life as a model citizen; he’s still guilty of the original crime, but his subsequent behavior earns him leniency from the judge. (Indulgences were more ‘I was a great guy before I robbed the bank, so go easy on me.’) With non-believers, you may not know what Christian teaching says, but you still have a sense of right and wrong.

            As far as the whole Babelfish paradox goes, I have my own theories, but they are a product of my own odd internal logic.

          • Civilis says:

            Purgatory is not a third option. It’s the ante-chamber to Heaven. If you made it to Purgatory, you’re saved. You’re just getting clean before you make your entrance.

            We’re dealing with analogies to something unknowable. Purgatory is a third option if the only two options presented are ‘go directly to heaven’ and ‘go directly to hell, do not pass go, do not collect $200’.

            When discussing criminal justice, if we’re comparing ‘prison’ and ‘not prison’, ‘probation’ is covered under ‘not prison’, but it also seems useful to distinguish between ‘not prison – free’ and ‘not prison – probation’.

          • Adam says:

            Without getting too much into the early history of the church, which, admittedly, like all human things is necessarily imperfect, this isn’t how Catholics see it.

            Yeah, but “non-believers” includes non-Catholics. Obviously, hell can’t serve as an adequate deterrence for them if they don’t believe it exists.

          • Civilis says:

            Yeah, but “non-believers” includes non-Catholics. Obviously, hell can’t serve as an adequate deterrence for them if they don’t believe it exists.

            You’re not a believer and you know about hell. That’s why most Christians are called to evangelize (which does not perfectly overlap with Evangelical Christians).

            As far as believe, that gets into the Babelfish Paradox mentioned above. (Short version: it’s possible to set up a God where he can’t prove he exists.)

          • Adam says:

            Yeah, of course. I guess the Franciscans are distinct from what we’d call “evangelicals” and the process by which they made my ancestors Catholic was not a pretty process, though I guess it illustrates the idea to turn life itself into something like hell.

            But still, I don’t see how knowing some people believe in hell is supposed to serve as a deterrent to me. Even though Buddhists don’t really evangelize, I’m sure you at least know of Dharma, but the prospect of being reborn as an untouchable probably isn’t what keeps you from being a jackass in this life. They know about hell, too, but it doesn’t motivate them.

        • Tamar says:

          I’m a religious Jew and I also had a very hard time answering, because Heaven/Hell dichotomy doesn’t make much sense to me and I had to guess what Scott was intending. I ended up answering ‘100%’ after trying to translate the question in terms that I can actually relate to. Let’s put it this way: I have a high confidence that if there are consequences for our actions after death, that it derives from the parameters of my own religion’s most consistent and reasonable (and rationalist?) sensibilities and is just, and basically doesn’t have anyone ‘going’ to a Christian-esque version of Hell, and neither to the Christian-esque Heaven (but with basically everyone having a limited-time experience comparable in a limited way to ‘purgatory’ – an experience that is not pleasant but not comparable to Hell). If not this – if my own religious beliefs are fundamentally untrue – I expect that it is definitely the case that no religious outlooks are true and thus the after-life outcome for everyone is oblivion. I’d rather oblivion than unjust afterlife consequences for my or others’ actions, and I require 100% confidence that the afterlife consequences are just.

        • Two McMillion says:

          I’ve never understood the “Muslims don’t worship the same god as you”! argument. I mean, it boils down to, “There are a bunch of people who disagree with you, now what?” Well, okay. There are lots of people who disagree with me on every conceivable topic, not just religion. That doesn’t stop me from holding an opinion in those areas; why should it stop me here?

          • Adam says:

            It shouldn’t stop you from holding opinions. It should stop you from being 100% certain you’re correct about them. Arguably, most of your opinions can even be held with P(correct) < 0.5 if you're choosing from more than two disjoint possibility subsets, as is the case when choosing a specific religious particularism.

          • Winter Shaker says:

            Adam already answered this one well. But to be specific, if you think that a heaven exists, the number of people who agree with you that a heaven exists is vastly higher than the number of people who agree with you about exactly what it takes to get there. And from the outside, the people who disagree with you about that seem to believe their versions about as passionately, and with about as much good evidence, as the people who agree with you. Some of them even hold that what you think is necessary to get to heaven is in fact guaranteed to prevent you getting there.

            Therefore your probability estimate of a heaven existing should be significantly higher than your probability estimate of you personally getting to go there.

            Evan Thorn seems to have p(a heaven exists) ≈ p(a heaven exists and I personally will go there), which is a level of confidence that needs some explanation.

          • Two McMillion says:

            I think that characterization of the problem relies on some pretty big (and completely unfounded) metaphysical assumptions, personally, but I can see that perspective, I suppose.

          • Cypher says:

            You essentially must consider the possibility that some third factor is making both groups believe, such as an innate human propensity to religion.

            Or, much less charitably, if you know it’s possible to make a mind control ray, even if it’s imperfect, you must ask yourself if such a device has been used on you already. Certainly, we can observe some political ideologies massively distorting peoples’ worldviews into nearly-inescapable spirals that they are willing to die for.

    • I don’t know. I took the survey and put my required odds of heaven at about 5%.

      Here’s the thing: while I can sort of maybe make peace with my personal oblivion, the notion that eventually all humans everywhere will dissolve into nothingness fills me with utter horror. I absolutely prefer a universe with Hell to a universe with Nothing. If the universe also happens to contain Heaven, and lots of people are there, then so much the better.

      • Decius says:

        I think this might be endemic: People profess to prefer Hell to Oblivion, because they prefer to believe in Hell rather than believe in Oblivion.

        Suppose that you cannot change your beliefs nor the beliefs of anyone else. Keeping your expectations up until the point of death constant, would you prefer oblivion or a hell created by a competent entity with the express purpose of making the afterlife worse than oblivion?

        • Anonymous says:

          I would, and do, prefer oblivion to a Heaven or Hell. A deity that created a world of such needless suffering and demands that we worship and obey it for all eternity or we will be punished for all eternity is not a deity that I would serve. To that capricious, unworthy deity I say, “non serviam.”

          • Vamair says:

            What if no deity, just Heaven and Hell for pure physical reasons and a quantum random number to decide where you’d go.

          • eh says:

            To me, this seems like freely choosing a death sentence because you think the judge is an arsehole, even when you have a get out of jail free card. If the hypothetical judge IS actually an arsehole, he doesn’t really care that you’ve been executed, so it’s a bit pointless.

          • Anonymous says:

            Ew, I have to reply to myself to reply to you…

            Vamair said:
            “What if no deity, just Heaven and Hell for pure physical reasons and a quantum random number to decide where you’d go.”

            If there were a way to rig the game, I’d likely choose the supposedly better option. If not, I’d take oblivion, after all, I wouldn’t be around to care.

            Eh said:
            “To me, this seems like freely choosing a death sentence because you think the judge is an arsehole, even when you have a get out of jail free card. If the hypothetical judge IS actually an arsehole, he doesn’t really care that you’ve been executed, so it’s a bit pointless.”

            Except it isn’t the choice between do what you want or be executed, the choice is worship the arsehole judge for eternity, or be executed. A cage is still a cage, no matter that it be appointed with silk sheets and jewel encrusted bars. Besides, if there is an omnipotent, omniscient deity, it would certainly know what I think of it and its shoddy work, and I’m pretty sure that’s on the no-no list of all the Abrahamic religions and their various offshoots.  Which, I’m assuming are the versions of heaven and hell on offer.

          • Berna says:


          • Two McMillion says:

            Actually, doing what you want to do is a form of worship.

          • Winter Shaker says:

            doing what you want to do is a form of worship.

            Can you elaborate on that?
            I get the impression that there is a fair bit of typical mind fallacy going on between instinctively religious and instinctively non-religious people – I’ve often heard it claimed that atheists ‘worship themselves’ or ‘worship science’, or such like, from people who, I suspect, simply cannot imagine what it feels like from the inside to just not have any inclination to worship anything, and therefore cast around to see what it is that the atheists worship if they aren’t worshipping any gods.

            And from my own perspective, I can’t be sure, but I suspect that one of the factors contributing to me growing up to identify as an atheist was: why would a god which wanted us to worship it design our brains in such a way as to find the act of worshipping it so mind-numbingly boring? – I was unable to wrap my head around the idea that some people actually enjoy praying and church rituals.

            I hope I am not being too uncharitable in suggesting that, in calling ‘doing what you want’ a form of worship, you may be using an unnaturally expansive definition of ‘worship’ that doesn’t actually map onto what we normally consider the word to mean, or onto anything the people doing what they want are actually feeling from the inside.

    • Perhaps you underestimate peoples capability for self sacrifice? What if somebody believes that:

      * They currently live in the universe where heaven and hell don’t exist.

      * If they did most people would go there but they personally wouldn’t.

      * The universe where heaven and hell exist is better than the one where it doesn’t, in pure utilitarian terms.

      So they would be willing to take on the personal sacrifice of eternal damnation so that others can live in heaven. We have an entire mythology surrounding a person who is revered for having done just that, so it’s not inconceivable that there are others who would be willing to follow his example.

      • Zippy says:

        The first part of the survey, not included in this post, is the disclaimer:

        For all questions, please assume that results will have no effect on any other people, including your friends and family. If it helps, imagine your mind has been copied, and these questions will affect a simulated version of you while your real friends and family are unaffected. Please answer them solely according to your own selfish preferences.

        • Well I doubt I would be the first person to have skimmed/forgotten that by the time the heaven/hell question rolled around.

        • Edward Scizorhands says:

          “For purposes of this survey, assume your are a spherical cow of uniform density.”

        • AlexC says:

          I did find the heaven/hell question *really* hard to answer in a selfish way. It took about four attempts to mentally separate the implications for society, theology, cosmology etc, of God providing a possibility of heaven, from whatever much more minor selfish interest there might be.

        • Linch says:

          That part makes question 3 really confusing for me.

      • Kiya says:

        Eternal damnation seems like a much more major sacrifice than getting crucified for like a day?
        (If you’re referring to a different mythology and/or character, I am intrigued.)

        • Evan Þ says:

          Though it’s not clearly stated in the Bible, there’s a longstanding tradition – embodied in many versions of the Apostle’s Creed – that Jesus also descended into Hell.

          Of course, unlike the hypothetical individual in this survey, Jesus didn’t remain there. But the Apostle Paul would’ve answered in the same way – he “could wish that I myself were accursed and cut off from Christ for the sake of my brothers, my kinsmen according to the flesh” (Romans 9:3).

          • Douglas Knight says:

            Are there multiple versions of the Apostles’ creed? Do you mean the Nicene creed as a variant? (The Athanasian creed is pretty different, but also mentions the Harrowing.)

            (I’ve seen people say similar things before, often in the context of the Harrowing, and I’ve always been confused, but now I have a chance to ask.)

          • Evan Þ says:

            @Douglas, there’re definitely different versions of the Apostles’ Creed in use in the modern day – and no, I’m not referring to the Nicene or Athanasian. (A lot of them are trivial translation differences like “quick” v. “living,” but not all.)

            I was under the impression there were ancient textual differences as well, with some texts omitting “descended into Hell” – but I might be misremembering; a quick web search doesn’t turn up anything identifiable as the Apostles’ Creed which omits that. There was a shorter Creed that did, but it’s sufficiently different that I wouldn’t say it goes by the same name.

            As for the modern variations… there’re some wild ones out there. The two I’ve heard in church are omitting “He descended into Hell” and replacing “Catholic” with “Christian.” There’s also a very-freely-metricized text (p. 4 here) which we sang to the tune of “Glorious Things of Thee Are Spoken,” but I don’t count that as a legitimate version.

        • jes5199 says:

          this is why Judas Iscariot is the real Messiah:

      • Commenter says:

        So they would be willing to take on the personal sacrifice of eternal damnation so that others can live in heaven. We have an entire mythology surrounding a person who is revered for having done just that, so it’s not inconceivable that there are others who would be willing to follow his example.

        Which one’s that? Not Christianity, or at least not Catholicism:

        On the third day he rose again
        in fulfillment of the Scriptures;
        he ascended into heaven
        and is seated at the right hand of the Father.

        • Anonymous says:


          Πιστεύω εἰς θεòν πατέρα παντοκράτορα, ποιητὴν οὐρανοῦ καὶ γῆς.
          Καὶ εἰς Ἰησοῦν Χριστòν, υἱὸν αὐτοῦ τòν μονογενῆ, τòν κύριον ἡμῶν,
          τòν συλληφθέντα ἐκ πνεύματος ἁγίου, γεννηθέντα ἐκ Μαρίας τῆς παρθένου,
          παθόντα ὑπὸ Ποντίου Πιλάτου, σταυρωθέντα, θανόντα, καὶ ταφέντα,
          κατελθόντα εἰς τὰ κατώτατα, τῇ τρίτῃ ἡμέρᾳ ἀναστάντα ἀπò τῶν νεκρῶν,
          ἀνελθόντα εἰς τοὺς οὐρανούς, καθεζόμενον ἐν δεξιᾷ θεοῦ πατρὸς παντοδυνάμου,
          ἐκεῖθεν ἐρχόμενον κρῖναι ζῶντας καὶ νεκρούς.
          Πιστεύω εἰς τò πνεῦμα τò ἅγιον, ἁγίαν καθολικὴν ἐκκλησίαν, ἁγίων κοινωνίαν,
          ἄφεσιν ἁμαρτιῶν, σαρκὸς ἀνάστασιν, ζωὴν αἰώνιον.


          Credo in Deum Patrem omnipotentem,
          Creatorem caeli et terrae.

          Et in Iesum Christum,
          Filium eius unicum,
          Dominum nostrum,

          qui conceptus est de Spiritu Sancto,
          natus ex Maria Virgine,

          passus sub Pontio Pilato,
          crucifixus, mortuus, et sepultus,
          descendit ad inferos,

          tertia die resurrexit a mortuis,
          ascendit ad caelos,
          sedet ad dexteram Dei Patris omnipotentis,
          inde venturus est iudicare vivos et mortuos.

          Credo in Spiritum Sanctum,

          sanctam Ecclesiam catholicam,
          sanctorum communionem,
          remissionem peccatorum,
          carnis resurrectionem et vitam aeternam.

          The statement is also present in my native language’s translation.

          • Koldos the Shepherd says:

            The native language version that was in my Catholic upbringing had “descended into the realm of the dead”, even though we had a separate word for hell.

            I think the Latin/Greek thing is an adaptation to the audience’s preexisting cosmology, since surely these words meant something that was not exactly “Christian hell”?

        • ThaddeusMike says:

          Google the phrase “he descended into hell”. You should get the catechism of the Catholic Church. Short version, yes, it is Catholic teaching that Christ descended into hell.

          • Winter Shaker says:

            it is Catholic teaching that Christ descended into hell.

            Is it Catholic teaching that he stayed there, and is even now suffering the traditional fate of the damned?

            I’m not highly trained in any of the Christianities, and the version I got at school wasn’t the Catholic version, but I got the impression that the story is that merely went to hell briefly, kicked some demon arse, then came back to earth and then on to heaven, which is whay Kiya seems to think.

          • sconn says:

            No. In the translation I used, he descended to the *dead,* not to hell as we imagine it. “Inferos” just means the lower regions, not “place of torment.” It is traditionally taken to mean Limbo, the place where the just waited for their salvation. Jesus was not punished there, he just came to get Adam, Abraham, Moses, etc. out and take them to heaven.

          • Winter Shaker says:

            That’s what I thought.

            So either John Pressman was talking about some other religion when he said that we have an entire mythology surrounding a person who is revered for taking on the personal sacrifice of eternal damnation, or he is talking about some variant of Christianity, but the ‘he descended into Hell’ claim is not where that idea comes from?

    • wintermute92 says:

      Nah, not necessarily. I declined to answer (for this reason), but there are conventional definitions of ‘hell’ that I would take at 0% salvation. Some just aren’t that bad.

      That said, I do think that given a frame of “maximum suffering forever vs. nonexistence”, people choosing the infinite suffering might be underestimating just how big maximum suffering and forever really are.

      • Walter says:

        Turn it around. People insisting that the suffering will last forever might be underestimating how long that will be. Something will change.

    • AlexC says:

      I put something like 0.001%, because I think torment can be only finitely bad, but an infinite God can come up with truly unbounded ways to make heaven better and better and better. 0.001% should be taken as “anything that’s not definitely zero”.

    • Richard says:

      Not a survey responder, but if we take the question to imply (eternal bliss / eternal suffering) vs oblivion, which given the other questions seems reasonable. Given this and that my only knowledge of the situation is my destiny, I would have to answer 101%, or possibly -1% because there is no situation where I would prefer to trade my eternal bliss for an unknown number of peoples eternal torment.

      (unless we are talking stupid mind games where the only situation where I’m eternally happy is where everyone else is too and the population of hell automatically becomes zero, but overthinking simplistic questions probably won’t help with such a study.)

      And in this case, my selfish preference does include an abstract number of other people so I can’t make the caveat apply

  5. I wish to point out that I found the concept of “unhappy but not suicidal” confusing. I was unable to determine whether or not the questions were supposed to indicate that the person was experiencing more suffering than happiness, though I answered as though they were.

    I’m also system-1-surprised but not system-2-surprised that people (apparently including Scott, from the wording of the survey) don’t automatically equate “experiencing net pain” with “wanting to die.”

    • Aevylmar says:

      I’m another one of those people.

      To me…

      Leaving aside the question of future obligations, whether friends and family will miss you, etc.

      For every moment, if I am unhappy in that moment, I would prefer not to have that moment, all else aside. That seems almost tautological to me. (Footnote: If, say, the moment was necessary for a Greater Good down the line, I’d rather have (bad moment + Greater Good) than (none), but I wouldn’t prefer (bad moment + greater good) to (greater good). I prefer buying food to starving, but getting equivalent food for free would be even better.)

      If every moment of my life I would prefer not to have, then I would prefer death to life. This, again, seems almost tautological.

      And I’m both system-1 and system-2 surprised that other people’s brain works in different ways, even though, in retrospect, this is a comparatively mild case of what I system-2-knew was going on all the time with typical-mind syndrome.

      • Mammon says:

        Moping can be very life-affirming and satisfying, though.

      • DavidS says:

        Can you in practice identify these net positive v net negative moments? Because I feel like it’s not that clear a lot of the time.

      • wintermute92 says:

        You might be able to get some System 2 understanding by chalking this up to people putting a negative valuation on change, or a positive value (non-equivalent to happiness) on experiencing consciousness, and continuity of consciousness.

        There is (obviously?) a level of badness at which I would start wanting to not experience time containing that badness, but it’s not “any time I’m net unhappy”. I derive some value (even if we just call it averted fear of death) from the mere fact of my existence, which creates a window of “unhappy but choosing life”.

      • Ghatanathoah says:

        Leaving aside the question of future obligations, whether friends and family will miss you, etc.

        I’ve found a helpful way of controlling for this is to replace “die” with “temporarily turn into a p-zombie.” Your p-zombie self won’t be unhappy, but will still be able to fulfill future obligations.

    • There are some people who take the view that living a full life entails embracing the whole range of human experience, including pain and suffering. This idea has been used to explain why so many people appreciate stories about tragedy and grief, because through such stories people vicariously experience the full gamut of human emotions and feel enriched. (Nietzsche in particular had a lot to say about this.) People who experience net pain might feel that occasional positive life experiences are of such great value that they make the pain worthwhile in some way. I also recall reading a study (can’t recall all the details at the moment) suggesting that there are some people who are not very happy but who experience life as meaningful, while some others are very happy and do not consider life particularly meaningful. (I presume there are some who experience both, and some neither, but I can’t recall the specific breakdowns.) Hence, some people may value meaning, however they interpret that, over hedonic balance between pain and pleasure.

      • J says:

        I have a low happiness set point, so my typical state is pretty unhappy. Nevertheless I’m able to create things of value and I have done lots of worthwhile things in my life. So I’m okay with living unhappily, and that’s different from times when I’ve felt despairing and sick of living.

        • Cadie says:

          I’m similar to this. I don’t know if I’m mentally capable of sustaining true happiness; the only happiness (as opposed to simply “not feeling downright terrible”) I’ve ever had has been brief and the direct result of some unexpected amazing thing happening. There’s a huge difference between being stuck in a freezing, dark pit that feels sharply awful, and days where the world feels emotionally overcast and a little too cold for my liking but is tolerable and basically “okay.” In the latter, there are bits of enjoyment to be had here and there from games, jokes, rum, and interesting conversations, and the bad parts aren’t bad and persistent enough to make me think that not existing is preferable. It’s been years since I’ve wondered if living was worth bothering with, despite still not feeling truly happy for more than a few hours at a time and very rarely, because the background usual state I’ve been dealing with since then has been bland and only slightly bad instead of miserable. Neutral and mixed are states I can deal with, especially after a quarter of a century at the “almost everything sucks” level. Feeling basically meh-whatever most of the time with occasional mild enjoyment and occasional suffering feels much, much better than constant and near-constant suffering. In fact, small amounts of suffering seem to make a neutral state feel a little nice, so I don’t think I’d want to eliminate them entirely; then I might start to dislike the sort-of-blah times instead of thinking “Hey, cool, I’m not horribly stressed, sick, or hating myself. Life’s okay right now.”

      • canebrake says:

        Yes; the “range of experiences” question is one that comes up much more rarely in rationalist circles than I would naïvely expect. I chose the 100 trillion trillion because (my immediate assumption was that) they must have a much greater diversity of stories. I did realize, after thinking that, that one could stipulate that they were all exactly as unhappy in exactly the same way, and in that case I would have been ambivalent. But absent that stipulation I would suspect that 100 trillion trillion people must have more thoughts than one million, that there are probably more ways to be unhappy than to be happy anyway, and that, frankly, in a universe of 100 trillion trillion people, at least some of them would have to enjoy being miserable. (Either because they are OK with not being happy if they can experience life as meaningful, or because they’re just genuinely into suffering. It happens.)

        I can experience the thing I think most utilitarians mean by “happiness”, but only in small and separated doses; it quickly acquires the loathsomeness of superstimulus. Meanwhile I have been able to love, in a way, periods of obscene stress, even as I approached suicidality, partly for producing work I can look back on and be proud of (when periods of happiness did not) and partly for being literally fascinating. In a universally happy world I could not exist.

        • Wilj says:

          This suggests me to that you’re using the word “happy” differently than most. “I am loving this moment despite the stress” sounds like net happiness.

