Related to: Is That Your True Rejection?

"If you’re interested in being on the right side of disputes, you will refute your opponents’ arguments.  But if you’re interested in producing truth, you will fix your opponents’ arguments for them.  To win, you must fight not only the creature you encounter; you must fight the most horrible thing that can be constructed from its corpse."

   -- Black Belt Bayesian, via Rationality Quotes 13

Yesterday John Maxwell's post wondered how much the average person would do to save ten people from a ruthless tyrant. I remember asking some of my friends a vaguely related question as part of an investigation of the Trolley Problems:

You are a doctor in a small rural hospital. You have ten patients, each of whom is dying for the lack of a separate organ; that is, one person needs a heart transplant, another needs a lung transplant, another needs a kidney transplant, and so on. A traveller walks into the hospital, mentioning how he has no family and no one knows that he's there. All of his organs seem healthy. You realize that by killing this traveller and distributing his organs among your patients, you could save ten lives. Would this be moral or not?

I don't want to discuss the answer to this problem today. I want to discuss the answer one of my friends gave, because I think it illuminates a very interesting kind of defense mechanism that rationalists need to be watching for. My friend said:

It wouldn't be moral. After all, people often reject organs from random donors. The traveller would probably be a genetic mismatch for your patients, and the transplantees would have to spend the rest of their lives on immunosuppressants, only to die within a few years when the drugs failed.

On the one hand, I have to give my friend credit: his answer is biologically accurate, and beyond a doubt the technically correct answer to the question I asked. On the other hand, I don't have to give him very much credit: he completely missed the point and lost a valuable effort to examine the nature of morality.

So I asked him, "In the least convenient possible world, the one where everyone was genetically compatible with everyone else and this objection was invalid, what would you do?"

He mumbled something about counterfactuals and refused to answer. But I learned something very important from him, and that is to always ask this question of myself. Sometimes the least convenient possible world is the only place where I can figure out my true motivations, or which step to take next. I offer three examples:


1:  Pascal's Wager. Upon being presented with Pascal's Wager, one of the first things most atheists think of is this:

Perhaps God values intellectual integrity so highly that He is prepared to reward honest atheists, but will punish anyone who practices a religion he does not truly believe simply for personal gain. Or perhaps, as the Discordians claim, "Hell is reserved for people who believe in it, and the hottest levels of Hell are reserved for people who believe in it on the principle that they'll go there if they don't."

This is a good argument against Pascal's Wager, but it isn't the least convenient possible world. The least convenient possible world is the one where Omega, the completely trustworthy superintelligence who is always right, informs you that God definitely doesn't value intellectual integrity that much. In fact (Omega tells you) either God does not exist or the Catholics are right about absolutely everything.

Would you become a Catholic in this world? Or are you willing to admit that maybe your rejection of Pascal's Wager has less to do with a hypothesized pro-atheism God, and more to do with a belief that it's wrong to abandon your intellectual integrity on the off chance that a crazy deity is playing a perverted game of blind poker with your eternal soul?

2: The God-Shaped Hole. Christians claim there is one in every atheist, keeping him from spiritual fulfillment.

Some commenters on Raising the Sanity Waterline don't deny the existence of such a hole, if it is intepreted as a desire for purpose or connection to something greater than one's self. But, some commenters say, science and rationality can fill this hole even better than God can.

What luck! Evolution has by a wild coincidence created us with a big rationality-shaped hole in our brains! Good thing we happen to be rationalists, so we can fill this hole in the best possible way! I don't know - despite my sarcasm this may even be true. But in the least convenient possible world, Omega comes along and tells you that sorry, the hole is exactly God-shaped, and anyone without a religion will lead a less-than-optimally-happy life. Do you head down to the nearest church for a baptism? Or do you admit that even if believing something makes you happier, you still don't want to believe it unless it's true?

3: Extreme Altruism. John Maxwell mentions the utilitarian argument for donating almost everything to charity.

Some commenters object that many forms of charity, especially the classic "give to starving African orphans," are counterproductive, either because they enable dictators or thwart the free market. This is quite true.

But in the least convenient possible world, here comes Omega again and tells you that Charity X has been proven to do exactly what it claims: help the poor without any counterproductive effects. So is your real objection the corruption, or do you just not believe that you're morally obligated to give everything you own to starving Africans?


You may argue that this citing of convenient facts is at worst a venial sin. If you still get to the correct answer, and you do it by a correct method, what does it matter if this method isn't really the one that's convinced you personally?

One easy answer is that it saves you from embarrassment later. If some scientist does a study and finds that people really do have a god-shaped hole that can't be filled by anything else, no one can come up to you and say "Hey, didn't you say the reason you didn't convert to religion was because rationality filled the god-shaped hole better than God did? Well, I have some bad news for you..."