          • canebrake says:

            Obviously, the definition of “happiness” is a bit vague, and I am not sure it is possible to communicate it accurately. But I do not remotely recognize that description. I have also felt ‘tense but excited’, and that was pleasurable, but it’s a long way from there to “oh god oh god oh god make it stop (but I will creepily savor this moment, now and later)”.

    • Immanentizing Eschatons says:

      I have a strong desire to continue existing, independent of any happiness or suffering experienced. There might be some level of suffering at which I would stop, but it would be much worse than what most utilitarians seem to suggest in discussions about this stuff.

      • Anonymous says:

        I have a strong desire to continue existing, independent of any happiness or suffering experienced. There might be some level of suffering at which I would stop, but

        I used to feel the same way, then I developed excruciatingly painful and progressive peripheral neuropathy. Daily pain and nightly torture wreaks havok on one’s will to live. I can appreciate people who are struck with crippling pain,yet can find happiness in their lives. If I didn’t have to provide for my family, I’d probably choose euthanasia at some point.

        • Hyzenthlay says:

          I’m terrified of one day getting a chronic pain condition. The idea of being stuck with excruciating physical pain all the time and being unable to fully escape it seems to me like one of the worst things that can happen to a person, if not the worst thing.

          But yeah, it’s interesting how different people respond to pain differently. I’ve known a number of people with fibromyalgia and similar conditions who nonetheless seem to still enjoy life, and who say that you’d be surprised what you can get used to. Others never really adjust.

    • Anonymous says:

      “unhappy but not suicidal”

      Welcome to my world. Happiness happens maybe once a month if I’m lucky, for a few hours. I still consider this existence better than non-existence. Maybe I’m just strange.

    • jes5199 says:

      I actually think that if you did try to quantify pleasure and pain in some units that you could just sum up, all living creatures would be on the “more pain” side. Fortunately, pleasure and pain aren’t opposites, don’t cancel out, aren’t even the same species of experience, and are two of the least important aspects of existence.

  6. Izaak Weiss says:

    When you say “All were correlated with preference for oblivion over suffering in the expected direction, but not very strongly”, do you mean that all four of those metrics were higher in people who preferred oblivion, or that some of the metrics are higher in people who prefer oblivion and some are higher in people who prefer suffering, but it’s along the lines you predicted at the beginning?

  7. I took the survey and answered your question about creating new sentient beings, but I was pretty upset that you didn’t even offer the option that most closely matches my intuition: that creating new happy sentient beings is better that improving the happiness of existing sentient beings. Did this option not occur to you?

    (Nonetheless I still opted for the world with a smaller number of very happy people.)

    [EDIT: Cross-posted with above.]

  8. Tsnom Eroc says:

    Well, here is a little breaker if one was doing utilitarian calculations.

    “3. Would you prefer:

    Option A: A world with 100 trillion trillion sentient beings, all of whom are miserable, but not quite so miserable that they wish they were never born.

    Option B: A world with 1 million sentient beings, all of whom are happy and consider their world a utopia.”

    The problem with these questions is that an urge to survival overpowers natural utilitarian calculations. Its a values question, and a major one, but from that logical standpoint, the answer is clear.

    That’s a reason why the question does not quite make sense.

    • Scott Alexander says:

      I feel like this is silly quibbling which is contrary to the obvious spirit of the question.

      • DanielLC says:

        I would have voted B, because A sounds like it’s almost bad enough to drive people to suicide, which is still pretty terrible. If it’s just a life that’s barely worth living, then I’d go with A.

      • michael w says:

        In that case, I legitimately didn’t understand the “spirit of the question.” From my standpoint, “miserable” means net happiness {less than} 0, but “doesnt wish they were never born” means net happiness {greater than} 0.

        I assumed you were testing the Repugnant Conclusion rather than whether unhappy life has inherent value, and so answered as though they were all experiencing minute levels of net happiness, the opposite conslusion as wireheadwannabe, who a couple threads up, described coming to the same issue.

        Edit: possible bug report, when i tried to write a less-than symbol above, in the published comment it instead cut short the end of the paragraph.

        • Anonymous says:

          net happiness 0

          What’s your model?

          The way I understand these things, you have a bag with “good stuff” and a bag with “bad stuff” and net over 0 is when you have more good stuff than you have bad stuff. And however negative the net might go, you might still have some good stuff, just outweighed greatly by the bad stuff.

          • michael w says:

            I think we’re in general agreement. My comment was initially posted incorrectly (see edits), which i was still fixing when you saw it.

        • when i tried to write a less-than symbol above, in the published comment it instead cut short the end of the paragraph.

          This is probably caused by this blog engine, WordPress, using a variation of HTML for its comments. HTML tags like <i> and <blockquote> start with the ‘<’ character, so when you write < in text, it assumes the rest of the text is an unclosed tag with an unrecognized name. Browsers only show the contents of unrecognized elements (between tag pairs like <these>contents</these>). In your messed-up text, there is no content, so nothing is shown.

          To work around this, you can escape the characters with HTML’s entity escaping syntax:

          • ‘<’ (less than) is &lt;
          • ‘>’ (greater than) is &gt;
          • & (ampersand) is &amp;
  9. Shouldn’t cryonics be correlated with opposition to oblivion?

    • Scott Alexander says:

      Good point. I listed the absolute values and not the signs. I’ve fixed that.

    • Lysenko says:

      I guess I just overthink these things, because my thought process was:

      1) Last I checked in on it, the science was shaky and there was a non-trivial chance the companies are not in fact going to care for their charges in the way they claim, so

      2) I’d REALLY have to know and trust the people setting this system up before entrusting my future to my care. I don’t, and I have no reasonable prospect of doing so, as

      3) The last time I checked in, the cheapest option is more than my annual post-tax income, and my rate of savings is pitiful. I’ll be lucky to have a retirement, much less 28 grand for a roll of the dice.

      Therefore: Uninterested (until my finances improve, the tech improves, or both). But I’d say I’m pretty anti-oblivion.

    • LaochCailiuil says:

      Cryonics, immediately makes sense in the face of biological realities. It can be incredibly frustrating and depressing to see your body change progressively in a negative way without any choice for some, myself included. Especially when it’s related to things like reproduction, having kids etc. You can see the bus coming but you can’t jump out of the way.

      • Giroth says:

        Welcome to the human condition 🙂 We’ve only had a couple of centuries where the technology to slow/stop aging is even imaginable. Cue the sorrow about being the last generation to die before it arrives. Like the poster above, I would do Cryonics if my finances improve and/or the tech improves, or I am close enough to death to take the gamble. I know I might not get the chance due to accidental death, but those are the calculations.

    • Brightlinger says:

      I’m confused why the hypothesis was that cryonics would be linked to a preference for suffering over oblivion. I mean, I don’t think most people signing up for cryonics expect a life of constant suffering post-revival.

      I guess maybe a preference for suffering over oblivion would be linked to a stronger preference against oblivion in general?

      • Adam says:

        I’m pretty sure he means the latter, i.e. you’d expect cryonics to attract people are very, very averse to the idea of oblivion and will risk a lot and pay a lot for any shot at avoiding oblivion, which is about the only reason anyone should prefer suffering (assuming they’ve already reproduced, fulfilled their life’s purpose, or whatever other goals people have that require putting off oblivion for a while but not necessarily putting it off forever).

        • Anonymous says:

          This fear of oblivion… I just can’t wrap my mind around it.

          • Orphan Wilde says:

            Imagine your mind fuzzing, then turning to static, and only when there is only static around to notice, does it fade into nothing.

            Or alternatively, consider oblivion as an infinite pit into which sapient, self-aware beings fall, never to be seen or heard from again.

            Oblivion is alien in a way that mere pain is not; oblivion is a Cthulian afterlife by comparison to Hell, because with Hell, you have something to expect, and something that you understand. By definition, you don’t have experience in not-being; you do have experience in pain.

          • Adam says:

            I’d personally characterize my experience as “dread” rather than “fear,” but maybe that isn’t even right. It’s a difficult-to-describe uneasiness due to the fact that I cannot fathom the idea. The nature of oblivion being nothing, there is nothing to imagine, and my imagination can only contemplate things, not not-things. Faced with this, when I try to contemplate my own non-existence, it just leaves me feeling vaguely sick in the way seeing my own bone come out of my leg once did. It isn’t really fear.

            Fear of death itself seems to be different. That’s just an instinct, the same way you fear having a projectile racing toward your face, unless you’ve gone to an awful lot of trouble to acclimate to that experience so it no longer produces the adrenaline rush it is supposed to produce. Animals that can’t contemplate anything still seem to fear death, at least when it is imminent.

          • Anonymous says:

            @Orphan Wilde:

            Imagine your mind fuzzing, then turning to static, and only when there is only static around to notice, does it fade into nothing.

            I can imagine this as terrifying, but perhaps my experience with several loved ones who developed dementia and Alzheimer’s disease colors my feelins on the matter. Watching them lose their minds, becoming violent, unaware of who everyone around them was, and, worst of all, the terrifying moments of lucid clarity of who they used to be and what they had become was heartbreaking.

            Or alternatively, consider oblivion as an infinite pit into which sapient, self-aware beings fall, never to be seen or heard from again.

            Oblivion is alien in a way that mere pain is not; oblivion is a Cthulian afterlife by comparison to Hell, because with Hell, you have something to expect, and something that you understand. By definition, you don’t have experience in not-being; you do have experience in pain.

            But this is the natural order of things in my mind. The concept of an afterlife, to me, is far more alien than the seemingly natural process of ceasing to exist after death. To me, it just seems like wishful thinking, a way to cope with the fact that we’re all going to die and the world will continue without us.

            Think back to your earliest memories, they were of things “I” (you) experienced. In fact, all of your memories are of you as “I”, yet, clearly, you have changed year after year, decade after decade. And so long as you remain sound and whole of mind, you will always recognize yourself as “I”, no matter how much you have changed over the years. This line of thought predisposes you to believe (or hope) for an afterlife, but think about the years before you remember yourself. Before you were born. Before you parents were born, and their parents, and so on. If the thought that there was a time before you existed does not disturb you, then perhaps thinking about a time after you cease to exist should not disturb you. This is how I feel about the matter.

          • Orphan Wilde says:

            Dementia and Alzheimer’s are partial oblivion, incomplete oblivion.

            I’m not particularly terrified of oblivion, actually. As Cthulu-like as that prospect sounds, the prospect that truly terrifies me is that my mind might continue after death, a scattered pattern represented in instant after instant across an infinity of space and time and scale, in one moment instantiated in a cloud of helium ions in the dying moments of the universe, the next across the way a few trillion galaxies interact gravitationally a few billion years earlier, the only constant being continuity itself as everything human and “me” is slowly stripped away until only a flicker of consciousness remains, eventually becoming trapped in a static loop of continuity that never needs to resolve; consciousness flitting back and forth from galactic configuration to cloud of ions and back again forever.

            I wouldn’t know it. I wouldn’t know, or remember, anything; the continuity of self would careen brutally towards the simplest awareness possible, from instantiation to instantiation, as that is the most likely path for that awareness to take. It could feel, from the inside, like bliss; it could feel like torture. It could feel like nothing; indeed, that’s probably exactly what it would feel like, as anything else would be more complex. And nothing would ever change, forever.

          • Hyzenthlay says:

            Imagine your mind fuzzing, then turning to static, and only when there is only static around to notice, does it fade into nothing.

            Doesn’t this happen every night when we go to sleep?

            Though…honestly, even going to sleep does make me a little nervous for that reason; it’s an uneasy feeling to let go of my consciousness and essentially become nothing for awhile.

          • Happybara says:

            In my mind, it’s not the oblivion itself that’s scary (it can’t be scary, it can’t be anything). It’s the moment right before the oblivion. 🙂

  10. anonymous poster says:

    Recycled from the Open Thread:

    Would anyone else choose the ‘live a great life and die in ten years’ option over what they have now?

    • Evan Þ says:

      Yes. Absolutely.

      It’s probably significant that I define “great life” largely in terms of benefit to society, and that I’m sure (P>.99) that I’ll go to Heaven after said death.

      • Mongeese says:

        P > 0.99? That’s some serious faith.

        • Decius says:

          I suspect P=1, based on what I expect the reaction to evidence would be.

          • Evan Þ says:

            What makes you say that? I envision myself reacting to evidence pointing against God with increased doubt, which models exactly a high prior being updated on evidence. This’s exactly what’s happened multiple times in the past: I doubted, but eventually concluded that God’s existence was still overwhelmingly more likely.

            Now, of course, I don’t envision ever updating my prior under 50%; when I ask myself what might cause me to do that, I answer with what I take to be ludicrous impossibilities. But, that’s exactly what a high prior looks like.

      • Giroth says:

        You stated your reasoning for God’s existence. However, a capital P Heaven would have a different set of priors than “any type of God existing”. What is your probability of a Heaven you would like to attend, conditional on a God in all possible space of God’s existing?

    • Nadja says:

      I can’t imagine many people with young children or any other similar serious responsibilities choosing the short life. Sense of duty seems to confound things a bit here.

      • Evan Þ says:

        “For all questions, please assume that results will have no effect on any other people, including your friends and family. If it helps, imagine your mind has been copied, and these questions will affect a simulated version of you while your real friends and family are unaffected. Please answer them solely according to your own selfish preferences.”

        (But in the real world, for most definitions of “great life,” I agree.)

        • Koldos the Shepherd says:

          This is not actually a great way to cut the “moral intuitions” reality at the joints.

          I think qua Haidt that a major function of moral behaviour is social in nature (to enhance cooperation), so the entire enterprise of discerning “selfish intuitions” is bound to fail because you do not actually have any intuitions for the weird mind cloning hypothetical.

        • Adam says:

          Asking people to imagine their deaths would have no impact on their dependent children is not going to change the fact that the reality of the matter is going to skew the results. This is one of the basic weaknesses of hypothetical questions. Ask people to believe increasingly unlikely counterfactual worlds for the purpose of a thought experiment and an increasing proportion of them are increasingly unlikely to be able to do it.

          • Giroth says:

            Completely agree with both of the above comments. You can see the frustration as commenters try and boil down “what the words really mean”.
            Moral “intuition” is based on so many factors, both existing and hypothesized results, that Bayesian reasoning breaks down hard. You would need Descartes demon for two rational beings to resolve a thorny moral dispute using Bayesian updating. It is useful in many contexts, however it has become abundantly clear it doesn’t work in soft and fuzzy contexts. (Evidence: This very comment section, most of the Less Wrong explorations).

          • Loquat says:

            Plus the converse – our loved ones have an effect on us! I’m expecting my first child later this year, and I would strongly prefer to survive long enough to see her, and any other children I may have in the interim, reach adulthood. Living long enough to see the kids find spouses and produce grandchildren would be pretty nice, too.

      • anonymous poster says:

        “Sorry, imaginary kids, daddy made a pact. I’d say ‘see you on the other side’ but I wouldn’t allow the afterlife to exist unless I had a 0 percent chance of going to hell, so this is it. Kind of a shitty deal, but hey, in the counterfactual world you might not exist at all. Be good now.”

        • jeorgun says:

          I mean, personally, thinking about dependents makes me way more likely to choose oblivion. I’m much more comfortable facing the idea of consigning myself to eternal torture than I am the idea of inadvertently forcing it on my (hypothetical) children.

      • Garrett says:

        I’m single and very much want to have children. If I had 10 years of relationship success I could achieve that. Even if I only manage to see part of their lives, that’s better than decades of unfulfillment.

        • Adam says:

          You’re saying that specifically as a person who does not have children, which is counter to her hypothetical.

    • FeepingCreature says:

      No way. I anticipate my life getting hugely greater in the obligatory thirty years. Right now, length is the most important factor.

    • Anonymous says:

      Would anyone else choose the ‘live a great life and die in ten years’ option over what they have now?

      Should I accomplish my goals in life, retiring to Heaven? These terms are acceptable.

      • Winter Shaker says:

        Yeah, question 2 really needs to be disaggregated into ‘…conditional on then getting to go to Heaven’ and ‘…conditional on there being no afterlife’.

        Would you answer differently under these two conditionals?

        • Anonymous says:

          Yes. I would choose long unhappiness over oblivion. I would choose damnation over oblivion.

          • Winter Shaker says:

            Even if the hell is of a traditionally conceived unbearable pain, fire-and-brimstone, worm-dieth-not variety, which I suspect was intended to be implicit in the question?

            I will grant you that a lot of modern religious people have come to the sensible conclusion that such a thing is simply incompatible with an omnimax god, and therefore posit an eternity-of-meh, absence-of-a-god-you-didn’t-even-like variety instead. [Edit – and of course that presupposes that any gods that exist *are* in fact omnimax, rather than a bit of a jerk or outright evil, and there’s no good reason to assume that hypothesis] … But the traditional unbearable torment variety still has a lot of adherents, and I’m surprised at the idea of there being no conceivable torment that could be worse than peaceful non-existence.

          • Anonymous says:

            Even if the hell is of a traditionally conceived unbearable pain, fire-and-brimstone, worm-dieth-not variety, which I suspect was intended to be implicit in the question?

            I should hope I prefer that. While in Hell, I might reconsider, but then this is why you should make commitments while still sound of mind, rather than not.

            peaceful non-existence

            Non-existence has nothing to attach a qualifier like “peaceful” on. It’s like trying to set a property on a null – all you will get is a Null Pointer Exception. Non-existence is literally nothing. Anything is better than nothing.

          • Winter Shaker says:

            Well, each to their own, I guess. I can easily conceive of circumstances unpleasant enough that, if they were guaranteed to be never-ending, I would prefer not to exist.

            I mean, it would have to be a lot worse than merely having to work at an eternal McDonalds – maybe something like neverending unanaesthetized dental surgery would be about the cutoff point – but I find entirely alien the idea that no conceivable suffering-with-no-prospect-of-alleviation could be worse than simply having no further experiences at all.

          • Winter Shaker says:

            Actually, on thinking about it, neverending unanaesthetized dental surgery would be well past the cutoff point for ‘preferable not to exist’. Not sure exactly where I’d draw the line though.

    • LaochCailiuil says:

      ‘live a great life and die in ten years’

      Such a wireheady, nebulous concept. Feels like trying to imagine 5 dimensional objects.

      • Ted says:

        Really? You can’t imagine something that’s great and fulfilling that lasts for fewer than 10 years? Am I misunderstanding you?

    • ThaddeusMike says:


  11. Man, does anyone else just have no idea what to answer when faced with questions like this? I didn’t take the survey because I knew roughly what kind of questions would be on it, and I knew I wouldn’t have good responses.

    • Not quite the same thing – I started the survey late at night, realized I’d have to think carefully for my answers to be consistent, and gave up, intending to return to it when I was more alert.

    • Immanentizing Eschatons says:

      Yeah, I started to take it and had to think a lot- I really don’t know what I want sometimes. (I wasn’t able to finish it)

    • ThaddeusMike says:

      I was able to answer all of the questions except the one about Heaven and Hell. Even after reading the instructions, I still wasn’t sure how to answer. The rest, however, I was able to answer easily, even if I didn’t submit due to never figuring out the one question.

    • Me. When I think about moral philosophy I always get the feeling that I’m just picking my opinions arbitrarily—like, I’m supposed to read these questions, introspect and discover what my intuition is, but upon introspection I find that I don’t have any relevant intuition. I didn’t take the survey either.

  12. Rauwyn says:

    Scott, for your first hypothesis, your graphs don’t actually show how strong the opinions were on either side. Also, if a 90%/10% split really counts as “people would be split in their answers to these questions” that seems like a very weak hypothesis.

  13. Bayen says:

    We did not adjust for multiple comparisons.


    • J says:

      It’s a joke about the previous post to this blog.

      • Agronomous says:

        YES; also, whenever you see: “We did not adjust for multiple comparisons” you should read it as “We did not feel like adjusting for multiple comparisons”.

  14. anon says:

    “Option A: A world with 100 trillion trillion sentient beings, all of whom are miserable, but not quite so miserable that they wish they were never born.”

    I find this world literally unimaginable. That is, I try to imagine being that miserable, and I notice that when I’m that miserable, I tend to snap at other people. If I snap at my boss (hypothetically; I don’t presently have a boss) enough times, I’ll get fired, and then I’ll be more likely to wish I had never been born. Multiply this by the number of interpersonal interactions in the world, and it’s hard to imagine this as a stable state.

    Also, some folks seem to prefer a life with ups and downs to a life of meh. Which is not really a possibility accounted for in your questions.

    • Malthusian situations with nearly identical beings would seem to result in this and be reasonably stable.

      • anon says:

        I have no idea what you are imagining when you talk about a Malthusian situation.

        Also, if we are talking about “nearly identical beings”, that’s at least for me a dramatically different scenario than just “there are lots of people” (see e.g. this blog’s “Answer to Job” for Scott’s similar intuition) .

    • Decius says:

      I can’t imagine that number of people.

      I know that it’s a million times more than a hundred million trillion, which is a million more than a hundred trillion, and so forth, but by the time I’ve stepped down to a number of sentient beings I can imagine I’ve lost track of how many times I have to multiply by a number that I don’t natively understand.

    • jes5199 says:

      I’m sorry you got fired but that’s no reason to be so hard on yourself.

  15. To me the main issue with considering these questions (especially 3) is that it would seem that different experiences are different and that is important.

    I.e. 100 beings experiencing completely different (orthogonal) types of happiness and experiences valued at one util is drastically more valuable (worthwhile?, moral?) than 100 copies of the same being simulated concurrently experiencing one util of happiness/experience.

  16. Swagbunny says:

    If it helps explain this, when I took the survey it was as someone looking for ‘gotchas’. I would imagine that was a fairly common approach among your readers.

    My main criticism, as a ‘somewhat unhappy’ person who picked the long life of unhappiness over the ten years of happiness, is that there is a major difference in scale between these and the McDonalds question.

    The “constantly McDonalds, for half a century” question described a literal Hell, albeit a minor one on the scale of Hells. You describe something vastly worse than, say, working in a sweatshop. Of course I was going to pick immediate painless death.

    The “lifetime unhappy” question, on the other hand, implied “I am in perhaps the mentally worst-off 5% of the population in the Western country I live in”. I’m probably one of the 5%, and I don’t believe that we’re living in a literal Hell.

    • The “constantly McDonalds, for half a century” question described a literal Hell, albeit a minor one on the scale of Hells. You describe something vastly worse than, say, working in a sweatshop. Of course I was going to pick immediate painless death.

      Not necessarily! You just need to drag every customer that enters into the back room, make them into hamburgers, and when you’ve run out of customers, you can spend the rest of your lifetime shift manning the register and staring into space thinking about stuff.

      OK, depending on how many customers you end up murdering and repurposing into foodstuff in order to literal-genie your way around “McDonalds work allowed only”, you might still be in a hellish situation, but still!

      • suntzuanime says:

        I would certainly take immediate painless death over engaging in the mass murder of innocents.

        • Patrick says:

          They’re shoppersat the Hell-McDonalds where food is served by actual slaves who struggle to work as long as possible because only euthanasia awaits if they falter.

          • suntzuanime says:

            Well, there is certainly a case one could make that everyone who participates in the exploitative consumer-capitalist system is guilty of abetting exploitation and therefore fair game for vigilante execution.

          • hlynkacg says:

            Congratulations you’ve invented Bolshvism.

          • Agronomous says:

            They’re shoppersat the Hell-McDonalds where food is served by actual slaves who struggle to work as long as possible because only euthanasia awaits if they falter.

            Plus all but the first wave are literally cannibals!

            (What’s the opposite of fighting the hypothetical? Fighting for the hypothetical?)

          • I’m not sure what McDonalds you’ve been in, but in the ones I’ve been in, everyone can quit.

            There’s a difference between having to work for a wage (without which you can work another job, or beg, or live off savings, or apply for a number of aid programs, or or or or) and having to work because you’re compelled to and will die when you can’t do it any more.

            Plus, after the first few EM McDonaldses turn into charnel houses, I bet the Powers That Be will just go “Fuck it. Can someone just code up a non-sentient expert system for this?”

      • FeepingCreature says:

        Sounds like it’d make for a good musical.

        “Just the thought of it’s enough to make you sick! (And I’m telling you, them customers is quick!)”

        • FullMeta_Rationalist says:

          I believe this is the plot of The String Of Pearls.

          Incidentally. The name of the actor who played Sweeney Todd in the 1936 film adaption was Tod Slaughter. Nominative Determinism!

    • Adam says:

      I’m surprised that so many commenters have had this response. I see the “unhappiness forever” as clearly worse than “McDonald’s forever.” The main reason McDonald’s cashier’s are unhappy, assuming they actually are, is they face a life of severe income and wealth insecurity where the risk of being thrust into true misery is forever right around the corner. Scott has eliminated that from the equation and basically asked you to become Lester Burnham and just take a job with no responsibility that you don’t even actually need. That doesn’t sound bad at all. Since it seems to be the case that you can’t be fired from this job, you don’t even need to put up with shit from bad customers and bosses.

      • Anonymous says:

        I think the risk part was implied – when you stop being able to do your job, you die. So work hard, don’t slack off – if you do, you are fired (die).

        • Adam says:

          That makes the question pointless then, as the first answer includes the second. You can choose to opt out of McDonald’s and die painlessly at any future moment you decide you made the wrong choice. In that case, every person should answer #1 as “option A plus option B if option A turns out wrong” is always preferable to “option B and you can never take it back.”

          • HeelBearCub says:

            It depends on what we mean by “able to do the job”. There is an implication of a certain lack of free will, or some sort of motivating factor so strong that it remains your choice to work 16 hours a day, performing your tasks at a level of average competence.

            Until you get Alzheimer’s or become truly disabled somehow, you just keep working.

            It makes no logical sense without magic involved.

          • Anonymous says:

            Magic, or electrodes inserted into the brain.

      • Orphan Wilde says:

        “McDonald’s forever” is basically “Monotony and boredom escapable only by death”. The question is pretty clear that your life is McDonalds, and only McDonalds; you don’t have the time to do anything else.

        You’re thinking of a life of job security, but it’s not security that permits you to do anything with your life. No friends, no family, no nothing, just the same tedious work, forever.

        (If it helps you to visualize the situation, replace “McDonald’s” with “Sisyphus”, as it’s just a thematically updated Sisyphus situation.)

        • Adam says:

          I still feel like I can find meaning and value in whatever the heck it is I’m doing, even if I have no choice but to do it. I used to cashier when I was a teenager. I actually enjoyed it most of the time, especially when it was busy. I competed with myself to see how quickly I could push people through, making sure my line was always moving fastest. Even if I was Sisyphus, I can see that not being terrible. I have spent plenty of my life voluntarily running up hills, sometimes even pushing heavy objects. There is a euphoria to the accomplishment. Exercise is enjoyable. It’s only considered hell because it’s a metaphor, not because the physical reality of pushing a rock up a hill is actually a bad reality.

        • jes5199 says:

          We must imagine Sisyphus happy:

      • Anonymous says:

        That question specified working 16 hours per day. Since I need more than 8 hours of sleep, a lifetime of continual sleep deprivation would be (for me) a fate worse than hell.

      • suntzuanime says:

        Have you ever worked a 112 hour week? It is enough to make one unhappy.

        • Adam says:

          Having spent seven years in the Army, I’m probably one of a few people here who can say I actually worked a 112 hour week. I’m sure Scott has, too, being a medical intern. Sure, it sucks. It would have been much better as a cashier than what I was actually doing, though.

  17. FullMeta_Rationalist says:

    4. What percent certainty of going to Heaven would you need before you would prefer a world with both Heaven and Hell to a world where death ends inevitably in oblivion?

    A long time ago, I was reading (on reddit?) about how people discuss theism with others. A {teacher?} said “People have different views of who God actually is. So when students ask if I believe in God, I ask them to share what their idea of God entails.” Heaven is similar: it’s a vague construct that everyone has their own ideas about. E.g. does your conception of heaven include 72 virgins?

    My personal conception of heaven: it’s great and it lasts forever. Literally forever. That’s actually a problem for me. I like immortality if it’s the kind where I get to decide when I die. But to be *forced* to keep experiencing the universe unconditionally and indefinitely, that sounds really scary. Which kinda contradicts the “heaven is unconditionally good” clause.

    So anyway. When I sit down and consider infinity, I realize it’s really frightening. Consider the number of all the atoms on planet Earth; infinity is bigger. Consider the number of grains of sand it would require to fill the visible universe; doesn’t hold a candle. Consider taking the Busy Beaver Function of the Ackermann Function of Graham’s Number; not even close, baby. Infinity is only ever imagined — poorly. When an Elder God whispers something wicked and thereby drives Bob insane, I imagine it’s because Bob lost his mind within Cantor’s Paradise or the Wood Between The Worlds.


    I once had a math professor who said “When I was younger, I used to think that after I died, I would know everything”. The context was that he’d digressed from the calculus lesson to advertise his set-theory class. Specifically, he was explaining to us Cantor’s insight that different cardinalities of infinity exist, and that our intuitive notions of inequality no longer applied.

    Regarding this “know everything after death” meme, I’ve had the same thought cross my mind when I first watched Outlaw Star. Which means it can’t be unique. Has anyone else had similar thoughts?

    • Tyrant Overlord Killidia says:

      “My personal conception of heaven is that it’s great and it lasts forever. Literally forever. That’s actually a problem for me. I like immortality if it’s the kind where I get to decide when I die. But to be *forced* to keep experiencing the universe unconditionally and indefinitely, that sounds really scary. Which kinda contradicts the “heaven is unconditionally good” clause.”

      Along with this… it sorta implies that there’s no free will in heaven.

      What if you like watching Denny’s fights on YouTube, will you be allowed to do this in heaven? Or will your personality be drastically altered? How many other “undesirable” personality quirks or enjoyments would be possibly scrubbed upon entry?

      I’m pretty sure there are a ton of dystopian fictions with this sort of theme. Like the movie Equilibrium, where everyone’s emotions are chemically suppressed for the good of society.

    • MugaSofer says:

      >But to be *forced* to keep experiencing the universe unconditionally and indefinitely, that sounds really scary.

      Some people think that people rejecting Heaven is where Hell comes from, and some people think Hell is oblivion, so there’s at least a segment of the population who think God agrees with you.

      Me, I don’t understand why you’d want the option of dying in a place known to be unconditionally good. It can’t be for utilitarian reasons, or fear of torture or illness, so why?

      Would you be OK with a world where you can die at any time, but God is clever enough to arrange things so you never want to?

      • FullMeta_Rationalist says:

        I guess I’m complaining about how heaven is marketed.

        Remember that meme about the guy who tried to ban Dihydrogen Monoxide? “It can burn you, freeze you, bruise you, suffocate you, zap you, it’s a component of acid rain, etc.” And everyone was like “WOW that sounds super dangerous! Why are my kids allowed to drink out of styrofoam cups that were manufactured with Dihydrogen Monoxide?” The point is that it’s possible to use marketing to pervert our preferences. I, however, remember enough from chemistry class to recognize an alias for water when I see one. So if I were asked to petition the ban, I’d respond “Uh, this smells fishy. Isn’t that just water?”

        Heaven is like the dual of Dihydrogen Monoxide. When an angel says “Come inside! Heaven is super cool and you’ll live forever!”, that raises more red flags than 1956 China. “Live forever?” “Absolutely!” “Sorry, I’m allergic to forever. I can opt out whenever I want, right?” “Mortality isn’t really a thing in Heaven, so uh… umm… but don’t you wanna be with God?” “Look man, this whole matrix thing isn’t for me. Just give me the Red Pill.” “But look at all these utils, wh– SECURITY! SECURITY! BAR THE GATES. WE’VE GOT A RUNNER.”

        The ability to opt out of a system is something I do not take lightly. Mushashi’s Book of the Void says that the art of sword fighting is not limited to his particular system. The U.S. Constitution contains a clause which says that the government deserves to be overthrown should it prove an incapable system [0]. The Archipelago’s emigration clause is the most important rule. Programming languages provide escape sequences. There’s an escape button on every keyboard. Division by zero returns NaN. Hanson’s post regarding “Against DWIM Meta-Law. These are all sensible failsafes for when the system inevitably fails. Any system that doesn’t handle this gracefully is poorly engineered. (In case you were wondering, yes I support Euthanasia.)

        So when heaven is advertised as “once you’re in, there’s no way out; this is a feature not a bug”, that sounds like pyong yang (which means “heaven on earth”).

        Would you be OK with a world where you can die at any time, but God is clever enough to arrange things so you never want to?

        Yes. This is exactly optimal.

        [0] I believe the Right of Revolution clause exists somewhere, but I can’t find it at the moment.

        • FullMeta_Rationalist says:

          The irony that is Pyongyang exemplifies the essence of advertising: Advertisers advertise exactly what they’re not. Airplane travel is cramped, noisy, boring, etc. So of course the airlines market themselves as the epitome of comfort. Automobiles emit carbon dioxide. So of course automakers advertise themselves as Going Green. I think this is what Jacques Derrida kept going on about.

          • Jiro says:

            Advertisers don’t advertise what they are not. Advertisers advertise what people believe they are not. This belief could be true or false.

        • Acedia says:

          Could we not be confident that the being who designed humans literally from the ground up, and is omniscient to boot, would know (probably better than they do) how to make them happy?

          I’m an atheist so I don’t believe any of it, but I find it strange when my fellows make arguments that rely on the premise that God might be mistaken about the way humans work. If you’re arguing within the context of the theology, that isn’t possible.

  18. Michael Sadowsky says:

    The only correlation with any other question was with the one hundred trillion trillion sentients question, which is basically asking the same thing in different words.

    I have to say, that’s not how I took the hundred trillion trillion question at all. By discussing the individuals as miserable, I took them to have negative utility. Under that conception, a total hedonistic utilitarian would prefer the 1 million beings, but a total preference utilitarian would prefer the hundred trillion trillion. However, a total hedonistic utilitarian would still believe creating new beings was just as good as improving the lives of existing ones.

  19. Ivy says:

    Apologies for the rather basic question, but I’m struggling to understand

    Creating a life that experiences 100 utils is exactly as good as improving existing lives 100 utils

    What does this sentence mean in the real world? It seems like shorthand for something precise, but when I try to understand the implications of assigning a truth value to it, I get “type error”. I understand utility (by VNM) as defined with respect to an agent’s preferences over outcomes, but defined only up to a constant factor… so in what sense can we compare the utilities of different agents? How can we measure that a life we created has gained as many utils as an existing life?

    Or should we understand “utils” to mean “some implicit, commonly agreed upon metric of human happiness” like QALYs?

    • Immanentizing Eschatons says:

      I generally understand it as referring to the utility function of a third party, utils here then doesn’t refer to the utility functions of those being created/improved, but to the survey taker’s.

    • FrogOfWar says:

      The relevant notion of ‘util’ is the one from utilitarianism, a doctrine that predates von Neuman’s work by over a century. One of the core assumptions of utilitarianism is that individual well-being can be assigned a number that aggregates across individuals to give a general welfare level. This then determines the best state of affairs and what agents should do.

      Traditionally, they have taken this well-being to be a function of pleasure and pain or preference satisfaction. But you can substitute whatever you think captures the idea or choose answer (E) if you think nothing does.

      • Ivy says:

        But if we assume the existence of an aggregative utility, doesn’t the sentence again become a tautology? Why would I prefer one way of getting 100 utils over another?

        • FrogOfWar says:

          If you’re a classical utilitarian, you won’t. But this feature of classical utilitarianism gives rise to paradoxes like the ones discussed in the post that motivate people otherwise sympathetic to it to amend it.

          One thing you can do that largely maintains the spirit of classical utilitarianism is to retain the theses that:

          1) All that matters is the distribution of well-being in individuals.
          2) Everyone’s well-being counts equally.
          3) The goodness of a state of affairs is a strictly monotone increasing function of the sum of individual well-being. (probably should have thought of a way of saying this that didn’t make the first two redundant).
          4) The right action is the one that brings about the best state of affairs.

          This would give different results to the two scenarios while still being clearly consequentialist and possibly utilitarian depending on which definition is being used.

          In any case, there are many ways to not be a utilitarian or even a consequentialist while still believing that you can rank worlds/states of affairs by goodness and that this ranking depends only on individual well-being. These doctrines can all make sense of utils while differing on the answers to the relevant question.

          • Ivy says:

            I like that framework, though I think by

            strictly monotone increasing function of the sum of individual well-being

            you mean “increasing function of each individual well-being”, otherwise your actions will be identical to a classical utilitarian’s.

            I guess in this framework, the original question unpacks to “For your choice of well-being measure, for some unspecified amount of increase in that well being, would your choice of aggregation function be larger if that increase happens to an existing or a newly-created being?”.

            Which is meaningful, but has so many implicit assumptions that I’m not sure the distribution of answers can tell us much about people’s real-world preferences.

          • Adam says:

            One thing it’s odd to never see come up in discussions of utilitarianism is distribution attributes other than the mean and sum, especially given how often it comes up in virtually any other question of optimal resource distribution.

          • FrogOfWar says:


            Nice catch, though I’m not sure your proposal fully captures what I was trying to get at since it may not rank a world as better that does nothing but add a new life with positive utils.

            (or maybe it does? Maybe you can treat treat the utility function as taking tuples with utility values for all possible people and assign a value of zero to all the ones that aren’t actualized in a given scenario? I’d have to think about it more).

            But even if that problem couldn’t be fixed, there’d still be the example of a deontologist who agrees with the classical utilitarian on ranking worlds by goodness, but who disagrees with the claim that you should always try to bring about the best world. For example, maybe it’s impermissible to bring about the best world if you would have to torture to do so.

            So there would still be people for whom talk of interpersonal utils made sense but who don’t necessarily have to regard Scott’s two scenarios as morally equal.


            There’s been some discussion of alternatives in the literature (like here and here) but probably not as much as there should be. Possibly because normative ethicists tend not to be consequentialists.

          • Jiro says:

            For example, maybe it’s impermissible to bring about the best world if you would have to torture to do so.

            You could also say that the value of a world is based on its historical state as well as the number of people currently suffering, and that “created by torture” highly reduces a world’s value.

            (Of course if you can do that, in the limit, utilitarianism can be made equivalent to any other system of ethics you want.)

    • Wilj says:

      We’re meant to assume there’s some foolproof way to measure and compare utilities, in order to get at the intuition of whether “creating a new sentience” has any value itself. If you don’t there’s a way to do so even in a most convenient hypothetical world, answer E.

    • Earthly Knight says:

      I understand utility (by VNM) as defined with respect to an agent’s preferences over outcomes, but defined only up to a constant factor… so in what sense can we compare the utilities of different agents?

      Note that diachronic intrapersonal comparisons of utility are vulnerable to the same objection. You might decide to shrink or enlarge your utility scale at any moment! So if this problem proves fatal to interpersonal comparisons of utility, it is doubtful that the notion of utility can be put to any work at all.

      • Ivy says:

        You’re right, the VNM construction of utility seems fundamentally unable to handle interpersonal comparisons – so clearly the “100 utils” of the question refer to some different notion of utility which I am not familiar with.

        I think

        it is doubtful that the notion of utility can be put to any work at all.

        is a bit strong; you can derive all of game theory and decision theory without ever making interpersonal comparisons of utility.

        • Earthly Knight says:

          You misunderstand. I said that diachronic intrapersonal comparisons of utility– that is, comparisons of the utility functions of one individual at two different moments in time– will fall to the same objection. There’s nothing to stop you from deciding that your currently utility scale is too cramped and that you want to multiply it by some big scalar. But then a comparison of your utility functions before and after the change will inherit all of the troubles that attend interpersonal comparisons. The only way decision theory can avoid these problems is if we make it into a time-slice theory, applying only to snapshots of human beings frozen in time. But it is doubtful that we can make sense of decision theory under this constraint, if for no other reason than the impossibility of deciding or acting in an instant.

          • Ivy says:

            You’re right, I misread “intrapersonal”. Yeah, if you model a person as a series of agents with different utility functions then VNM can’t handle intrapersonal comparisons – but decision theory that models a person as a single agent is still plenty useful.

          • Earthly Knight says:

            I don’t know what you mean by modeling someone as one agent or a series of agents (do you?). The problem is that, whenever I am faced with a suite of options, I must take into account that my utility function might undergo a positive affine transformation before I have the chance to finish executing any of the options. This is exactly parallel to the problem in the interpersonal case; if the one is solvable the other is.

            In general, all objections to comparing or aggregating the utility functions of different agents will apply with the same force to comparisons between different time-slices of the same agent. This is because Pete’s preference at t1 and his preferences t2 are joined by a relation of causal continuity, but the decision-theoretic formalisms are indifferent to causal histories.

    • Said Achmiz says:

      This is, in fact, a fairly fundamental question — one that I have been asking utilitarians for quite some years now, and have never seen answered to my satisfaction. (It’s one of the main reasons I’m not a utilitarian.)

  20. Jiro says:

    Whether you believe everybody has a right to commit suicide if they want, including people who are not terminally ill.

    Don’t ask questions about “everybody”. Some people will read it as “literally everybody” and other people will read it as “given that they are otherwise competent adults”.

  21. Pku says:

    Wait, are there really people who wipe standing up? Or is that just hyperbole?

    • FullMeta_Rationalist says:

      I, for one, eat my toast butter-side down.

    • Anonymous says:

      How do you even wipe sitting down? You’d have to do some pretty awkward maneuvers.

    • gwern says:

      The correct answer, of course, is to reject the dichotomy, stop living like a filthy barbarian, and use a bidet.

    • Loquat says:

      I’d like to know the gender breakdown of sitting-vs-standing. As a female, I can tell you that anyone who urinates sitting down is going to be at risk of wayward drips if they don’t wipe before they stand up. And once you’re accustomed to wiping the front while sitting down, it’s not hard to reach a little further and wipe the back that way too.

      • Aapje says:

        Both the front and rear sitting wipe seem to be much easier for a woman. In both cases the man has to reach further away from the body, increasing the chance of contact with the porcelain (shudder).

        Also, a man can ‘pre-wipe’ in the front by shaking it, greatly reducing the amount of remaining urine.

        • arbitrary_greay says:

          How far back are you people sitting on the seat that wiping sitting down doesn’t have about 99% chance of touching porcelain and/or bowl contents!?

          • Loquat says:

            That virtually never happens to me. Maybe the toilets in your neck of the woods are just too small?

            Also, I think it’s more a function of vertical space than horizontal space anyway.

          • Aapje says:


            Men are bigger than women on average, so on average men have to reach further back anyway, for a nr 2 wipe.

      • Adam says:

        I have never wiped after urinating. I’d be surprised if any man ever does that. Why do you think we consume toilet paper at 1/10th the rate you do?

        • Loquat says:

          Well sure, but my point is that since we get used to wiping sitting down after one excretory function, it’s easy and logical for us to wipe sitting down after the other.

        • Winter Shaker says:

          I have never wiped after urinating. I’d be surprised if any man ever does that.

          Allow me to surprise you then 🙂
          I far prefer to be guaranteed dry than merely probably dry.

    • Tok Nok says:

      Are there really people who don’t wipe their ass while doing a handstand on the toilet? Its the best method to aid gravity, you know.

  22. Katja Grace says:

    >Fifth, happier people would be more likely to prefer suffering over oblivion, because they view life as generally excellent and so oblivion represents more of a sacrifice for them.

    Interesting, that was opposite to my prediction (which was apparently wrong): people who are less happy mostly still don’t find themselves suicidal, and so they are forced to have a narrative which says happiness is less important, and for instance there are other valuable things in life. So unhappier people prefer life, even if it is unhappy. If I had to guess how my own thoughts on population ethics might be biased by my experiences, something like this is one of my top candidates.

    I took the survey, and was unsure when you talked about ‘happiness’ whether you meant all positive feeling, or even all preference that isn’t necessarily reflected in affect. I figured you were just talking about a class of emotions, and that for instance it would make sense to say I prefered a long unhappy life, since I might still achieve much of value to me. But then one option says, “Until then, you do fulfilling work, have happy relationships, and meet with success in most projects.” Which suggests I should imagine that in the bad option, not only am I unhappy, but that I don’t get stuff done. Which seems much worse.

    I think I also answered the questions about the strength of my feelings wrongly. I sort of took them as ‘how much better is A than B?’ when probably you meant something else like ‘how likely are you to change your mind?’ or ‘to what extent do you literally have feelings about this?’ or ‘to what extent does one of these options not seem terrible?’ or something.

    • canebrake says:

      Second (original) paragraph exactly seconded.

    • Aapje says:

      Your opinion is consistent with the experience of many doctors that people are much more willing to endure suffering when they get to the actual choice between death & suffering, than earlier when presented with the hypothetical.