Another easy answer is that your real answer teaches you something about yourself. My friend may have successfully avoiding making a distasteful moral judgment, but he didn't learn anything about morality. My refusal to take the easy way out on the transplant question helped me develop the form of precedent-utilitarianism I use today.

But more than either of these, it matters because it seriously influences where you go next.

Say "I accept the argument that I need to donate almost all my money to poor African countries, but my only objection is that corrupt warlords might get it instead", and the obvious next step is to see if there's a poor African country without corrupt warlords (see: Ghana, Botswana, etc.) and donate almost all your money to them. Another acceptable answer would be to donate to another warlord-free charitable cause like the Singularity Institute.

If you just say "Nope, corrupt dictators might get it," you may go off and spend the money on a new TV. Which is fine, if a new TV is what you really want. But if you're the sort of person who would have been convinced by John Maxwell's argument, but you dismissed it by saying "Nope, corrupt dictators," then you've lost an opportunity to change your mind.

So I recommend: limit yourself to responses of the form "I completely reject the entire basis of your argument" or "I accept the basis of your argument, but it doesn't apply to the real world because of contingent fact X." If you just say "Yeah, well, contigent fact X!" and walk away, you've left yourself too much wiggle room.

In other words: always have a plan for what you would do in the least convenient possible world.

New to LessWrong?

New Comment
203 comments, sorted by Click to highlight new comments since: Today at 9:31 PM
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

I think a better way to frame this issue would be the following method.

  1. Present your philosophical thought-experiment.
  2. Ask your subject for their response and their justification.
  3. Ask your subject, what would need to change for them to change their belief?

For example, if I respond to your question of the solitary traveler with "You shouldn't do it because of biological concerns." Accept the answer and then ask, what would need to change in this situation for you to accept the killing of the traveler as moral?

I remember this method giving me deeper insight into the Happiness Box experiment.

Here is how the process works:

  1. There is a happiness box. Once you enter it, you will be completely happy through living in a virtual world. You will never leave the box. Would you enter it?
  2. Initial response. Yes, I would enter the box. Since my world is only made up of my perceptions of reality, there is no difference between the happiness box and the real world. Since I will be happier in the happiness box, I would enter.
  3. Reframing question. What would need to change so you would not enter the box.
  4. My response: Well, if I had children or people depending on me, I could no
... (read more)

I find a similar strategy useful when I am trying to argue my point to a stubborn friend. I ask them, "What would I have to prove in order for you to change your mind?" If they answer "nothing" you know they are probably not truth-seekers.

Namely, the point of reversal of your moral decision is that it helps to identify what this particular moral position is really about. There are many factors to every decision, so it might help to try varying each of them, and finding other conditions that compensate for the variation.

For example, you wouldn't enter the happiness box if you suspected that information about it giving the true happiness is flawed, that it's some kind of lie or misunderstanding (on anyone's part), of which the situation of leaving your family on the outside is a special case, and here is a new piece of information. Would you like your copy to enter the happiness box if you left behind your original self? Would you like a new child to be born within the happiness box? And so on.

This seems to nicely fix something which I felt was wrong in the "least convenient possible world" heuristic. The LCPW only serves to make us consider a possibility seriously. It may be too easy to come up with a LCPW. Asking what would change your mind helps us examine the decision boundary.
Great, David! I love it.
The happiness box is an interesting speculation, but it involves an assumption that, in my view, undermines it: "you will be completely happy." This is assuming that happiness has a maximum, and the best you can do is top up to that maximum. If that were true, then the happiness box might indeed be the peak of existence. But is it true?
Okay, well let's apply exactly the technique discussed above: If the hypothetical Omega tells you that they're is indeed a maximum value for happiness, and you will certainly be maximally happy inside the box: do you step into the box then? Note: I'm asking that in order to give another example of the technique in action. But still feel free to give a real answer if you'd choose to. Side you didn't answer the question one way or another, I can't apply the second technique here. I can't ask what would have to change in order for you to change your answer.
What if we ignore the VR question? Omega tells you that killing and eating your children will make you maximally happy. Should you do it? Omega can't tell you that doing X makes you maximally happy unless doing X actually makes you maximally happy. And a scenario where doing X actually makes you maximally happy may be a scenario where you are no longer human and don't have human preferences. Omega could, of course, also say "you are mistaken when you conclude that being maximally happy in this scenario is not a human preference". However, 1. This conclusion that that is not a human preference is being made by you, the reader, not just by the person in the scenario. It is not possible to stipulate that you, the reader, are wrong about your analysis of some scenario. 2. Even within the scenario, if someone is mistaken about something like this, it's a scenario where he can't trust his own reasoning abilities, so there's really nothing he can conclude about anything at all. (What if Omega tells you that you don't understand logic and that every use of logic you think you have done was either wrong or true only by coincidence?)
This would depend on my level of trust in Omega (why would I believe it? Because Omega said so. Why believe Omega? That depends on how much Omega has demonstrated near-omniscience and honesty). And in the absence of Omega telling me so, I'm rather skeptical of the idea.
For my part, it's difficult for me to imagine a set of observations I could make that would provide sufficient evidence to justify belief in many of the kinds of statements that get tossed around in these sorts of discussions. I generally just assume Omega adjusts my priors directly.