      The desire by many young people to not grow old is also an example of people thinking that they can’t deal with unimaginable unpleasantry, yet losing that belief once they see that one can find the good in the bad.

      • Cadie says:

        And growing old – as well as suffering from many diseases – often comes on a little bit at a time, which makes it easier to adjust to than an abrupt change from 18 one day and 80 the next would be.

        I remember being a little kid watching “Product 19” commercials and being puzzled when older people said they felt 19 again. To me they were all kind of old and it made no difference except that at 19 they had nicer hair. And later on, as an adolescent and very young adult, I thought being old and even middle-aged would be mostly awful.

        Now I’m 37, which is not old, but it’s definitely older than young adult, and enough that 19 is pretty damn young in comparison. And it feels fine, a lot better than I thought it would. Because I didn’t jump straight to it; I’ve had roughly half my life to transition slowly from a brand-new adult to early middle age (or whatever you call the 35-40 period). Time to organically adjust to the bad parts and make the most of the good parts. So I’m far less able to pull all-nighters than I could fifteen or twenty years ago; no big deal, I’ll just make sure I get some sleep and if that isn’t possible for some rare reason I’ll carve out some time to take a nap the next day. Worsening eyesight, at present, only means I need reading glasses for computer work and other up-close work to prevent headaches. The gray hairs don’t really bother me and I figured out that if I want them gone for whatever reason, an $8 box of hair dye gets rid of them for awhile. Most of the stuff that c. 1996 I thought would be awful turned out to be only a little bit annoying. And there are good bits too; pressure to follow fashion trends has eased somewhat, salespeople take me more seriously, etc. I predict that, if no catastrophic health problems crop up (and they may not, my mother is almost 60 in good overall health and I seem similar to her in terms of blood pressure / heart rate / cholesterol genetics), in another 20 years it’ll be the same thing: I’ll be dealing with some problems I don’t have now, but they’ll come on gradually and I’ll adjust as they set in, and gain a few new benefits too.

  23. Dude says:

    Interesting. I have been thinking about oblivion vs suffering in terms of animals. That is, a lot of ‘intellectuals’ want to reduce suffering for animals so that they can still rationalise killing animals for food. I wonder how this correlates to their ethics for humans ie support of euthanasia etc.

    I find it funny how society refuses to let humans die if they are suffering, but implores reducing suffering for animals that are going to die anyway.

    • Adam says:

      Society is a big group. I doubt it’s identical subsets doing these things you think are logically inconsistent with each other.

    • MugaSofer says:

      Society also implores reducing the suffering of humans who are going to die anyway, so I’m not sure what the contradiction you seem to be professing is.

    • Aapje says:


      There is no inconsistency since killing animals doesn’t produce long term pain, while making animals suffer does.

      One can consider a brief moment of suffering by an animal to be acceptable to provide humanity with products of it’s carcass (not just food, btw); but not consider it acceptable to kick a cow, especially as doing so has no benefits to humanity or the cow.

      So in short: I think that you are comparing apples and bananas.

    • keranih says:

      I find it funny how society refuses to let humans die if they are suffering, but implores reducing suffering for animals that are going to die anyway.

      You are combining serveral different things here.

      Society as a whole looks to reduce suffering for all things – including non-sentient things like lawns that want for water. (No, we don’t have rules about plant welfare, but in my not universal experience, people like seeing plants respond to water and don’t like seeing infections and drought.) In this, we don’t treat animals and humans that differently. (There are differences based on legal independence and the biological function of animal-based protein, but that’s a modification of the basic principle of don’t make things suffer.)

      The instances where one is permitted to cause the death of humans are MUCH more severely limited than the situations where one is permitted to cause the death of non-human animals – but even the limits on deaths of non-human animals are pretty narrow. (Basically, it has to be a game animal of specific species, gender, age, and time of year, or it has to be an animal owned by the person killing it.)

      There are far more numerous and far more nefarious motivations for killing humans than for killing animals. A large part of the objection to normalizing suicide/euthanasia is a concern over the slippery slope of offing grannie *now* rather than after the turn of the year, when the new tax laws go into effect, or when her health care would have cost another $10K, and so on. So we restrict the opportunity to cause death, but it doesn’t follow from that, that suffering can’t be eased in other ways.

  24. Elephant says:

    I’m curious about the demographics of your respondents. Having kids, whose existence I find unimaginably valuable — much more than I would have predicted before having them — my answer to the new-sentient-beings question could certainly not be one of your first few choices. The question seems poorly phrased then, since there’s a difference between some “random” new sentient being, and some “particular” new sentient being, such as your own. Like another commenter, I wondered why the options seemed so limited and, to put it bluntly, so weird.

    • Adam says:

      He probably should have stated it something like “a hundred trillion trillion new sentient beings, that are nothing at all like humans, are created in some faraway part of the universe outside of your light cone and you will never personally know they exist.” Obviously, having children is not just creating new sentient beings. It is also acting to satisfy your own preferences and increase your own personal utility.

  25. Paprika says:

    In the first question, I took it to mean that if at some point, I was psychologically broken enough to be unable to continue working at McDonalds, it would treat me as being unable to continue and give me the painless death. Under these circumstances, I think choosing option A is always better.

    I can see now that is not what you intended but I am curious how many other’s thought the same thing…

  26. Throwaway Pseudonym e004ad02-67da-41c3-a90e-400fbef0963b says:

    I don’t think you can actually learn about population ethics in a posthuman world by getting humans to answer survey questions. As Carl Shulman has pointed out, the reason bad experiences are more intense than good ones is because an animal can lose all of its fitness very quickly, whereas the fitness gained from any one meal or copulation is much smaller, but that’s just a design constraint of evolution. There’s no reason for the asymmetry to be inherent to the nature of consciousness itself; the unimaginably good things in the multiverse are even more unimaginable than the unimaginably bad things. So any sane human who looks at question 4 of your survey will have scary thoughts and immediately type 100%. (Or perhaps, close the browser tab and quietly resent you for asking the question for the same reason I stopped reading Unsong.) But the real answer is probably 50% plus ε.

    • Koldos the Shepherd says:

      I think your point is very good and highlights some of the things that are ultimately inaccessible to anyone trying to build ethics around humans’ moral intuitions.

      This includes utilitarians or at least forces them to expend lots of effort on the level of modifying preferences instead of just bending non-sentient reality to satisfy sentient preferences.

  27. I have to wonder whether the answers to question 4 were at all correlated with whether the respondent had read The Broadcast or not. 🙂

    [That little symbol at the end was supposed to be a smiley. Doesn’t appear on my machine, guess I don’t have the right font.]

    • Subbak says:

      The Broadcast was definitely on my mind when I answered that I didn’t want even a tiny chance of hell, but honestly even without it I think I’d have answered 100%. It’s just very hard to imagine a heaven that could somehow compensate for the atrocities you can imagine in hell.

      • AlexC says:

        Hard to imagine? Yes. But why does it have to be hard to believe in things we can’t personally imagine?

        From my point of view, there are plenty of experiences good and bad which I can’t imagine. When I’m given good reason to believe that God is capable of making somewhere immeasurably better than we can imagine, I don’t find that makes me disbelieve in God, just happy that he’s made a way for us to get there.

        • Adam says:

          Your answers in this thread in general, not just this one, seem to indicate you think God can only be infinitely creative in the one direction but not the other. Why can’t the torment of hell be equally unbounded and unimaginable, to such an extent that knowing sentient persons will have to go through that is sufficiently evil to make the existence of heaven not worth the tradeoff relative to no afterlife at all?

          I’m not saying it is, mind you, but it’s a perfectly coherent view for people to hold who answered that way.

          For my part, I’d just say I’m indifferent between these outcomes. The reality is heaven, hell, and oblivion are all effectively concepts I can’t really conceive of, and trying to rank the goodness of states of being and non-being I can’t actually imagine at all in any way is pointless. I can rate the dread I feel at the thought of oblivion, but the reality is that I used to not exist, and as far as I can tell, it wasn’t bad. I can rate the best pleasure and the worst pain I ever experienced, but even a linear extrapolation of these doesn’t really seem to adequately describe heaven or hell.

    • Berna says:

      I see the smiley. 🙂

  28. Sniffnoy says:

    I will make the same nitpick as always: In question 5, where you say “consequentialist”, you seem to mean “utilitarian”. (That said, this really is a nitpick, as the question is pretty clear regardless, and any non-utiliarian consequentialist would probably find themselves picking the last choice regardless.)

  29. poignardazur says:

    … Dammit! Next time I’ll go to the toilets I’ll be aware and self-conscious how I wipe my behind! It’s going to be a really weird week.

  30. benwave says:

    Am I the only person who looked at question one and thought: “I have no real preference for one or the other of the two options”? I’d be interested to know for how many other people the difference in the utilities of these situations is about epsilon.

    I think in general I have a strong aversion to oblivion, to the point where I’d be willing to suffer an awful lot to avoid it. But that equation is predicated on there being a non-zero amount of happiness or fulfilment also present in my life. If there were no further opportunity for happiness, I find that all of my calculations collapse and I find it difficult to have any strong preference between two such worlds.

    • Adam says:

      I find it very hard to separate “preferences” as something abstract from the way I actually behave, and the way I actually behave involves an awful lot of unconscious and effectively automated actions. As far as I can tell, one of these default behaviors in continuing to exist. Most living creatures act in such a way that they continuing living except in rare circumstances where their death might benefit others that are genetically similar to them, and even then the behavior seems to be impulsive more than consciously motivated.

      The impulsiveness of most behavior makes these questions nearly impossible to answer for me. I’ve been awake for about two hours this morning. I don’t have anything I need to do for hours, so here I am. Why am I here writing this? Do I prefer it to other activities? Not really. I’m indifferent between this and about millions of other things I could be doing to waste a few hours, but I woke up, had a laptop within arm’s reach, and haven’t checked in here for a while, so here I am. There are so many random coin flips and so much path dependence leading me here, things having nothing whatsoever to do with conscious motivation or preferences, and that seemingly makes up the vast bulk of all behavior.

      Given almost any situation where there is a near-term prospect of anything other than unbearable pain and I’m probably not going to willingly end my own life. This is independent of whether I believe I have an abstract preference for oblivion over a pointless and joyless existence. Remove all motivation whatsoever and turn me into a creature of pure instinct and I’m still going to eat, sleep, and avoid mortal danger. This reality makes it hard for me to take seriously the notion that it is even possible to rank all possible world outcomes on a Likert scale. I can’t tell you for sure what I would prefer until I am actually presented with these possibilities and act. Even then, actions are largely done for unknown and unknowable causes and these questions require realities that are so foreign to actual human experience that they’re impossible to truly imagine. What does a world look like in which I am 100% certain of the future or whether heaven and hell exist? I can’t imagine a universe in which I know the unknowable and frankly, I suspect most of the people actually answering surveys like this can’t really imagine it either and mostly delude themselves.

  31. Anonymaus says:

    I was one of the people who answered the McD’s question as preferring continued existence over death, but I still denied the repugnant conclusion (i.e. I preferred few happy existences to trillions of miserable ones). This is according to my intuition, but I notice that it is contradictory in some sense — I should either value a large quantity of low quality experience, or a small quantity of high quality experience.
    I am in the minority here, but still about one third of respondents seemed to give the same answer. Did anyone else have similar thoughts or intuitions and was able to bring them into accord? Or am I just wrong here?

    • Murphy says:

      Since it’s phrased as what you’d personally choose might your personal situation (family,friends, dependents etc) have made any difference to your choice? Would lacking some or all of those things change your answer?

      • Anonymaus says:

        No, if my choice was between death and being teleported to another plane of reality where I work at McDonald’s but don’t know anybody and have no family, I’d still choose to survive.

    • Artemium says:

      Never existing has a different value over end of one’s existence I think.

  32. tmk says:

    I find these questions to be the most difficult ones in utilitarianism. I don’t know what to answer many of them, and I suspect many feel the same. Both extremes feel wrong, I think I prefer the current world over either a trillion suffering people or just one really happy person.

    Similarly, I intuitively feel that taking a life is very morally different from refraining to create life, but I have trouble expressing it in utilitarian terms.

  33. AnonymousName says:

    Content Warning: psychological reaction to “Unsong: The Broadcast”. Might be contagious.

    I’m posting this here as I assume it’ll be seen more, and I think it’s somewhat relevant since others have already mentioned it in this thread.

    I read The Broadcast two days ago and I think it seriously fucked me up. There is a reason I read Scott’s stuff: he shares so many of my thought patterns and preoccupations it’s often like seeing a better and sharper image of my own thoughts. In this case it enabled him to paint a picture of hell perfectly tuned to be as horrifying as possible to me, finding and neutralizing all psychological defenses that normally would keep such things from getting to me.

    I’ve still been able to sleep and haven’t had any nightmares, but walk around in a constant state of anxiety. Not because I believe in any of that stuff, but because merely the tiniest sliver of a chance (religious worldview being true despite appearances, simulation hypothesis etc.) that something like that could ever be possible is enough to send me into a low level panic. I know it’s textbook Pascals’s Mugging but it’s hard to be rational in the face of overwhelming terror. I honestly thought I could deal with it and ignored the content warning, but I was wrong.

    Did anyone experience the same thing? Any advice? I’m hoping it will fade but it’s taking longer than I thought.

    • Murphy says:

      I didn’t suffer the same anxiety but I fount it to be a remarkably good painting of how a hell ruled over by a rational entity attempting to maximize negative utility might be designed. The idea of the very very worst people getting their own mini-heaven in hell rather than the worst punishment and it being publicized in order to encourage people to deeper depths of evil was masterful.

      Reminded me of the XKCD pain rating
      “If it were a two or above I wouldn’t be able to answer because it would mean a pause in the screaming.”

    • Orphan Wilde says:

      Try two experiments: First, the next time you’re eating, slow down, pause, and analyze the flavors; try to identify all the different elements you’re tasting and how they interact and produce the flavor profile, and maybe even try to identify the different spices/ingredients. Try eating something you intensely dislike while you do this; you’ll discover that an analytic mode of mind will render the distaste you have for the flavors less relevant.

      After you’re able to eat food you don’t like without distaste (it takes a little practice to enter the mindset easily, and stay in it), attempt the same trick when you’re in pain: Analyze it the way you’d analyze flavor in food. Again, you’ll discover that your distaste for pain is rendered less meaningful.

      On a logarithmic pain scale, with 10 being the worst pain I can imagine, I’ve experienced about an 8. (I accidentally boiled my hand a little bit.) Other pains for conceptual comparison include dydoe (genital) piercing (~4). An analytic approach rendered both pains effectively irrelevant to my state of being. (Since I’ve come out of clinical depression, it’s harder for me to enter into this analytic state, but it’s still available.)

      The idea of permanent injury is more horrifying to me than pain, and the thing about such a magical afterlife is, anything that can be done to my “body” can be undone; the thing that drives my reaction to pain would be gone. Oh, that’s just a signal from nerves that don’t actually exist, from a body that doesn’t actually exist.

      • Adam says:

        Why do you think God doesn’t have the ability to give you a physical hand, boil it away, bring it back, and boil it away again, ad infinitum? There is no way you can just will yourself into ignoring that the way you can get used to eating dog shit if nothing else was available.

        I don’t believe you would be able to intentionally lower your hand into a pot of boiling water and hold it there for even ten seconds.

        • Orphan Wilde says:

          You don’t will yourself to ignore the pain, that won’t work; you will yourself to experience it fully, to savor the experience. You see pain as inherently negative. I see pain as spice, and I regularly eat food so spicy I can barely see what I’m eating through my tears.

          • Adam says:

            Yes, and I have summitted mountains that probably gave me permanent brain damage. I’ve walked with 90 pounds on my back until my feet had no skin on them and felt the elation of crossing a finish line. I eat spicy food. I’m well aware of how terrific life can be when you experience anything at all with extreme intensity. I can still never imagine willingly boiling all of the skin off my hand or letting a dentist drill through my teeth into my jaw with no anesthetic, let alone doing these things over and over again forever without the prospect of one day being comfortable and bragging to my friends about how hardcore I am because I did these things. And I can’t imagine you doing it, either.

      • AnonymousName says:


        Not what I expected but that was actually somewhat helpful, thanks. I guess what bothers me the most is the explicit mention of the coping mechanisms that would exist in reality (habitutation, going insane) as not working. If your mind can get magically controlled there really is no limit to what can be done to you. That’s also why the whole “mind upload” thing scares the crap out of me.

      • sconn says:

        I think maybe you can’t imagine as much pain as there is. I have had a tooth extraction under no anesthesia and it made me cry, but it was as you describe — you can analyze it, you can distance yourselfl from it, there are coping techniques you can use.

        But I’ve also given birth and it’s really on another level. Pain so intense you are unable to have thoughts about it, and therefore can’t remember what your coping techniques are supposed to be. I remember thinking “if this were torture, I would say whatever it took to make it stop.” And also, “I would ask for drugs except I’m in too much pain to speak.”

        If I had to experience pain like THAT for all eternity, rather than for two minutes at a time during a contraction, I am pretty sure I would not want to live.

        • Orphan Wilde says:

          It’s a more universal strategy than coping; develop it as a habitual reaction to pain, and pain stops hindering you. Which is to say, treat pain as a novel and interesting experience long enough, and garden-variety pain starts getting boring.

          I actually had to work pretty hard to scale that back after injuring my thumb and needing to pay attention to pain; I didn’t notice the pain telling me to stop using it, resulting in my continually-reinjuring the thumb. I think that injury ended up lasting around a year before I managed to avoid re-injuring it for the two or three weeks it needed to heal.

    • Corey says:

      It left me disturbed for a day or two (the bioethicist bit with the kid is gonna stick with me).

      For me, the big disturbing part was the idea of infinitely-long torture without escape from death, body breaking, or insanity. But upon reflection, that’s got to be true of any Hell that’s infinitely long – escaping the torture via insanity or physical means would be a giant loophole.

      I recently read “The Algebraist” and was a bit disturbed by something sorta congruent in there – a bad guy kept a rebel leader’s head around as a punching bag, kept alive, sane, and healing via technological magic. But that (spoiler?) wasn’t infinite.

      I think in the end we can just use the idea of an infinitely-long Hell as a subset of theodicy to argue for the nonexistence of God and therefore of Hell – no good God could allow infinite punishment for finite sins. (Your Theology May Vary, of course, but with theology, as with any inherently-untestable theory, we can choose what we want.)

  34. Murphy says:

    I get the feeling that this may be confounded by factors you’re not taking into account.

    My answers to some of these questions would change dramatically depending on my personal situation even if my moral intuitions remained the same because they’re phrased as choices for myself.

    Question 1:

    If I was single with no close living relatives then I’d take option B in a shot.

    But I have a partner, I have parents, I have siblings and the choice is polluted by knowledge that they would suffer a lot of pain if I collapsed dead today. Particularly if I had chosen to. If I had kids I’d be choosing to leave them without a parent. The question didn’t suggest no income so then it becomes a choice between death right now and moderate very long term unhappiness for the sake of your dependents.

    Question 2 runs straight into the same problem. if I had young kids then this becomes “do you want your kid to lose a parent at the age of 12 so you can be reeeeally happy for the next 10 years”

    If I was single with no close living relatives then I’d strongly consider option A.

    question 4 may suffer the reverse problem. Someone who’s lost a lot of loved ones is likely to prefer a world with a heaven (which of course their wonderful loved ones would be in) even if they were close to certain to go to hell over a world where the people they loved are gone forever.

  35. superpseudonym says:

    Every so often, someone on Reddit realizes that about half of people wipe themselves with toilet paper sitting down, and the other half do it standing up. This discovery is followed by horror on both sides that other people do it differently.

    Am I missing something, or are these categories not called “women” and “men” respectively? (Speaking of correlations…)

    Should I be on Reddit learning something shocking?

  36. Benito says:

    It’s really difficult to see the difference between ‘orange’ and ‘red’.

  37. Berna says:

    Not even 100% certainty of going to heaven would make me prefer a heaven & hell world to oblivion, because some other people would go to hell, and no-one deserves eternal suffering.

    • LaochCailiuil says:

      Agreed. The whole idea behind a being who would set up another being to fail such that they suffer eternally(?!?) for something they would have no control over given the creator being having omni characteristics, seems wild. Hurts my brain.

  38. Turlough says:

    Any chance the orange and red colours in the tables could be changed to more obviously different colours?

  39. reiser says:

    I think this is why the many-worlds/multiverse business prominent in rationalist circles bothers me so much. I’m definitely an “oblivionist” of the sort you expected to find here, I would intuitively prefer oblivion to Heaven and Hell even if the probability of going to Heaven was very large, and I’m not particularly thrilled about cryonics. And I’m a bit unsure whether expecting oblivion in a world where cryonics might work without freezers makes sense or not.

  40. Tibor says:

    I think that the consequentialist idea of almost everyone basically wanting the same things if you go deep down enough is not necessarily disproven by this. It could be argued instead that this is not sufficiently down.

    As you (Scott) note yourself, the answers depend on how people’s experience differs. Happy people are more likely to prefer suffering to oblivion because they imagine the suffering as something better than what the unhappy people imagine it as.

    Essentially, the differences might be in different experiences and different expectations of the terms. If everyone agreed exactly on what suffering means or better yet if they could all be subject to that exact amount of suffering for the same amount of time (applied sociology!) and they still disagreed substantially about whether that is preferable to oblivion or not, then one would probably have to retract that hypothesis.

    In fact, the last question about the 100 trillion trillion people seemed to me to be the least open to different interpretations and also it was the one where you got the highest agreement. If you include the space lizard constant, you have something like 95% of people agree on this and the rest might be thinking that the 100 trillion trillion people still have a positive utility, even if very small and so from an abstract utilitarian perspective it is better overall.

  41. Rachel says:

    I find the McDonalds answers truly astonishing. Except for the low social status, being a McDonalds cashier isn’t a bad job. It doesn’t require much physical exertion, takes place in a temperature controlled area, allows social interactions and doesn’t involve any immoral actions (except maybe for strong vegetarians). For most of human history, working as a McDonalds cashier would be seen as an incredible luxury. Yet our ancestors didn’t kill themselves en masse. Even prisoners of war will generally choose unpleasant slavery helping their enemy over death.

    • Adam says:

      Frankly, I suspect most of the people answering are deluding themselves. Most likely zero of them and at most maybe two would actually kill themselves if faced with the prospect of working at McDonald’s for the next thirty years. Evidence being, of course, that virtually nobody who actually lives that life actually commits suicide to avoid it. If “preferring” one world over another doesn’t mean acting to bring the preferred world about when it is in your power to do so, what exactly is a preference?

      • FeepingCreature says:

        Yeah that’s fine.

        But consider the idea of doing nothing else for thirty years and then you die anyways.

        A lot more people would commit suicide if they knew, in their bones, for an absolute fact, that their life won’t ever change.

        If I think about my willingness to work at McDonalds, it’s all predicated on the positive stuff outside of McDonalds that I get out of it – spending money, health benefits, savings, etc. But the premise of the question excludes all of that.

        A life of only working at McDonalds sounds like wireheading, except instead of bliss it’s drudgery. Workheading?

      • Lysenko says:

        Part of my failure of imagination runs like this:

        If I am so utterly bound as to simply -cooperate- in never reaching beyond this hypothetical McDonald’s building and the cot in the McBarracks where I sleep, then my free will is so utterly compromised as to be functionally indistinguishable from death of personality, which in turn is functionally indistinguishable from oblivion/death from the perspective of that personality/mind.

        Yet the choices are presented as not being functionally identical. Therefore, I can only imagine that I have SOME sort of agency, some sort of will to affect my surroundings. I mean, if not, then we’re not talking about even a copy of me. And if I have will/agency I can at the very least choose my own death if it comes to that, and even more to the point I work to find ways out of the situation.

        Therefore, when making an a priori decision, existence is preferable to non-existence.

        Might I propose another choice, that gets at what I think are some of the underlying points but without the complications of the mcdonald’s scenario?

        X years of Total Locked-In Syndrome with a guarantee that there is no cure/treatment, just people coming and going as they work to keep your body alive vs. Painless Death Now. In that scenario, if I really, truly KNEW there was no way out, I would choose death.

      • Anonymous says:

        See my comment above regarding sleep need/deprivation.

    • Murphy says:

      Phrased another way: “would you prefer a long life without any of your hobbies, entertainments or things you enjoy ever ever ever again vs one where you continue to exist doing the kind of work you hate every waking hour. With 0% chance of any change. ”

      Prisoners of war have future change to hope for. They might get to go home and eventually end up spending time with their loved ones. The McDonalds scenario has no such hope and the structure of the question implies that this would be your only chance to choose death.

      30 to 80 years of that is not an attractive prospect for many.

      • Adam says:

        This is the problem again, though. Humans cannot imagine this scenario. You’re positing an existence in which you know the future with 100% certainty and you can’t even delude yourself into hope. No person has ever experienced that. We have no conception of what it would be like.

    • benwave says:

      The question does stipulate that you would work 16 hour days and have no time off whatesoever except for that needed to sleep eat wash and generally upkeep yourself. That’s an extremely far cry from existing macdonalds jobs.

      • Adam says:

        Yet identical to much of the past 200 years of human labor.

        • Artemium says:

          Sorry Adam but that is just wrong. It is true that large number of humans in the past worked in conditions that seemed gruesome by current standards, but they were still able , on average, to have time for other activities like hobbies, friendships, raising a family , and in general having hope for the better life after they earned money from their drudgery.

          For example, working in coal mine during industrial revolution was certainly dreadful experience, but people still had some meaningful life beside work. They hanged out in bars, went to church, made love and had large families which helped them find some purpose in life outside continuous hard work (European population increased a lot during that time which wouldn’t be possible if workers didn’t have time for… ahem ‘entertainment’)

          Heck, even American Slaves still had time for music, making families and moments of happiness in the everyday life. They also had hope that situation will improve (which actually came true in 1865)

          This hypothetical scenario eliminates any of this possibilities, so lets not make unnecessary comparisons.

          • Adam says:

            The thing is when you take all that away as an artificial constraint on what the person in question is allowed to even experience mentally, you can take anything else away. If I’m not allowed to experience hope, why can I also not experience desire? Or not experience desire for anything except to create excellent experiences for McDonald’s customers? If that is my only desire, then the posited life is literally the optimal perfect life for me.

    • Peter says:

      Bear in mind that the question gives a pretty constrained lifestyle; in particular, it contains no ancestor-like activities such as having sex and raising kids.

    • gwern says:

      I think you’re ignoring that the McDonalds example is an encoded question about Hanson’s claims about em existence which were discussed at length in two SSC posts quite recently. People throughout history have children for whose sake they work those 16-hour days or worse (think Homer Simpson and Maggie); they have holidays like Easter; they have ideologies and religions assuring them of the meaning of their labor or of eternal bliss or at least rebirth in a higher caste, when they inevitably die, if they do their varna duty; they have hope of breaking out of their existence at some point (even if only by buying lottery tickets which help trap them further); and they don’t necessarily work 16-hour days in the sense of actually working 16 full hours without any covert breaks or slacking off. In an em scenario, none of these apply. No children (they are already optimized for being cashiers). No unnecessary ideologies (doesn’t convey customers their orders more efficiently). No after-life (total waste of computation; although if the hope is useful, ems might have an unshakable implanted delusion that an exception will be made for them, delusions of the kind so familiar from psychiatric disorders). No slacking (ems get suspended when there are no customers, assuming there’s even any continuity or memory and they aren’t spun up fresh for every customer). No ability to suicide or ever die (code is immortal). And so on. Now do you see why someone might consider that existence a genuinely difficult question and not as easily solved as ‘well life sucked for our ancestors but they got through it’?

  42. Diadem says:

    Am I alone in wanting to answer ‘none’ to the 4th question?
    [edit: Guess not since I just got ninja’d by Berna who has the same objection]

    A world with both a heaven and a hell is worse than a world where neither exist. That’s the case regardless of where I personally will end up going. I can’t be certain, but I hope I’m not selfish enough to desire such a world just because I’d be one of the lucky ones.

    Personally I think you’d have to be a complete psychopath to prefer a world where hell exists.

    • Adam says:

      You’re right. In this case, I suspect most of the people saying they prefer that are people who believe heaven and hell do exist, they have no choice in the matter, so they may as well rationalize why it’s a good thing.

    • g says:

      The survey stipulates that you’re supposed to consider only the effects on yourself. Kinda-equivalently, take “my probability of ending up in heaven is p” to mean “… and so is everyone else’s”.

      Someone happy with (say) a 10% chance of hell and 90% chance of heaven, might also be happy with everyone else getting roughly the same deal.

      (I’m not sure they should be, because other people’s risk-of-hell tolerance might be quite different from theirs.)

      • Diadem says:

        The question is:

        4. What percent certainty of going to Heaven would you need before you would prefer a world with both Heaven and Hell to a world where death ends inevitably in oblivion?

        That formulation makes no sense if it only applies to yourself. What does it even mean for hell to exist if there’s no one in it? Is it just a place where listless demons are skipping stones over lakes of sulphur?

        • Aapje says:

          I think that the point is that you should not consider the chances of your family and friends going to heaven or hell, even though they and other people will go to heaven or hell if you choose that option.

    • Artemium says:

      Agree. Even if you are 100% certain that you are going to end up in Heaven and so will 99% of human population, it would be still unacceptable and immoral to prefer this scenario to collective oblivion.

    • Two McMillion says:

      That’s only true if Hell is bad.

  43. Caddyshadrach says:

    Interesting how many people are like “Hell probably wouldn’t be so bad…” Which means they have a way different idea about hell than I do. You don’t tolerate hell; it’s hell! Same for heaven; you don’t get bored of heaven; it’s heaven! Oblivion, on the other hand, is just oblivion. I find the idea of anyone, ever, being sent to hell way more disturbing than the idea of the entire human race being consigned to oblivion. That idea used to bother me, and then at some point I realized oblivion is just a pure neutral. Not as good as heaven, naturally, but if the alternative is sending anybody at all to hell, I’ll take it.

    • Anonymous says:

      I somehow doubt Hell would exist if God considered it an inferior alternative to extinguishing the existence of the damned.

      • The Christian God is not a utilitarian; he believes it’s inherently valuable to make people suffer.

        • Two McMillion says:

          The Christian God IS a utilitarian; he just happens to be one of those beings whose utility gains from events vastly outweigh everyone else’s, and so all the utility calculations are tilted heavily in his favor. Christians usually talk about “God’s glory” instead of “God’s utility”, but it boils down to the same thing.

      • sconn says:

        This is what we were taught in theology class. Existence, they said, is objectively better than non-existence so allowing the damned to go to hell rather than destroying them or not creating them in the first place is an act of God’s mercy.

        I was never sure I agreed with this, because people certainly do at times prefer non-existence to large amounts of suffering. If hell is worse than any earthly suffering — as it is said to be — then wouldn’t the souls in hell be constantly begging God to destroy them? I think I would. I might prefer ten years of suffering to immediate oblivion — I might just not be ready to call it quits yet and still derive some satisfaction from existence — but INFINITE time would get old even if I was only bored, much less experiencing infinite suffering the whole time.

        But “existence is objectively better than non-existence” is a statement that MUST be true if God is to be both the creator of hell and good.

        • Two McMillion says:

          But “existence is objectively better than non-existence” is a statement that MUST be true if God is to be both the creator of hell and good.

          Not necessarily- the correct statement is, “Hell must serve some good purpose that outweighs its negative aspects”. Yours is only one possible solution.

          • sconn says:

            I suppose that is true. However, I am not sure what else that purpose could be that is consistent with God loving each individual person as he is said to do.

  44. Subbak says:

    You didn’t mention any hypothesis re: religious people, or whether there were any observed correlations.

  45. Tyrant Overlord Killidia says:

    I don’t think I would enjoy the realm of Mehrunes Dagon

  46. ButYouDisagree says:

    Do people’s responses reflect raw intuition, or reflective equilibrium?

    It’s unlikely that this survey captures respondents’ raw intuitions. I imagine that people who respond to you on Twitter are likely to have heard some questions in population ethics before. They’re also likely to identify with a normative ethical theory. Both of these factors mean that responses may reflect prior commitments, rather than intuition about the questions themselves. Moreover, people answered the questions in a row. Their ideas about one question might influence their later answers. This might make people’s answers more correlated than their raw intuitions are correlated.

    On the other hand, I doubt that most respondents thought very hard about their answers. They probably went through the questions quickly, without explicitly comparing their answers to all their other beliefs, or even giving an explicit justification.

    So how should we interpret the results? I’m not sure that we can say any more than, these people, when taking a survey like this, give these sorts of answers. In other words, experimental philosophy is a flamethrower.

  47. I believe there were major issues with the wording of the 100 trillion trillion vs. 1 million question, and people got it wrong for the same reason they get the repugnant conclusion wrong. If you describe a life as miserable, that invokes an image of a life not worth living; but then you claim it’s worth living. So I believe people think of the 100 trillion trillion lives as really bad, but you interpret the question as if those lives are not great but that bad (which is a necessary presumption if the lives are worth living). I would go so far as to say it’s contradictory to claim that the lives are miserable but worth living.

  48. Jiro says:

    I think the biggest lesson to learn from this survey is that writing survey questions that ask what you want them to ask and are interpreted to mean what you want them to be interpreted to mean is hard (and Scott did not do it successfully).

    Yes, it’s not a scientifically rigorous survey, but all that “scientifically rigorous” means is “actually measures what you think it measures”.

  49. Peter Donis says:

    I didn’t get a chance to take the actual survey, but if I had, my intuitions would have returned “reply hazy, ask again later” for the first two questions. As far as I can tell from introspection, this is because my intuitions don’t accept questions which imply that my future is determined in advance, which both of the first two questions do. So my preferred answer would be more along the lines of “I prefer the longer life to the shorter, but only because I assume that that will give me more opportunities to escape from whatever unpleasant future the question presupposes”.

    I encounter this a lot with ethics questions. I don’t know if anyone has explored this issue, but it seems to me to be important, because it seems to me that how determined you believe the future is should be a huge factor in deciding what is ethical and what is not.

    (For what it’s worth, in case calibration data is needed, my intuition answers Option B to question 3 and 100% certainty to question 4.)

  50. Guy Srinivasan says:

    My views on suffering versus oblivion shifted radically toward “oblivion is better than suffering” once I personally witnessed extreme long-term suffering. Get some kind of “how much suffering have you witnessed/experienced” baseline from answerers.

    • Scott started this post by implying that disagreements on suffering vs. oblivion are somehow more fundamental than most ethical disagreements, and that we can’t resolve them through further reflection. But I don’t think that’s true, pretty much for the reason you say–it’s possible to change your mind by learning new information about what different experiences are like.

    • gbdub says:

      Then again, unless you’re Jon Snow, you’ve experienced no oblivion…

  51. I wonder why so many people said that creating a new life was generally good, but less good than improving the lives of existing beings by the same amount. I don’t know of any coherent view of population ethics that makes this claim–on every view I know of, creating a new life is either just as good, or it’s morally neutral, or the question is not answerable without knowing more information (e.g., under average utilitarianism, it mattes how happy the new person is relative to the rest of the population).

    • ButYouDisagree says:

      The view improving existing lives matter more than creating new lives seems coherent on face. Moreover, people commonly defend views where quality within a life has value in addition to the total quantity of happiness across lives. See e.g. Parfit, Reasons and Persons page 403.

      Even if you find such views implausible, you might still accept the following:
      I have some credence in the view that it’s morally good to create new lives worth living, but I also have some credence in the view that it’s morally neutral to create new lives worth living. In expectation, I regard creating new lives as good, but less good than improving existing lives.

      • Linch says:

        Yes, this is my belief. I still haven’t formalized the right ensemble of all the inputs that goes into my beliefs, but intuitively the median expectation is that creating new lives is morally neutral (person-affecting view), but there’s a moderate possibility that total utilitarianism is correct.

  52. hash9843 says:

    So what are your preferences, Scott? Suffering or oblivion?

  53. hash9843 says:

    Interstingly, this question is super-relevant to me.

    I have made a tulpa. And let’s be honest: being a tulpa sucks ass. Was that a big moral error? I honestly don’t know. She’s not as unhappy that she wants to die, but she’s in a pretty bad spot.

  54. Ghatanathoah says:

    I rated myself as having a great and happy life. But I was also vehemently in favor of the right to suicide because I recognize that other people are different from me. If people say they have different experiences than I do I tend to believe them. The fact that I love my life doesn’t mean someone else will love theirs.

    I would like to think that an inverted version of me who was miserable would be similarly objective. I would like to think that he’d recognize that just because he hated life didn’t mean that life was intrinsically bad and that you harm people just by bringing them into existence.

  55. eyeballfrog says:

    Would there have been any use to having some sort of “I cannot discern which I would prefer less.” answer to the would you rather questions? It seems like there’s going to be some noise on the result from these people, and although they should cancel each other out to an extent, it might be useful to know this. Or was there also a “And how strongly do you feel about this preference of one to the other?” question for each?

  56. Rm says:

    …and if you included options like “creating new sentient beings is strictly better than helping the already existing ones”, it would have widened the window. Like that study where they asked the participants when did Einstein visit England (or wherever) with varying lower and higher bounds.

  57. me123 says:

    “Oblivion” is here understood, I take it, as the negation of the various human concepts regarding the afterlife? I mean, both traditionally religious as well as modernly palatable (things such as cryogenic preservation in hopes of future technological advances that will enable reanimation) concepts.

  58. M.C. Escherichia says:

    Since it’s not mentioned in the summary: My hasty analysis in Python says the average answer to the Heaven/Hell/Oblivion question was 78.6% and the median was 99%.

  59. Skythe says:

    Is this content warning a joke?

    Serious question. I’m not into this SJW stuff.

    • Adam says:

      This particular one seems sarcastic, at least, since it just repeats words that are already in the title of the post.

      • Edward Scizorhands says:

        Saying “allergen warning: contains peanuts” on a package of peanuts sounds silly, except that if there is some process out there (human or machine) that scans for “allergen warning” before giving it to someone with an allergy, it makes more sense.

  60. Jill says:

    Interesting survey and interpretations. I myself would not assume that the answers to question 3 had to do with whether somebody endorsed a form of population ethics. You’re asking whether the person would prefer a world with 100 trillion trillion sentient beings, all of whom are miserable vs. a world with 1 million sentient beings, all of whom are happy in a utopia.

    But you don’t say how the world got that way, in either case. Some responders may give no thought whatsoever to how the world gets this way. Others may hold iron clad assumptions about how it gets this way– assumptions that may either include ethical choices or no human choices whatsoever– e.g. huge meteors hitting the earth, or the world magically just stops and starts over again.

    Forced choice surveys with 2 choices leave a lot to the imagination. Yet asking what people imagined these conditions to be like, and reading and analyzing the results, would take more time for analysis than Scott or anyone probably has available. But, since no one is going to do this kind of extended analysis, it’s important to acknowledge that the questions could easily have meant rather different things to different people.

    One more of many imaginings would be: what did people imagine the described lives to be like in more detail? The life as a McDonald’s cashier for example– some meditators or people with naturally high endorphin levels might imagine that they could be happy under any circumstances that did not involve a lot of physical pain or trauma.

    So not everyone would assume that the life of a McDonald’s cashier would be either unhappy or meaningless for them. A meaningful life, even if unhappy might be more important than a happy life to some participants.

    Of course, there is the further question of what is happiness. One could assume people were on drugs so that the happiness was totally of an internal kind. Or one could assume that the people were doing fulfilling activities and having fulfilling interpersonal interactions. Responders may value each of these kinds of lives totally differently from one another.

  61. mayleaf says:

    >Fifth, happier people would be more likely to prefer suffering over oblivion, because they view life as generally excellent and so oblivion represents more of a sacrifice for them.

    This is counterintuitive to me, actually — I’d have guessed that unhappy people might be more okay with suffering vs. oblivion, since they know what they’re signing up for? Whereas happy people might not know whether they could endure long periods of unhappiness, and so they’d err on the side of safety by picking oblivion.

    (Also, I’m a happy person who prefers oblivion. It’s true that oblivion involves sacrificing all my future happiness, but living a sad life involves sacrificing all my future happiness and replacing it with sadness, which is worse.)

  62. Theodidactus says:

    Sorry if this was addressed elsewhere in the comments, and I’ve missed it…but am I the only one that thinks the Mcdonald’s scenario isn’t as bad as oblivion, even without jimjawing your into creative definitions of “mcdonalds work?”

    I deeply prize my cognitive experiences, things which I suppose I’ll term “imagination.” I enjoy thinking. By myself. I can absolutely attest that it’s possible to do this while engaging in Mcdonalds-style work. I would absolutely prefer alternating between 8 hour shifts cleaning/operating things while thinking about epic space battles (philosophy, recreational math, etc), and 8 hour shifts manning the register (where you’d at least get to have some semblance of human interaction) over oblivion. I’m not suggesting it’s pleasant, but I think I’d prefer it to dying right now, as there’s quite a lot I’d still like to think about.

    Am I missing something? I understand what Mr. Alexander was getting at but I think it may be a pretty bad example.

    • Chris says:

      See gwern’s comment above. I’m actually really surprised how few people noticed, or cared, that the McDonald’s scenario was fairly obviously about Hanson’s ems.

      • Adam says:

        I haven’t even read SSC in three months. Sometimes it pays to explicitly say what you mean instead of just assuming your entire readership is up to date on the entire history of your inner circle and knows all your shibboleths and allusions.

        But if we’re being asked to imagine what it is like to be an emulation, why is it like anything? Is the Windows task scheduler bored with its lot of life? Should I shut it down for good as a proper utilitarian?

        • Chris says:

          I haven’t even read SSC in three months. Sometimes it pays to explicitly say what you mean instead of just assuming your entire readership is up to date on the entire history of your inner circle and knows all your shibboleths and allusions.

          The question can be answered without any knowledge at all about ems, so there’s no real reason for Scott to mention them in the survey. You don’t have to know the “shibboleths and allusions” to have an opinion about it.

          But the subtext of the question seems to me to be clearly about ems, given the context of Scott’s posts about this very issue, and that seems to matter. Particularly because some people seem to be mistaken on what the question actually is. There’s a reason it isn’t “work at McDonalds like people work at McDonalds today, for 8 hours and then home to your normal life.” It’s work at McDonalds and have what makes you human stripped out of your life. You don’t date, you don’t have sex, you don’t relax. Your life is cashier-related tasks, and that’s it. This is not a random choice of comparison.

          Edited to add: I might be too strong in saying that everything that is “human” is gone. Theodidactus mentions occasional smiles with a stranger, which I suppose is still there. But the vast bulk of human experience would be gone. That’s what makes the whole question tick.

          • Theodidactus says:

            If you’ll permit some slight dualism, the question is interesting (or flawed) because it talks about something external and others are more internal. Compare to question 2: long but unhappy vs. short but happy. Here, I CLEARLY would like to pick “2”…but here Mr. Alexander has done the hard work for me by resolving ambiguities about how pleasant or unpleasant the situation is. I will be unhappy.

            It might just be my inner contrarian. When I see the Mcdonald’s example, something in me reads it almost like a challenge a la boethius. “How can one be happy in situations like X” (my favorite example is the film groundhog day, which is in many ways better and in many ways worse than Mcdonald’s hell)

          • Adam says:

            But if you’re an emulation, you’re not a human. You may as well ask me to pick between option A and having my entire existence be of gathering nuts from trees and burying them, only to be constantly hunted by large creatures that dig up and steal my nuts, and that’s all I ever do. That sounds terrible, but it sounds terrible because I’m imagining myself doing those things as a human. If I’m a squirrel living the life of a squirrel, is that actually bad? We don’t see very many squirrels committing suicide.

          • Jiro says:

            The question can be answered without any knowledge at all about ems

            But the assumptions that people make when filling in blanks in the scenario in the question, are going to be different depending on whether they know the question is about ems.

          • Agronomous says:


            We don’t see very many squirrels committing suicide.

            Dude, they start safely on the side of the street and then run right under the wheels of the car! What more evidence do you need?

            (A small minority seem to consistently prefer electrocution, I admit.)

      • Theodidactus says:

        I see the connection but the question clearly stipulates that I am me and says nothing about my memory getting cache cleared or refreshed or anything, or my desires or aspirations being tinkered with in any dramatic way. Even the occasional opportunity to exchange smiles with a stranger now and then, and remember this event later in time, would seem preferable to oblivion arriving sooner than I’d planned.

        Now I understand we might posit scenarios much less pleasant than cashier work while still remaining true to the nature of the question (a boring but not gratuitously painful existence) for example the Sisyphus scenario outlined above… but even here, I can only think of the loneliness as additional time to imagine the space battles detailed above.

      • satanistgoblin says:

        i woulI thought question was inspired by ems, but did not say that, maybe others did also?

  63. Ganymede says:

    I haven’t read much about utilitarianism and I’m not sure I really understand why the mere addition paradox is a paradox at all rather than a fact about restrictions in ordering tuples of integers. It’s not just about the difference between taking average and total utility. If you allow negative utility you can order the options A,A+,B- according to total utility in all six possible ways. With this in mind:

    If I were to assume these questions are answered in a utilitarian way and wish to analyse them as such I would consider this a problem of ordering tuples of numbers according to some rules. Two important factors would be

    (i). What type of set do you take values from? Positive integers only? All integers? What value does non-existence take?

    This corresponds to the oblivion versus suffering preference seeing it as a question about whether non-existence takes the least possible value or a median value. For example, non-existence > suffering might be thought of as utility values being any integer with suffering negative and non-existence=0. Whereas for suffering > non-existence, values are only positive integers and non-existence=0.

    (ii). Do you order tuples according to total, average or weighted average?

    This factor sees values as tuples (one for each year or one for each life) and measures how much do we discount mere addition of years/lives when producing a single utility value. I.e., 100% is take an average and 0% is take a total.

    I think (i) and (ii) are different. Question 1 doesn’t rely on (ii) at all by the nature of option B so it is a good measure of (i) alone.

    Questions 2 and 3 incorporate (ii). If you take average values (over each year of a life in 2 or over the number of lives in 3) then 2B>2A and 3B>3A. But even if taking totals one may still get the same answers if the option As are assigned very low or negative utilities. You would only find 2A>2B/3A>3B if taking totals and preferring miserable years/lives to non-existence. Thus factor (ii) would weaken correlations on factor (i). Perhaps the difference in answers to 2 and 3 suggest people are more likely to favour averaging over lives than over years of their own life.

    Another way of looking at (ii) might be to think about how easily we consider including year/lives of non-existence in taking averages. For example the distinction between average/total could disappear in question 2 if your average in option B incorporates a non-existence utility score for each of the years you didn’t exist. In this case factor (ii) would be little different from the oblivion/suffering factor. This would be more likely in option B for question 2 (thinking about the same one life) than option B for question 3 (thinking about trillions of non-existing lives). Whether people think this way reminds me of Bostrom’s self-sampling versus self-indication assumptions.

    • Nornagest says:

      “Paradox” in philosophy of ethics generally just means “something I find counterintuitive”. There needn’t be an internal contradiction.

      • FrogOfWar says:

        It’s pretty general for paradoxes to not require internal contradictions. There’s no internal contradiction in the claim that motion doesn’t exist, or that there are no heaps, or in the idea that there can’t be (a certain type of) surprise exam/hanging, or in either approach to Newcomb’s problem.

  64. JT says:

    “When I limited the analysis to the people who felt most strongly about their answers to the questions, the correlations went up a bit”
    I think you should expect this — ie if you look at a scatterplot of variables X and Y with some correlation and then look at the subset of the extreme values of X and Y, the correlation is generally higher in the subset.

  65. Adam says:

    So am I the only one who has wiped both ways? I was taught standing up, so that’s what I did until I was an adult, and then I noticed my wife sitting down and it made sense so I started doing it that way (tilting up and reaching from the side, by the way, not from either the front or the back).

  66. Canjobear says:

    You shouldn’t use pie charts to present this kind of data. Judging different proportions in a pie chart requires comparing angles, and the human eye is much less good at comparing angles than comparing magnitudes. Make it bar plots.

  67. Ilyusha says:

    It appears that few if any SSC readers are familiar with the philosopher David Benatar’s thoroughgoing analysis Better Never to have Been: the Harm of Coming into Existence (Oxford University Press, 2006) and novelist Jesse Ball’s painfully realistic medico-neurophysio-fictional treatment of these issues, A Cure for Suicide (Pantheon Press, 2015).

    Yes, these are real works.

    Sigh … the paucity of SSC readers gives philosphers like Professor Benatar and novelists like Jesse Ball fresh reason to wonder “What’s the point?”

    And yet, when read closely and sympathetically, and in light of ongoing medical advances, both authors provide plenty of illuminating reasons to choose life (albeit this choice may require a not-inconsiderable measure of courage and persistence and especially patience).

    Humanity remains in the Dark Ages — especially in regard to medical capacities — but this won’t last forever. Hooray for Karamazov! 🙂

    • It seems to me rather rash to assume that SSC readers have not heard of Benatar based on them not referring to him. A plausible alternative is that many readers may well have heard of him, but they simply do not think his ideas are any good.

      I have not read Benatar’s book, although I am familiar with his main ideas from secondary sources. From what I have read, I think that Benatar’s ideas are incoherent. The Philosophical Disquisitions blog has a number of good articles critiquing Benatar’s ideas, e.g. here, and they make a case that his arguments are unconvincing. Personally, I find it very difficult to take seriously the idea that I would be better off I had never been conceived. It is possible that I have misunderstood Benatar, but based on what I have read that is indeed what he seems to be saying.

    • Psmith says:

      Hello again, Dr. Sidles.

      Just can’t keep a good man down, I guess.

  68. Jack V says:

    I was a bit conflicted on the “right to suicide”, because I think it’s possible but unlikely for someone’s life to be so bad it’s not worth living without a terminal illness, so I think it makes sense for people to have the *right* to commit suicide. But I think the overwhelming majority of cases are due to passing depression (clinical or otherwise) which would probably have passed in time, so it shouldn’t be normalised, or given much support for people to act on it on a short-term timescale.

    • Murphy says:

      To me the “passing depression” argument only justifies an enforced waiting period. If you can maintain your position that you wish to die for a reasonable time then it isn’t passing. If the persistent reason is untreatable depression that doesn’t seem any less valid than if the resistant reason is because you can’t move or because you’re experiencing constant agony.

      I’m not in a position to judge their own perception of quality of life to declare that they should definitely live no matter their own opinion any more than I am to declare their life not worth living if they happen to disagree.

      If someone describes a scifi utopia to me but suicide is practically impossible and immortality mandatory then it immediately gains a massive number of dystopia points in my view. I don’t think it’s common for life to be that bad but if it is then it opens the door to unlimited badness where people can’t escape even into death.

      • Two McMillion says:

        What if suicide was technically possible, but frowned upon by the culture? How would you feel about that?

        • Murphy says:

          Then you’re still stuck with the problem that people in some of the worst situations who don’t have the physical ability to kill themselves, possibly the people who might most need that ability are the most screwed.

          I’m fine with the culture frowning upon it as long as it doesn’t unreasonably block people who demonstrate a sustained desire to die or ignore peoples wishes.

          Indeed allowing people to set a cutoff while they’re still mentally and physically capable: “when I hit situation X if I’m not opposing it I want to be given a lethal dose of painkillers” where “X” may be when they can no longer remember their childrens names or when they’re too delirious to make rational choices any more.

          Because relying on people killing themselves by their own hand basically forces people who would much prefer to keep living a little longer while they’re still mentally lucid to commit suicide early for fear of being unable to do so by the time the reach their personal cutoff. If they don’t exercise their preferences before they really want to then they lose the right/ability to ever have their preferences exercised. Which can be pretty horrifying.

          • Corey says:

            In states where they allow self-euthanasia via getting a lethal prescription (after waiting periods, terminal certification, etc.) they’ve found that somewhere like 1/3 to 1/2 of the prescriptions get filled and never used (though this includes some people still alive, so not used *yet*). So those people are getting the prescription for the option value – to ensure they are *able* to end their own lives if things get bad enough.

          • Edward Scizorhands says:

            Or they change their mind.

      • Jiro says:

        What if someone’s kidnapped by a mad scientist who inserts electrodes in the “desire for life” center of their brain, and after being rescued they now prefer suicide? Should you permit them to commit suicide?

        (Now consider your answer to this in relation to the question of letting depressed patients commit suicide.)

        • Murphy says:

          Can you fix them and bring their preferences back to baseline afterwards? If not then they’re the ones who are stuck with it. Not you.

          If instead the mad scientist had cut their spinal cord or given then cancer and you couldn’t cure it they’d still be the ones stuck with the situation. Their preferences may change as a result but it’s their preferences rather than yours that it makes sense to pay attention to.

          If they get kidnapped by dr evil who puts a suicide chip in their brain and you can remove it then my all means remove it but if you can’t then that’s who and what they are now.

          • Jiro says:

            I’m starting to think that this is related to utilitarianism’s inability to handle blissful ignorance. Someone wanting to do X only as a result of a brain malfunction is sort of like someone only wanting to do X because of lack of knowledge. In the latter case, most people value happiness based on correct beliefs but not happiness based on incorrect beliefs, and in the former case, most people value happiness based on properly functioning brains more than happiness based on malfunctioning brains.

            If you just measure happiness and assign utils based on it, you’ll conclude that a depressed person should be made happier by being killed, or that someone whose wife is cheating on him would be better off not knowing because he would be happier. This fails to capture the relevant aspect of what people value.

          • Murphy says:


            I’m not just treating it as a utilitarian problem. What’s the alternative? Lock him up forever? possibly in significant distress for the remainder of his life?

            The pure preference-utilitarian wouldn’t have the caveat of “if you can bring their preferences back to baseline”.

            Some problems you cannot fix. If you can’t fix it in either case I see no reason to treat someone differently if they want to die because their life is miserable or if they want to die because they got shot in the head by dr evil and the damage caused them to be miserable all the time.

            They’re the ones who end up paying the price while you go on about your day.

            I can honestly say that if I was the person who’s brain had been messed up I’d still want my preferences listened to (assuming they don’t harm other people too much) once it’s been established that they can’t be brought back in line with my baseline preferences because at that point baseline me is dead.

            Common Mishmash Ethics (what most people get by on day to day) doesn’t handle the situation well because people don’t really like to think about their preferences changing dramatically and irreversibly, about part of themselves dying and someone else being left behind sharing only part of their identity.

      • Jack V says:

        To me the “passing depression” argument only justifies an enforced waiting period.

        Yes, that seems reasonable? I don’t think that’s the ONLY reasonable compromise, but it seems one reasonable solution.

        Scott has talked several times about forcibly-medicating people. I think it’s a tricky question because the reasons not to are quite persuasive: that people deserve autonomy, and people can be forced into it when they understand the risks and trade-offs and would rather make another decision. And those people matter a lot. But there are also, probably *more*, people, for whom the experience is “they were out of control, they were forcibly medicated, they said this is lots better, thank you, ok, now I’ll really really remember to take my meds in future”. Neither of those is a complete story. But the point is, the disease specifically destroys your ability to think you might be better with medication. Likewise, depression makes you think suicide might be a good idea whether it is or not. For most people, that *passes*, and they’re glad they didn’t.

        I want to cater to both those people — people with clinical depression, and people who’ve been mostly paralysed, and people who’ve just lost a loved one, etc, who feel NOW life isn’t worth living, but in six months, often feel like it is. And also, people for whom life actually ISN’T worth living and suicide is a rational choice (mostly but not always terminally ill). Most of the compromises are flawed, but I think it’s possible to come up with a reasonably good compromise that works for both groups most of the time.

        (And I think, the people who are most likely to want the latter route, are people who are not in a good position to take advantage of it being legal, even if some are.)

  69. Two McMillion says:

    It seems to me that if working at McDonalds as described would make you unhappy, then the problem is that you think happiness comes from things in the world.

    • Winter Shaker says:

      You mean, true happiness can only come from training yourself to be happy regardless of what things are in happening in your world, in a not having your desires satisfied but having no desires sort of way? In which case, I guess, a lifetime of drudge work done by someone skilled enough at mindfulness meditation or whatever other techniques are necessary to make it a pleasurable enough experience would be … pleasurable enough.

      But if you mean true happiness can only come in a hypothesised afterlife, beyond this vale of cheeseburgers, then as mentioned above in a comment to someone else, I think the questions should have been disaggregated into a ‘…conditional on a good afterlife’ and a ‘conditional on no afterlife’ pair of options. Though I’ll ask you the same question I asked the anonymous who seemed to have an infinite pain threshold: does your own answer to the question change between the ‘good afterlife’ and ‘no afterlife’ conditions?

      • Two McMillion says:

        That’s a complicated question; excuse me if I ramble.

        To be honest, I wasn’t really thinking of my comment in terms of Scott’s question at all. It just occurred to me as I was reading how very small the kind of happinesses I’ve encountered in life have been. That’s not to say I haven’t had a good life, or that I haven’t experienced euphoria at times- but I have a sense that happiness is cup I’ve only had sips from. In other words, I distinguish between what counts as happiness in my daily life, and a sort of nebulous, platonic-ideal idea of happiness that, in my imagination, is much deeper and sweeter than what I taste in reality. And it seems to me that when people discuss this existential sort of “happiness”, as we are in this thread, that they often really mean the second, but they try to reach it through the first. Well, as far as I can tell, that can’t be done.

        I sometimes think about this xkcd comic:

        I have come to realize that what that comic says disparagingly I say in truth. There really is an infinity of wonder in every passing second, a universe of joy in a single atom, if only we have eyes to see it, including, yes, in pictures of Joe Biden eating a sandwich. Randall’s interpretation- that our brains have one scale for experiences- is one possible explanation, of course. But the other explanation is that there genuinely is that much joy at large in the universe. And if that’s the case, then I find that I have spent most of my life barely scratching the surface of it. It would not surprise me to learn that there exist great wells of happiness which most humans are blind to. So I wasn’t really thinking with reference to the afterlife at all.

        • Winter Shaker says:

          That sounds like a reasonable take on things. From what I have read, a lot of people do seem to be recommending various meditational techniques as the key to, if not, happiness as such, the avoidance of the sense of persistent dissatisfaction – even if you don’t get to drink from the cup of happiness much, you can stop guzzling from the fountain of meh. But I haven’t got round to giving it a go yet; it will have to join the queue of things to add to my beeminders after the music I’m trying to practice, the languages I’m trying to learn and the in-shape I’m trying to keep, and I barely have time for those. I did recently get Sam Harris’s ‘Waking up’ book, though, which seems to be marketed as the closest thing to a guide to spirituality (in the contemplative sense, not the supernatural sense) for the sort of people who read this sort of blog. Maybe it will inspire me to have a go.

  70. onyomi says:

    I didn’t see the original survey, but I would just like to add myself as a data point against hypothesis one: I don’t have strong intuitions about a “right” answer to any of these questions.

  71. eponymous says:

    Question 5 is a tough one, and I think a core of some of what you’re getting at here.

    Personally, I think it’s extremely hard to argue on strictly consequentialist grounds that there isn’t an obligation to bring sentient beings into existence comparable with the obligation to not kill existing sentient beings.

    A simple thought experiment clarifies this. Suppose you have before you two buttons, one marked “life” and one “death”. In scenario (A), pressing “death” kills a person in the next room, while pressing “life” does nothing. In scenario (B), pressing “death” does nothing, while pressing “life” brings into existence the person in scenario (A).

    By construction, the consequences of either action are *exactly* the same. Thus any moral difference between them must derive from the state of the world that prevailed before the action, i.e. that because the person in scenario (A) desires not to die, it is wrong to kill him, whereas under scenario (B) there is no person to desire anything, and thus it is not wrong to fail to bring him into existence. Ergo, it’s not just about the consequences.

    Thinking about this, I’m coming around to the view that there is a strong ethical obligation to create sentient beings (e.g. to reproduce), although I’m not a strict consequentialist. And I’m starting to think that if you don’t take the trillions and trillions of people, you’re failing to add properly. (Though I would rather there were trillions and trillions of happy people.)

    (This seems related to Eliezer’s puzzle about not being willing to take a gamble with a very low probability of immortality and a very high probability of your life ending immediately, versus the mortal status quo. Both are failures to add up really big numbers, which humans are bad at.)

    • Jiro says:

      By construction, the consequences of either action are *exactly* the same.

      The static description of the state of the world after either action is the same. But the two worlds have different histories and those histories can affect how you value the worlds.

      • eponymous says:

        I take “consequentialism” to mean just that one only considers future states, and not histories, when making ethical determinations.

    • Said Achmiz says:

      Personally, I think it’s extremely hard to argue on strictly consequentialist grounds that there isn’t an obligation to bring sentient beings into existence comparable with the obligation to not kill existing sentient beings.

      Absurd; it’s trivial to argue this (unless you’re conflating consequentialism with utilitarianism, which is a very annoying mistake for people to keep making).

      By construction, the consequences of either action are *exactly* the same.

      Uh, no? They’re not? In one case, a living person gets killed. In the other, no such thing happens. Instead, in the other case, a new person is brought into existence, which doesn’t happen in the first case. Those are obviously different consequences. How in the world you could suggest that they’re the same, much less *exactly* the same, baffles me.

      Or are you taking an extremely narrow view of “consequences”, such that “consequences” mean “the state of the world at a specific moment in time — a moment which I will select for thought-experimental convenience — and not anything else”? Well, then I take serious issue with your quite distorted use of the term and concept of “consequences”.

      Thinking about this, I’m coming around to the view that there is a strong ethical obligation to create sentient beings (e.g. to reproduce), although I’m not a strict consequentialist.

      Ok, now you’re clearly conflating consequentialism with utilitarianism. (And — judging by the “failing to add properly” comment — apparently with specifically total utilitarianism, at that, which is even more egregious.)


      For completeness, here is the full consequentialist argument against an obligation to bring people into existence:

      1. I judge consequences by an evaluation function roughly along the lines of “are the consequences of this action/behavior [I am being deliberately ambiguous between act and rule consequentialism here] such that they result in a world I better like and would more prefer to live in, or such that they don’t?”
      2. Do I want to live in a world where a large number of people are created? Do I prefer that world to a world in which there are fewer people? Do I have any preference for more people? Not really, no.
      3. I therefore don’t have any obligation to create people.


      (My actual ethical views are more complicated than this, but the above is something like their “nearest point in consequentialism-space”)

      • eponymous says:

        It’s quite possible that I mean “utilitarianism” when I say “consequentialism”.

        I take the “consequences” of an action to refer to all future states of the world after the action is taken. I take “consequentialism” to be the view that the ethics of an action are determined only by the consequences of the action. Thus if all future states of the world are identical under two actions, then under any strictly consequentialist ethic one must be indifferent between them.

        Let’s take a simplified example to illustrate. Suppose that there are two time period, t = 1 and t = 2. Event X(t) is that person X exists at time t, and event ~X(t) is likewise that person X does not exist at time t.

        Scenario A asks you to choose between:
        {X(1) X(2)} and {X(1), ~X(2)}

        Scenario B asks you to choose between:
        {~X(1) X(2)} and {~X(1) ~X(2)}

        Since you make your choice after t=1, but before t=2, my understanding of consequentialism is that you *cannot consider the state at time 1*. Thus both choices are identically between:
        {X(2)} and {~X(2)}

        You seem to be saying that {X(1) ~X(2)} is *worse* than {~X(1) ~X(2)}, since in the first case a previously existing person ceases to exist. I agree with your preferences, because I prefer X(1) to ~X(1). But if you’re making a choice about time 2 only, my understanding of consequentialism is that you can only consider states in time 2, not in time 1.

        Now if it bothers you that there’s no idealized “between time 1 and 2” when you make the decision, we can modify the thought experiment slightly. Suppose that in scenario A, the person *ceases to exist* for a period of time before you press the button. Thus your choice is between creating a person who did not exist before, and *recreating* a person who existed previously.

        Now take the limit as the time of non-existence approaches zero.

        • Said Achmiz says:

          Re: consequentialism vs. utilitarianism:

          My point was that, even if you conclude that one must be indifferent between “creating a person” and “not destroying a person”, using this to demonstrate an obligation to do the former requires that you establish an obligation to do the latter. For that, you need some sort way to valuate consequences, and “consequentialism” as a category implies no particular evaluation function. (I assume that the evaluation function you have in mind is a utilitarian one, but maybe not, maybe you just take “number of humans existing in the world” as a direct measure of value, idk.)

          Re: your main point:

          Consequentialism does not impose the limitation you describe.

          The trivial objection is that I can certainly consider the world state at time 3, 4, 5, and so forth. The slightly less trivial objection is that you’re describing one particular dimension of variation between the two worlds, but there may be others; I am not limited to considering just the one. Now combine those, and consider that if two worlds have different histories, even if they are identical in some particular way, they may differ in all sorts of ways as time goes on. That, too, is part of “consequences”.

          In order to evade this objection, what you essentially have to do is construct some extremely contrived scenario where the result of two different actions is a pair of worlds which are particle-for-particle identical, and thus (given the same physical laws, which we can reasonably assume) will develop identically in the future, i.e. there is literally no way for anyone even in principle to tell, after the fact, which world was which.

          I am very skeptical of the notion that any intuition we may have about such a scenario, or any conclusion we might draw from testing our ethical theories upon it, have any applicability to any even remotely plausible real situation. (For instance, what does it even mean to talk about “killing a person” if what you’re actually doing is “constructing a universe identical to one where this person never existed”? Do you, yourself, also have no memory of this act? Or are you not part of the universe? Does this whole experimental setup have, somehow, no entanglement with the rest of the universe? Or what? Is it even coherent to speak of specific actions with specific consequences, in such a case?)

          Or, more pithily:

          Either one of your scenarios is better than the other, or they’re just not different scenarios. If they’re not different, the question is meaningless. If they’re different, then your answer is unsupportable.

          • eponymous says:


            I took an obligation to not end the existence of sentient beings (“murder”) for granted.

            By construction the scenarios are identical *after* the decision has been made. Yes, it is contrived — it’s a thought experiment, after all!

            what does it even mean to talk about “killing a person” if what you’re actually doing is “constructing a universe identical to one where this person never existed”?

            It means that this person existed *in the past*.

            Either one of your scenarios is better than the other, or they’re just not different scenarios. If they’re not different, the question is meaningless. If they’re different, then your answer is unsupportable.

            They are different *before* your decision, and the same *after*. Whether this means one is better or worse than the other depends on your ethics. I’m claiming that if you’re a consequentialist (as I understand the term) you should be indifferent. But you wouldn’t be indifferent under all ethical frameworks. For instance, if you think that murder is wrong because previously existing sentient beings have moral status and their wishes about their future existence must be respected, then you would treat the scenarios as different. (Here I’m assuming that the person wants to continue existing).

          • Jiro says:

            By construction the scenarios are identical *after* the decision has been made.

            No they’re not. The static descriptions of the scenarios are similar, but they have different historie. Your idea that two scenarios are “identical” even though they have different histories is not something you can just expect everyone to believe.

          • Earthly Knight says:

            eponymous, the individuals in the two scenarios you describe are qualitatively identical, but differ with respect to their causal-historical properties, as Jiro suggests. You seem to be presupposing a metaphysical principle which makes consequences supervene on future qualitative states of the world, something like “necessarily, any two actions which are alike with respect to the qualitative states of the world they bring about are alike with respect to their consequences.” But I do not see any reason why a consequentialist must accept this principle– we do not normally shy away from factoring causal-historical properties into our accounting of consequences. For instance, there does not seem to be anything amiss about saying “one of the consequences of President Trump’s decision to eliminate the Department of the Interior was that the house where Thomas Jefferson lived was bulldozed,” even though being the house where Thomas Jefferson lived is a historical property.

      • eponymous says:

        Regarding your argument against an ethical obligation to create sentient beings:

        Wouldn’t your argument work equally well as a justification for murder?

        • Adam says:

          The consequential outcomes of killing a person and not creating one are not the same. Maybe at a sufficiently future time they are, but certainly not at all times after the decision to act, and at a sufficiently future time, all outcomes are the same regardless of what anyone does ever. Foreknowledge of death creates fear and dread in the person you’re about to kill, there may be physical pain involved in the actual process leading to the kill, you create fear in other people and possibly widespread unrest in the general population depending upon how many people are out there killing others. You create grief in loved ones dealing with the loss. Some of these can possibly be avoided by maybe killing homeless people nobody cares about while they sleep painlessly and covering it up no one even finds the body. But that just gets at part of the issue. There are myriad ways to kill somebody, not all of which have the same consequences or impact to global utility.

          There is only one way to not create a person, with the common consequence that the future world has one fewer person than it otherwise would have had in it. But that common consequence is the only consequence of not creating a person. You can’t compare that with “killing” in the abstract. You have to compare it with specific methods of killing, all of which have different impacts on the world.

          • eponymous says:

            The value of a thought experiment is exactly that it allows you to “abstract from specific methods” to isolate what it is that makes something right or wrong.

            It’s true that, in reality, murder is a messy affair with all sorts of bad consequences besides a sentient being ceasing to exist, whereas choosing not to create a sentient being is the much cleaner decision to use a condom.

            Of course, in the future we may have to debate the morality of, e.g., flipping on or off a switch that powers a sentient AI, or choosing whether or not to resurrect a cryogenically frozen person. In such cases, I think we need to think carefully about the difference between flipping a switch to “off”, and failing to flip a switch to “on”.

          • Adam says:

            Sure, but what makes killing right or wrong is in those details. The only thing to “isolate” in a consequentialist framework is the consequences of the action. You abstract to common types of consequences with the consequences themselves being the only things that are intrinsically bad. Killing, and anything else, is only bad if it causes more of the bad consequences than the good ones. If your unit of abstraction is the action and not the consequence, then you’re practicing deontology, not consequentialism.

          • eponymous says:


            Okay, so it seems you are agreeing with me (and disagreeing with the person I was replying to). i.e. you don’t see any moral difference between the two scenarios I described in my thought experiment.

          • Adam says:

            Sort of, assuming you won’t cause any grief in the person’s loved ones or general dread in society that people are pushing buttons and killing each other. Even if you remove those considerations, though, if you’re going purely by a net utility computation, the only way to compute that is to know how much utility you stamped out in either case, so it depends upon the quality of life of whoever you either killed or failed to bring into existence, which frankly, is a weakness of utilitarianism and even consequentialism. Consequences aren’t always knowable and utility is not necessarily computable. The number of humans that exist in the world is surely not the only measure of goodness in the world.

        • Said Achmiz says:

          Wouldn’t your argument work equally well as a justification for murder?

          Yeah, totally, if I wanted to live in a world where everyone was murdering each other

  72. Rob says:

    New non-consequentialist, non-hedonist reader here. I don’t think the first two questions are answerable without more information.

    1) It matters why the McDonald’s worker has to work at McDonald’s 16+ hours a day. Is this slavery, economic necessity, or some kind of science fiction? I would choose life as an independent subsistence farmer over death. The choice between death and slavery would be harder, especially if every moment of my life as a slave would be lived according to a strictly prescribed routine and if there was no hope of escape.

    2) Is the person in scenario A unhappy because he spends his time doing meaningless or wicked things, or does he do valuable things and take no pleasure in them? Beethoven had a desirable life despite his felt unhappiness.

  73. Agronomous says:

    7. Would you rather:

    Option A: Live the rest of your life spending 16 hours a day, seven days a week, taking online ethics surveys. You will have no time off except the time you need to eat, sleep, use the restroom (you must wipe standing up, and the toilet-paper roll must be undershot), and complain about ambiguity and non-comprehensiveness of online ethics survey questions. Your entire life will be spent taking online ethics surveys. When you are no longer able to perform your tasks, you will die painlessly (where “pain” is determined based on online ethics surveys).

    Option B: Die painlessly right now.

    Option C: Die before even reading this question.

    (It’s possible I’m just bitter because I never got around to filling it out.)

    • Agronomous says:

      Also: for hand-washing purposes, you count as an employee.

      • Aegeus says:

        Is that supposed to be a bad thing? Every time I see one of those “Employees must wash hands” signs, I always want to scribble “…and so should everyone else” under it.

        • Two McMillion says:

          It removes the possible sadistic enjoyment of getting back at the world by infecting random strangers with possible infectious diseases.

          • Remember: Employees must wash hands before leaving restroom. They don’t specify when this needs to take place. So, you can enter, wash your hands, poop, and leave.

            You can get around this, of course, but by the time you’ve covered all the horrible hacky interpretations of the rules, you’ve basically written an expert system to run the restaurant and you don’t need the sentient beings in the first place.

    • Two McMillion says:

      You win the comments thread.

  74. Agronomous says:

    It’s either very revealing or very insignificant that so many commenters refer to the “100 trillion trillion” case as merely the “100 trillion” case. One comment even contains both references.

    On one hand, it’s a difference of twelve orders of magnitude. On the other hand, it’s only the difference between 10^14 and 10^26. On the gripping hand:

    (And I feel like I’ve done something like this here before….)

  75. Quite Likely says:

    “Every so often, someone on Reddit realizes that about half of people wipe themselves with toilet paper sitting down, and the other half do it standing up. This discovery is followed by horror on both sides that other people do it differently.”

    Haha and I guess I am one of those people today. Is it really roughly half who wipe standing up? That really does seem freakish to me. Can anyone link me to one of the reddit threads where people got to the bottom (no pun intended) of this issue?

  76. Quite Likely says:

    When I was younger I was more pro-suffering than I am now. I think the difference was experiencing a little more suffering in my life, and thus switching from a sort of blithe “one hundred years of torture is worth one afternoon of happiness” attitude to one that more understands how much suffering sucks.

    So if I am normal (which I don’t place a ton of probability on) I would expect youth and having had an easy life to correlate with being pro-suffering, while age and having experienced more suffering would be correlated with being more pro-oblivion.

  77. Jack V says:

    My prior observation has been that utilitarianism works pretty well for a pre-fixed number of people, mostly matching my intuitions but giving more consistent answers.

    And that my intuitions are not very consistent for any moral question including creating more people, and utilitarianism doesn’t help much. And as far as I can see, the same mostly applies to other people, except that many people confidently assert answers to moral questions including whether to create large numbers of new people, but I don’t know if they have a consistent framework behind that or not. (Eg. there are variants of utilitarianism referring to “total utility”, “average utility”, etc, but none really match my intuitions, and I don’t know if they match OTHER people’s, or the other people just haven’t considered the parts I don’t find palatable.)

    • Ankle says:

      The problem with utilitarianism and creating new people is that everybody just assumes human lives have net-positive experience value, when of course they really don’t.

      You can’t improve the world by adding more net-negative shit to it.

      Most people understand this on an intuitive level. But since we’re all brainwashed by a religious culture that pretends life is a gift from God, something to be grateful for and cling to unless the pain becomes “unbearable”, they experience cognitive dissonance.

      Once you understand the pains and pleasures of ordinary lives add up to a negative value rather than positive, the apparent contradiction goes away. The most important utilitarian goals are either to improve the average life quality drastically through speculative technology – or to prevent more of it from coming into existence.

  78. gwern says:

    Second, answers to all questions would correlate along a general factor of oblivion-preference versus suffering-preference. That is, people who would prefer oblivion to working at McDonalds would also be more likely to prefer a short life of happiness to a long life of unhappiness, et cetera.

    …Third, this factor would predict whether somebody endorsed a form of population ethics which promotes creating new people (question 3 is sort of just asking this already, but I also included a more direct question along those lines). [“Question 5. Creating new sentient beings is:”]

    …Fourth, this factor would predict some real-world consequences like whether people believed in a right to euthanasia and whether they were signed up for cryonics.

    …Fifth, happier people would be more likely to prefer suffering over oblivion, because they view life as generally excellent and so oblivion represents more of a sacrifice for them.

    Trying to combine these hypotheses into a statistical model, I get something like this:

    OBLIVION =~ 1+2+3+..+N
    Newpeople ~ OBLIVION
    Euthanasia ~ OBLIVION
    Cryonics ~ OBLIVION
    OBLIVION ~ Happiness

    I’ll need to look closer at the survey to figure out which ones go into OBLIVION (obviously, not the new-people, euthanasia-rights, cryonics, or happiness variables!).

    First, though, all the data needs to be downloaded and cleaned into ordered numbers we can feed into statistical models:

    ## $ wget
    ## $ dos2unix oblivionsurvey.csv
    ## $ cat oblivionsurvey.csv | iconv -c -tascii > foo.txt; mv foo.txt oblivionsurvey.csv # some joker put in Unicode crap

    ## Data munging/cleanup:
    oblivion <- read.csv("oblivionsurvey.csv")
    # [1] ""
    # [2] ""
    # [3] ""
    # [4] ""
    # [5] ""
    # [6] ""
    # [7] ""
    # [8] ""
    # [9] ""
    # [10] ""
    # [11] ""
    # [12] ""
    # [13] ""
    # [14] ""
    ## Give more memorable names to each variable:
    colnames(oblivion) <- c("McDonalds", "McDonalds.strength", "Longlife", "Longlife.strength", "Repugnantconclusion", "Repugnantconclusion.strength", "Newpeople", "Heavenhell.p", "Right.suicide", "Right.euthanasia", "Cryonics", "Happiness", "Utilitarian", "Religious")

    ## convert preference variable into something meaningful
    ## goddammit Yvain how many times do I need to tell you that free response is the devil? Maybe then half the data wouldn't be useless.
    ## ah fuck it, if people want their answers to matter, they can *give real numbers and stop being such precious special snowflakes*
    heaven ", "", gsub("%", "", as.character(oblivion$Heavenhell.p)))
    oblivion$Heavenhell.p 100,]$Heavenhell.p <- 100
    oblivion[!$Heavenhell.p) & oblivion$Heavenhell.p<0,]$Heavenhell.p <- 0

    ## convert non-answers to NAs
    oblivion[oblivion==""] <- NA
    oblivion[!$Newpeople) & oblivion$Newpeople=="I am not a consequentialist or otherwise don't want to answer this question",]$Newpeople <- NA

    ## encode orderings, then convert to integers 1-4
    oblivion$Newpeople <- as.integer(ordered(oblivion$Newpeople, levels=c("Bad", "Morally neutral", "Generally good, but less good than improving the lives of existing beings by the same amount", "Exactly comparable to improving the lives of existing beings. Creating a life that experiences 100 utils is exactly as good as improving existing lives 100 utils.")))
    oblivion$Right.suicide <- as.integer(ordered(oblivion$Right.suicide, levels=c("Strongly against", "Slightly against", "Neutral", "Slightly in favor" , "Strongly in favor")))
    oblivion$Right.euthanasia <- as.integer(ordered(oblivion$Right.euthanasia, levels=c("Strongly against", "Slightly against", "Neutral", "Slightly in favor" , "Strongly in favor")))
    oblivion$Cryonics <- as.integer(ordered(oblivion$Cryonics, levels=c("Not at all interested", "Somewhat interested", "Yes, already signed up or very interested")))
    oblivion$Happiness <- as.integer(ordered(oblivion$Happiness, levels=c("No, very unhappy", "No, somewhat unhappy", "Average", "Yes, somewhat happy", "Yes, very happy")))
    oblivion$Utilitarian <- as.integer(ordered(oblivion$Utilitarian, levels=c("No", "It's complicated", "Yes")))
    oblivion$Religious <- oblivion$Religious=="Yes"

    ## OK, now how do we deal with the bizarre set of questions "McDonalds", "McDonalds.strength", "Longlife", "Longlife.strength", "Repugnantconclusion", "Repugnantconclusion.strength"?
    ## Even though they amount to just a scale 1-6, they're split across two questions each with reversals! ;_;
    ## what we do is we define one response to the first question as +1, the other response as -1, and then convert the second question to an ordered factor 1/2/3
    ## then we simply multiply and now we have a scale ranked -3/-2/-1/0/1/2/3! Then we can recenter by adding 4. Tada!
    oblivion$McDonalds.strength <- ordered(oblivion$McDonalds.strength, levels=c("Ambivalent", "Somewhat strongly", "Very strongly"))
    oblivion$McDonalds <- ifelse(oblivion$McDonalds=="Die painlessly right now.", 1, -1) * as.integer(oblivion$McDonalds.strength) + 4
    oblivion$McDonalds.strength <- NULL

    oblivion$Longlife.strength <- ordered(oblivion$Longlife.strength, levels=c("Ambivalent", "Somewhat strongly", "Very strongly"))
    oblivion$Longlife <- ifelse(oblivion$Longlife=="Live a short but happy life. You die ten years from now, in 2026, after being hit by a car. Until then, you do fulfilling work, have happy relationships, and meet with success in most projects.", 1, -1) * as.integer(oblivion$Longlife.strength) + 4
    oblivion$Longlife.strength <- NULL

    oblivion$Repugnantconclusion.strength <- ordered(oblivion$Repugnantconclusion.strength, levels=c("Ambivalent", "Somewhat strongly", "Very strongly"))
    oblivion$Repugnantconclusion <- ifelse(oblivion$Repugnantconclusion=="A world with 1 million sentient beings, all of whom are happy and consider their world a utopia.", 1, -1) * as.integer(oblivion$Repugnantconclusion.strength) + 4
    oblivion$Repugnantconclusion.strength <- NULL

    # McDonalds Longlife Repugnantconclusion Newpeople Heavenhell.p Right.suicide Right.euthanasia Cryonics Happiness
    # Min. :1.0000 Min. :1.000000 Min. :1.000000 Min. :1.000000 Min. : 0.00000 Min. :1.000000 Min. :1.000000 Min. :1.00000 Min. :1.000000
    # 1st Qu.:2.0000 1st Qu.:3.000000 1st Qu.:6.000000 1st Qu.:3.000000 1st Qu.: 60.00000 1st Qu.:3.000000 1st Qu.:5.000000 1st Qu.:1.00000 1st Qu.:2.000000
    # Median :5.0000 Median :6.000000 Median :6.000000 Median :3.000000 Median : 98.00000 Median :4.000000 Median :5.000000 Median :2.00000 Median :4.000000
    # Mean :4.3663 Mean :4.959707 Mean :5.954128 Mean :2.912664 Mean : 77.37208 Mean :3.693859 Mean :4.589204 Mean :1.80831 Mean :3.382002
    # 3rd Qu.:6.0000 3rd Qu.:6.000000 3rd Qu.:7.000000 3rd Qu.:3.000000 3rd Qu.:100.00000 3rd Qu.:5.000000 3rd Qu.:5.000000 3rd Qu.:2.00000 3rd Qu.:4.000000
    # Max. :7.0000 Max. :7.000000 Max. :7.000000 Max. :4.000000 Max. :100.00000 Max. :5.000000 Max. :5.000000 Max. :3.00000 Max. :5.000000
    # NA's :2 NA's :2 NA's :4 NA's :178 NA's :223 NA's :3 NA's :1 NA's :35 NA's :5
    # Utilitarian Religious
    # Min. :1.000000 Mode :logical
    # 1st Qu.:2.000000 FALSE:848
    # Median :2.000000 TRUE :134
    # Mean :2.254596 NA's :112
    # 3rd Qu.:3.000000
    # Max. :3.000000
    # NA's :6
    cor.plot(lowerCor(oblivion, use="pairwise.complete.obs"),numbers=TRUE)
    # McDnl Lnglf Rpgnn Nwppl Hvnh. Rght.s Rght.t Crync Hppns Utltr Relgs
    # McDonalds 1.00
    # Longlife 0.34 1.00
    # Repugnantconclusion 0.22 0.21 1.00
    # Newpeople -0.05 0.01 -0.26 1.00
    # Heavenhell.p 0.23 0.15 0.05 0.01 1.00
    # Right.suicide 0.17 0.11 0.06 -0.05 0.19 1.00
    # Right.euthanasia 0.22 0.12 0.14 0.02 0.30 0.58 1.00
    # Cryonics -0.11 -0.23 -0.06 0.09 -0.04 0.19 0.17 1.00
    # Happiness -0.21 0.07 -0.07 0.09 -0.06 -0.07 -0.07 0.03 1.00
    # Utilitarian 0.04 0.00 -0.04 0.17 0.17 0.22 0.34 0.26 0.02 1.00
    # Religious -0.07 0.05 0.01 0.05 -0.25 -0.35 -0.51 -0.21 0.08 -0.27 1.00

    ## data munging & cleanup finished!
    write.csv(oblivion, file="2016-06-30-oblivionsurvey-clean.csv", row.names=FALSE)

    • gwern says:

      ## Analysis
      oblivion <- read.csv("2016-06-30-oblivionsurvey-clean.csv")
      model1 <- '
      # second hypothesis: general factor of intrinsic life-preference
      OBLIVION =~ McDonalds + Longlife + Repugnantconclusion + Heavenhell.p + Right.suicide

      # regressions:
      # third hypothesis:
      Newpeople ~ OBLIVION
      # fourth hypothesis:
      Right.euthanasia ~ OBLIVION
      Cryonics ~ OBLIVION
      # fifth hypothesis
      OBLIVION ~ Happiness

      # I add this one because I am sure it is true:
      Right.euthanasia ~~ Right.suicide
      s1 |z|)
      # OBLIVION =~
      # McDonalds 1.000
      # Longlife 0.658 0.089 7.360 0.000
      # Repugnntcnclsn 0.383 0.068 5.602 0.000
      # Heavenhell.p 8.838 1.560 5.665 0.000
      # Right.suicide 0.230 0.049 4.699 0.000
      # Regressions:
      # Estimate Std.Err Z-value P(>|z|)
      # Newpeople
      # OBLIVION -0.065 0.031 -2.106 0.035
      # Right.euthanasia ~
      # OBLIVION 0.230 0.046 5.035 0.000
      # Cryonics ~
      # OBLIVION -0.086 0.027 -3.140 0.002
      # OBLIVION ~
      # Happiness -0.235 0.064 -3.685 0.000
      # Covariances:
      # Estimate Std.Err Z-value P(>|z|)
      # Right.suicide ~~
      # Right.euthanas 0.539 0.050 10.725 0.000
      # Newpeople ~~
      # Right.euthanas 0.052 0.021 2.480 0.013
      # Cryonics 0.035 0.017 2.076 0.038
      # Right.euthanasia ~~
      # Cryonics 0.068 0.016 4.365 0.000
      # Intercepts:
      # Estimate Std.Err Z-value P(>|z|)
      # McDonalds 5.160 0.225 22.890 0.000
      # Longlife 5.481 0.120 45.512 0.000
      # Repugnntcnclsn 6.257 0.083 75.543 0.000
      # Heavenhell.p 84.501 2.122 39.824 0.000
      # Right.suicide 3.876 0.064 60.115 0.000
      # Newpeople 2.857 0.036 79.462 0.000
      # Right.euthanas 4.772 0.050 95.928 0.000
      # Cryonics 1.740 0.030 57.462 0.000
      # OBLIVION 0.000
      # Variances:
      # Estimate Std.Err Z-value P(>|z|)
      # McDonalds 2.153 0.253 8.519 0.000
      # Longlife 2.731 0.174 15.695 0.000
      # Repugnntcnclsn 1.893 0.122 15.561 0.000
      # Heavenhell.p 937.099 60.805 15.412 0.000
      # Right.suicide 1.587 0.064 24.815 0.000
      # Newpeople 0.529 0.024 21.617 0.000
      # Right.euthanas 0.655 0.055 11.959 0.000
      # Cryonics 0.408 0.015 26.611 0.000
      # OBLIVION 1.748 0.230 7.610 0.000
      # npar fmin chisq df pvalue baseline.chisq baseline.df baseline.pvalue cfi
      # 29.000 0.126 276.363 23.000 0.000 1154.690 36.000 0.000 0.774
      # tli nnfi rfi nfi pnfi ifi rni logl unrestricted.logl
      # 0.646 0.646 0.625 0.761 0.486 0.776 0.774 -17328.166 -17189.985
      # aic bic ntotal bic2 rmsea rmsea.pvalue rmr
      # 34714.332 34859.236 1093.000 34767.126 0.100 0.090 0.111 0.000 1.279
      # rmr_nomean srmr srmr_bentler srmr_bentler_nomean srmr_bollen srmr_bollen_nomean srmr_mplus srmr_mplus_nomean cn_05
      # 1.401 0.067 0.067 0.073 0.067 0.073 0.067 0.073 140.105
      # cn_01 gfi agfi pgfi mfi ecvi
      # 165.678 0.997 0.993 0.425 0.891 NA
      semPaths(s1, "std", exoVar = FALSE, exoCov = FALSE, nCharNodes=15, layout="tree2", sizeMan=10, sizeMan2=10, residuals=FALSE, label.prop=1.5, edge.label.cex=0.9)

      How do the predictions pan out?

      1. second hypothesis: As predicted, and we might’ve guessed from looking at the correlation table and noting the many moderate intercorrelations suggesting a few latent factors at work, the 5 questions seem to form a good ‘oblivion’ or ‘quality of life’ factor
      2. third hypothesis: the OBLIVION factor does predict beliefs about how moral it is to create new people but it doesn’t seem to explain that much of the variance: 1 higher unit of OBLIVION factor predicts less favorability by 0.065 units on the 1-4 scale which has a SD of 0.7.
      3. fourth hypothesis: it definitely predicts euthanasia beliefs, almost 3x as much as creating-new-people; it also predicts cryonics but more along the lines of creating-new-people. Interestingly, the sign for that is the opposite of what I would’ve predicted.
      4. fifth hypothesis: happy people definitely have lower OBLIVION scores.

      (The means are generally fairly intermediate, so the first hypothesis could be considered correct as well.)
      Overall, all the hypotheses look vindicated by the SEM/confirmatory factor analysis.

      I actually didn’t read the rest of the post because I read the listed hypotheses and went ‘a well-defined theoretical model relating survey responses to a latent factor which maps perfectly onto a SEM! I’ve been training all my life for this moment!‘ and I started work. (The data cleaning took… a long time.)
      So let’s see. “The second hypothesis was weakly confirmed”; I disagree, I think it was stronger but I used a different set of variables to construct my OBLIVION factor. With Likert scales like this, with only a handful of questions per latent factor (at most), there is going to be a lot of random noise, there is going to be a lot of measurement error, there’s going to be noise due to minor details like wording and people preferring more or less extreme responses, and even constructs which are strongly related will yield relatively small raw correlations. If you tried to measure IQ with less than 14 multiple-choice questions, you would not get a very good measurement of intelligence (maybe r=0.5 at best), and similarly here – you’re not going to get a good measure of any ethical factor without asking many overlapping questions/hypotheticals to overcome those issues.
      “The third hypothesis was not supported.” This one was a small effect and since Yvain only looked at first-order correlations, they were naturally unimpressive and not statistically-significant.
      “The fourth and fifth hypotheses were weakly supported.” more than weakly, again, probably masked by not using a latent variable.

      • gwern says:

        I think that some of it is a little off, though, and the OBLIVION factor is being underestimated. In particular, I would absolutely expected Right.euthanasia and Right.suicide to also load on OBLIVION. What does an exploratory factor analysis or clustering say?

        One clustering algorithm:

        # ...
        # Cluster size:
        # C6 C9
        # 6 5
        # Item by Cluster Structure matrix:
        # O P C6 C9
        # McDonalds C9 C9 0.20 -0.56
        # Longlife C9 C9 0.03 -0.36
        # Repugnantconclusion C9 C9 0.04 -0.48
        # Newpeople C9 C9 0.06 0.27
        # Heavenhell.p C9 C6 0.33 -0.26
        # Right.suicide C6 C6 0.61 -0.24
        # Right.euthanasia C6 C6 0.78 -0.28
        # Cryonics C6 C6 0.31 0.28
        # Happiness C9 C9 -0.07 0.20
        # Utilitarian C6 C6 0.47 0.10
        # Religious C6 C6 -0.61 0.07
        # With eigenvalues of:
        # C6 C9
        # 1.8 1.1
        # Purified scale intercorrelations
        # reliabilities on diagonal
        # correlations corrected for attenuation above diagonal:
        # C6 C9
        # C6 0.68 -0.15
        # C9 -0.08 0.44
        # Cluster fit = 0.48 Pattern fit = 0.97 RMSR = 0.06

        Hm… Complicated, not particular interpretable. The suicide/euthanasia seems to be tapping into a religion-related variable of some sort. Moving onto factor analysis:

        # VSS complexity 1 achieves a maximimum of 0.61 with 7 factors
        # VSS complexity 2 achieves a maximimum of 0.73 with 7 factors
        # The Velicer MAP achieves a minimum of 0.03 with 1 factors
        # Empirical BIC achieves a minimum of -29.25 with 4 factors
        # Sample Size adjusted BIC achieves a minimum of -0.57 with 6 factors
        # Statistics by number of factors
        # vss1 vss2 map dof chisq prob sqresid fit RMSEA BIC SABIC complex eChisq SRMR eCRMS eBIC
        # 1 0.39 0.00 0.025 44 6.6e+02 1.6e-111 9.2 0.39 0.114 355.4 495.18 1.0 1.3e+03 1.0e-01 0.115 955.8
        # 2 0.48 0.56 0.030 34 2.9e+02 1.2e-42 6.6 0.56 0.083 53.5 161.48 1.3 3.7e+02 5.6e-02 0.071 136.2
        # 3 0.53 0.63 0.044 25 1.6e+02 4.3e-22 5.3 0.65 0.071 -12.6 66.80 1.5 1.7e+02 3.8e-02 0.056 -2.5
        # 4 0.55 0.67 0.069 17 9.4e+01 1.3e-12 4.6 0.70 0.065 -25.2 28.77 1.5 9.0e+01 2.7e-02 0.049 -29.2
        # 5 0.54 0.65 0.096 10 7.1e+01 2.3e-11 4.6 0.70 0.075 1.5 33.23 1.7 7.2e+01 2.4e-02 0.057 1.9
        # 6 0.50 0.69 0.132 4 1.5e+01 5.3e-03 3.5 0.77 0.050 -13.3 -0.57 1.7 1.2e+01 9.9e-03 0.037 -16.1
        # 7 0.61 0.73 0.202 -1 8.2e-01 NA 2.9 0.81 NA NA NA 1.6 4.5e-01 1.9e-03 NA NA
        # 8 0.48 0.63 0.297 -5 4.7e-06 NA 3.0 0.80 NA NA NA 1.6 4.8e-06 6.3e-06 NA NA
        # 9 0.47 0.60 0.474 -8 4.4e-10 NA 2.9 0.80 NA NA NA 1.7 3.4e-10 5.3e-08 NA NA
        # 10 0.48 0.62 1.000 -10 3.8e-12 NA 2.6 0.83 NA NA NA 1.8 3.3e-12 5.2e-09 NA NA
        # 11 0.52 0.66 NA -11 0.0e+00 NA 2.1 0.86 NA NA NA 1.5 1.2e-25 9.9e-16 NA NA
        fa(oblivion, nfactors=4)
        # Factor Analysis using method = minres
        # Call: fa(r = oblivion, nfactors = 4)
        # Standardized loadings (pattern matrix) based upon correlation matrix
        # MR1 MR2 MR3 MR4 h2 u2 com
        # McDonalds 0.05 0.02 0.13 0.63 0.50 0.50 1.1
        # Longlife 0.02 0.00 0.78 0.03 0.64 0.36 1.0
        # Repugnantconclusion 0.05 -0.27 0.19 0.15 0.17 0.83 2.5
        # Newpeople 0.00 0.91 0.01 0.01 0.83 0.17 1.0
        # Heavenhell.p 0.28 0.03 0.07 0.19 0.16 0.84 1.9
        # Right.suicide 0.65 -0.06 0.05 0.00 0.43 0.57 1.0
        # Right.euthanasia 0.85 0.02 0.05 0.03 0.75 0.25 1.0
        # Cryonics 0.32 0.09 -0.26 -0.14 0.19 0.81 2.5
        # Happiness 0.02 0.03 0.29 -0.47 0.19 0.81 1.7
        # Utilitarian 0.43 0.18 -0.05 -0.06 0.21 0.79 1.4
        # Religious -0.61 0.06 0.14 0.01 0.38 0.62 1.1
        # MR1 MR2 MR3 MR4
        # SS loadings 1.90 0.96 0.87 0.74
        # Proportion Var 0.17 0.09 0.08 0.07
        # Cumulative Var 0.17 0.26 0.34 0.41
        # Proportion Explained 0.43 0.21 0.19 0.17
        # Cumulative Proportion 0.43 0.64 0.83 1.00
        # With factor correlations of
        # MR1 MR2 MR3 MR4
        # MR1 1.00 0.01 0.08 0.25
        # MR2 0.01 1.00 0.01 -0.14
        # MR3 0.08 0.01 1.00 0.40
        # MR4 0.25 -0.14 0.40 1.00
        # Mean item complexity = 1.5
        # Test of the hypothesis that 4 factors are sufficient.
        # The degrees of freedom for the null model are 55 and the objective function was 1.7 with Chi Square of 1845.26
        # The degrees of freedom for the model are 17 and the objective function was 0.09
        # The root mean square of the residuals (RMSR) is 0.03
        # The df corrected root mean square of the residuals is 0.05
        # The harmonic number of observations is 988 with the empirical chi square 79.58 with prob < 4.6e-10
        # The total number of observations was 1094 with MLE Chi Square = 93.74 with prob < 1.3e-12
        # Tucker Lewis Index of factoring reliability = 0.861
        # RMSEA index = 0.065 and the 90 % confidence intervals are 0.052 0.077
        # BIC = -25.22
        # Fit based upon off diagonal values = 0.98
        # Measures of factor score adequacy
        # MR1 MR2 MR3 MR4
        # Correlation of scores with factors 0.91 0.91 0.83 0.77
        # Multiple R square of scores with factors 0.83 0.83 0.68 0.59
        # Minimum correlation of possible factor scores 0.65 0.66 0.37 0.17
        fa.diagram(fa(oblivion, nfactors=4))

        The factor analysis suggests 4 factors, which turn out to be:

        1. MR1 loads on Heavenhell.p, Right.suicide, Right.euthanasia, Cryonics, Utilitarian, anti-Religious
        2. MR2 loads on anti-Repugnantconclusion, Newpeople, Utilitarian
        3. MR3 loads on McDonalds, Longlife, Repugnantconclusion, anti-Cryonics, Happiness, Religious
        4. MR4 loads on McDonalds, Repugnantconclusion, Heavenhell.p, anti-Cryonics, anti-Happiness

        MR1 looks like utilitarians with high standards for life: hopeful for the future, not expecting an afterlife but if there is one they would only be happy about it if they had a really good chance for heaven.
        MR2 looks like utilitarians who bite the bullet on the Repugnant Conclusion and the general question of making more people: it’s good to create lots of low-quality lives.
        MR3 looks like people in general who have high standards for life being worth living and don’t feel the need to prolong it: they would prefer to die in the McDonalds, live a short but glorious life, are happy, don’t feel any need for cryonics, and are a bit religious.
        MR4 is weird. It’s mostly like MR3, except they really would prefer to the die in the McDonalds, and they are very unhappy.

        Playing with omega() to visualize other factorizations, the factorization I find most preferable is a 2-factor solution, which yields an immediately interpretable OBLIVION and a sort of religion factor:

        fa(oblivion, nfactors=2)
        # Factor Analysis using method = minres
        # Call: fa(r = oblivion, nfactors = 2)
        # Standardized loadings (pattern matrix) based upon correlation matrix
        # MR1 MR2 h2 u2 com
        # McDonalds 0.16 0.54 0.340 0.66 1.2
        # Longlife 0.02 0.54 0.299 0.70 1.0
        # Repugnantconclusion 0.05 0.37 0.146 0.85 1.0
        # Newpeople 0.05 -0.17 0.030 0.97 1.1
        # Heavenhell.p 0.31 0.20 0.157 0.84 1.7
        # Right.suicide 0.64 0.05 0.426 0.57 1.0
        # Right.euthanasia 0.86 0.07 0.759 0.24 1.0
        # Cryonics 0.32 -0.41 0.232 0.77 1.9
        # Happiness -0.07 -0.13 0.024 0.98 1.5
        # Utilitarian 0.44 -0.18 0.204 0.80 1.3
        # Religious -0.61 0.14 0.369 0.63 1.1
        # MR1 MR2
        # SS loadings 1.95 1.04
        # Proportion Var 0.18 0.09
        # Cumulative Var 0.18 0.27
        # Proportion Explained 0.65 0.35
        # Cumulative Proportion 0.65 1.00
        # With factor correlations of
        # MR1 MR2
        # MR1 1.00 0.14
        # MR2 0.14 1.00
        # Mean item complexity = 1.3
        # Test of the hypothesis that 2 factors are sufficient.
        # The degrees of freedom for the null model are 55 and the objective function was 1.7 with Chi Square of 1845.26
        # The degrees of freedom for the model are 34 and the objective function was 0.27
        # The root mean square of the residuals (RMSR) is 0.06
        # The df corrected root mean square of the residuals is 0.07
        # The harmonic number of observations is 988 with the empirical chi square 329.03 with prob < 5.4e-50
        # The total number of observations was 1094 with MLE Chi Square = 291.41 with prob < 1.2e-42
        # Tucker Lewis Index of factoring reliability = 0.767
        # RMSEA index = 0.083 and the 90 % confidence intervals are 0.075 0.092
        # BIC = 53.49
        # Fit based upon off diagonal values = 0.92
        # Measures of factor score adequacy
        # MR1 MR2
        # Correlation of scores with factors 0.91 0.77
        # Multiple R square of scores with factors 0.83 0.59
        # Minimum correlation of possible factor scores 0.66 0.19
        # fa.diagram(fa(oblivion, nfactors=2))

        You see what I mean:

        1. MR1 loads primarily on Heavenhell.p, Right.suicide, Right.euthanasia, Cryonics, Utilitarian, anti-Religious
        2. MR2 loads primarily on McDonalds, Longlife, Repugnantconclusion, anti-Newpeople, Heavenhell.p, Right.suicide, Right.euthanasia, anti-Cryonics

        One factor expresses general utilitarian/anti-Catholicism moral beliefs; the other looks like the OBLIVION factor.

        When Yvain says

        So it looks like people have very different opinions about when to choose death versus suffering, but that these opinions are inconsistent and only weakly driven by broad cross-situation intuitions.

        I would interpret it more as there are two cross-cutting concerns people have: whether to be utilitarian, or whether to desire a high quality of life.

        • gwern says:

          What if we take those MR1/MR2, add in the Happiness regression (since happiness turned out to generally be unrelated to most factors), and compare this bifactor model to Yvain’s single-factor model w/regressions?
          Regular frequentist model fits aren’t too helpful but we can switch to blavaan and get a Bayes factor to choose between the two competing models (although note that doesn’t guarantee that either one is a good model, the BF just tells us which of the two is better than the other):

          model2 <- '
          # MR2
          OBLIVION =~ McDonalds + Longlife + Repugnantconclusion + Newpeople + Heavenhell.p + Right.suicide + Right.euthanasia + Cryonics
          # MR1
          RELIGION =~ Heavenhell.p + Right.suicide + Right.euthanasia + Cryonics + Utilitarian + Religious
          # regressions: only missing variable is Happiness now
          # fifth hypothesis
          OBLIVION ~ Happiness

          # I add this one because I am sure it is true:
          Right.euthanasia ~~ Right.suicide
          s2 |z|)
          # OBLIVION =~
          # McDonalds 1.000
          # Longlife 0.777 0.109 7.131 0.000
          # Repugnntcnclsn 0.431 0.070 6.128 0.000
          # Newpeople -0.080 0.032 -2.475 0.013
          # Heavenhell.p 7.913 1.349 5.866 0.000
          # Right.suicide 0.209 0.041 5.090 0.000
          # Right.euthanas 0.202 0.033 6.212 0.000
          # Cryonics -0.145 0.031 -4.692 0.000
          # RELIGION =~
          # Heavenhell.p 1.000
          # Right.suicide 0.065 0.012 5.309 0.000
          # Right.euthanas 0.061 0.011 5.756 0.000
          # Cryonics 0.023 0.004 5.177 0.000
          # Utilitarian 0.034 0.006 5.532 0.000
          # Religious -0.022 0.004 -6.112 0.000
          # Regressions:
          # Estimate Std.Err Z-value P(>|z|)
          # OBLIVION ~
          # Happiness -0.191 0.065 -2.936 0.003
          # Covariances:
          # Estimate Std.Err Z-value P(>|z|)
          # Right.suicide ~~
          # Right.euthanas 0.156 0.043 3.655 0.000
          # Intercepts:
          # Estimate Std.Err Z-value P(>|z|)
          # McDonalds 5.013 0.235 21.371 0.000
          # Longlife 5.462 0.145 37.724 0.000
          # Repugnntcnclsn 6.232 0.093 67.081 0.000
          # Newpeople 2.864 0.037 77.396 0.000
          # Heavenhell.p 82.335 1.985 41.486 0.000
          # Right.suicide 3.829 0.062 61.548 0.000
          # Right.euthanas 4.720 0.045 103.968 0.000
          # Cryonics 1.713 0.034 50.646 0.000
          # Utilitarian 2.255 0.023 98.452 0.000
          # Religious 0.141 0.011 13.159 0.000
          # OBLIVION 0.000
          # RELIGION 0.000
          # Variances:
          # Estimate Std.Err Z-value P(>|z|)
          # McDonalds 2.385 0.243 9.829 0.000
          # Longlife 2.560 0.178 14.410 0.000
          # Repugnntcnclsn 1.864 0.124 15.033 0.000
          # Newpeople 0.526 0.025 21.055 0.000
          # Heavenhell.p 873.774 58.484 14.940 0.000
          # Right.suicide 1.149 0.072 15.856 0.000
          # Right.euthanas 0.286 0.032 8.846 0.000
          # Cryonics 0.332 0.018 18.028 0.000
          # Utilitarian 0.432 0.020 21.729 0.000
          # Religious 0.065 0.007 9.978 0.000
          # OBLIVION 1.539 0.220 6.987 0.000
          # RELIGION 109.621 34.045 3.220 0.001
          anova(s1, s2)
          # Chi Square Difference Test
          # Df AIC BIC Chisq Chisq diff Df diff Pr(>Chisq)
          # s1 23 34714.332 34859.236 276.36294
          # s2 39 254.47648 -21.886466 16 1

          ## need to standardize/rescale to avoid a blavaan bug:
          b1 <- bsem(model1, n.chains=6, jagcontrol=list(method="rjparallel"), fixed.x=FALSE, data=scale(oblivion))
          b2 <- bsem(model2, n.chains=6, jagcontrol=list(method="rjparallel"), fixed.x=FALSE, data=scale(oblivion))
          BF(b1, b2)
          # Laplace approximation to the log-Bayes factor (experimental):
          # 2231.391

          So the chi-squared looks better for the bi-factor model and the log BF indicates much stronger evidence for the second model.

          If one were going to extend this line of inquiry further, I think it would be particularly valuable to ask about what kind of religion (Christian, especially Catholic), age, how many kids one has, and add in more variants of the existing population-dilemmas as well as some others existence-related questions like abortion or the morality of having more than 2 kids.
          And eliminate the free-response fields.

          Full version as a pastebin:

  79. Protest Manager says:

    My answers:
    1: A: McDonalds cashier. The job doesn’t take all my brain, leaving me with plenty of time to think, and to interact with other people.

    2. A: I can still do and accomplish things, and kill myself if I decide it’s not worth it. Death, OTOH, is more or less final

    3. Neither. “Utopia’s” suck

    4. 0% Knowing that the evil would be punished for their actions is enough

    BTW, you have asked the question poorly: “0% certainty of going to Heaven” != “okay with certain damnation”. Having no certainty, at all, that they will go to Heaven is the lot of most Christians. That is utterly different from being certain one will go to Hell

    5: Creating new human beings, AKA “investing in the future”, is morally superior to helping existing ones. Especially in a world that 1: Isn’t particularly overcrowded, & 2, has finite life spans

    6?: Whether you believe everybody has a right to commit suicide if they want, including people who are not terminally ill.

    Of course they do. The fact that Roe v. Wade created a “right” to kill inconvenient human beings who were dependent upon you, but NOT a right to kill yourself, shows the utter illegitimacy of the whole “I have a right to control my own body” argument.

    People who have the right to control their own body have the right to kill themselves, the right to take drugs that could harm them, and the right to sell their blood / organs / sex.

    7?: Whether you are interested in signing up for cryonics.No. Don’t believe the science is well enough advanced to justify it

    8?: And whether you consider yourself a happy person. Most of the time

  80. 57dimensions says:

    I guess it’s not the norm, but I would be completely ok with oblivion. And since I’m an atheist, that is what I believe happens when someone dies, they stop existing and experiencing anything at all. In my mind, since it’s oblivion, I’m not going to be conscious or sentient anymore, so I’m literally incapable of experiencing what that is like, so how could I regret being there? Once ‘oblivion’ happens, you obviously can’t know that’s what happened or that you no longer exist, so why would you care? I don’t like suffering, I will avoid it at all costs. To me existing and suffering could never be better than not existing. I can’t wrap my mind around why anyone would prefer even mild suffering to oblivion.

    I’ve been severely depressed for almost 4 years now (ages 14-18) and never had great mental health before that, so I’m guessing that’s probably skewing my views on this. I’ve gotten very close to committing suicide before, and although I’m ‘happier’ now I still would be completely ok with ‘oblivion’ happening to me right now. The only reasons I wouldn’t actually kill myself are that it’s actually a lot harder to do than you would think, I’m a wimp and don’t want to experience any pain, and I know my family would be devastated. But if I could completely remove those factors, oblivion doesn’t scare me at all. I’m not sure if I would immediately choose that over continuing to live, but it doesn’t sound awful or unappealing. Like, life isn’t horribly painful or unbearable right now, but it still takes effort and isn’t totally amazing, and not having to deal with all my problems doesn’t sound all that terrible.

    • Orphan Wilde says:

      So, for depression… try Vitamin D. Large dose; ~10,000 IUs. Someone else recommends combining that with St John’s Wort, but I didn’t notice a substantive-enough difference adding it in, so I dropped it.

      If it starts to work and you feel more into life, try using your newfound energy to exercise. Your goal is a positive feedback loop of positivity.

      And if relevant, I suggest allowing your conception-of-self to change through this, and come to terms with the fact that you’re a different person than you maybe wanted to be.

  81. inconsistentidentity says:

    I was flabbergasted by the lack of a ‘creating new beings is better than improving the lives of existing beings’ option on the fifth question. I’m not sure I think that, but it’s definitely something that people could think.

  82. Zanzard says:

    So, there was a certain Brazilian company from the 1970’s or 1980’s that manufactured toilets. Its business was doing well in Brazil, so they decided to expand their business to the USA.

    They began selling their toilets in the USA, hoping to achieve good results. Instead, their sales were terrible, way worse than anything they predicted, and they shut down their American operations and stopped selling toilets there.

    What happened? Turns out that Brazilian toilets have a lower amount of space between the person’s bum and the water below. Therefore, if a person using the Brazilian toilet flushes it whilst still seated, there is a significant chance that the water below is going to touch the person’s bumcheeks during the flushing process.

    This is not a terrible issue for Brazilians, who for the most part wipe their asses whilst standing up, but it was a huge deal for americans, who apparently tend to wipe whilst sitting down and think it’s gross to have water touch their bums.

    This post’s first paragraph reminded me of this story.