I'm not sure if I'm evading the spirit of the post, but it seems to me that the answer to the opening problem is this:

If you were willing to kill this man to save these ten others, then you should long ago have simply had all ten patients agree to a 1/10 game of Russian Roulette, with the proviso that the nine winners get the organs of the one loser.

While emphasizing that I don't want this post to turn into a discussion of trolley problems, I endorse that solution.

In the least convenient possible world, only the random traveler has a blood type compatible with all ten patients.

This is fair, because you're using the technique to redirect us back to the original morality issue. But i also don't think that MBlume was completely evading the question either. The question was about ethical principles, and his response does represent an exploration of ethical principles. MBlume suggests that it's more ethical to sacrifice one of the lives that was already in danger, than to sacrifice an uninvolved stranger. (remember, from a strict utilitarian view, both solutions leave one person dead, so this is definitely a different moral principle.) This technique is good for stopping people from evading the question. But some evasions are more appropriate than others.
I'd go with that he's the only one who has organs healthy enough to ensure the recipients survive.
MBlume knows this, he's just telling us what he was thinking.
2Said Achmiz11y
What if one or more of the patients don't agree to do this?
Then you let him die, and repeat the question with a 1/9 chance of death.
1Bruno Mailly6y
To me the logical answer is that it depends on how much value is attributed to "a" life vs respect of individual freedom/integrity. It is fairly reasonable : do no evil, do not instrumentalize people, especially if not involved ; because this is a very slippery slope. But it is unworkable to enter such a game of value accounting : Whose value system should be used ? Apple-and-orange value ? My practical answer meets yours : If one is ready to kill the stranger, one should have anticipated this and done something along those lines long ago, like kill a criminal or comatose.
The technical creativity of this solution reveals the limits of rationality. This is a solution only in a world of dice. But in a world of minds and psyches there are problems. The nine survivors have killed a man so they themselves can live. The "dice-argument" that it was voluntary and everyone had an equal chance of dying or surviving is irelevant. The survivors pulled the trigger on the victim in order that they could survive. That is their legacy, that is their gulit and only a "self-deceiving rationalist" would be able to suppress this guilt by rejoicing in the numbers.

Throwing a die is a way of avoiding bias in choosing a person to kill. If you choose a person to kill personally, you run a risk of doing in in an unfair fashion, and thus being guilty in making an unfair choice. People value fairness. Using dice frees you of this responsibility, unless there is a predictably better option. You are alleviating additional technical moral issues involved in killing a person. This issue is separate from deciding whether to kill a person at all, although the reduction in moral cost of killing a person achieved by using the fair roulette technology may figure in the original decision.

But as a doctor, probably you will have to choose non-randomly, if you want to stand by your utilitarian viewpoint, since killing different people might have different probabilities of success. Assuming the lest convenient possible world hypothesis, you can't make your own life easier by assuming each one's sacrifice is as likely to go well. So in the end you will have to assume that one patients sacrifice will be the "best", and will have to decide if you kill them, thus reverting to the original problem.

There are real life examples where reality has turned out to be the "least convenient of possible worlds". I have spent many hours arguing with people who insist that there are no significant gender differences (beyond the obvious), and are convinced that to assert otherwise is morally reprehensible.

They have spent so long arguing that such differences do not exist, and this is the reason that sexism is wrong, that their morality just can't cope with a world in which this turns out not to be true. There are many similar politically charged issues - Pinker discusses quite a few in the Blank Slate - where people aren't wiling to listen to arguments about factual issues because they believe they have moral consequences.

The problem, of course - and I realise this is the main point of this post - is that if your morality is contingent on empirical issues where you might turn out to be wrong, you have to accept the consequences. If you believe that sexism is wrong because there are no heritable gender differences, you have to be willing to accept that if these differences do turn out to exist then you'll say sexism is ok.

This is probably a test you should apply to all of your moral beliefs - if it just so happens that I'm wrong about the factual issue on which I'm basing my belief is wrong, will really I be willing to change my mind?

That raises an interesting question: is it possible to base a moral code only on what's true in all possible worlds that contain me?
To do that would require that "all possible worlds that contain me" be a coherent concept. What does it mean, to identify as "me" some agent in a world very different from our own?
I think that it is not. All possible worlds include worlds where every tuesday the first person you meet in a crowded place just happens to attack you. That would lead to a personal moral code of stabbing the first person you meet on tuesday. I think we can only have a moral code that works on most worlds at best
You could have a personal moral code of stabbing anyone who you're 90% certain would otherwise attack you. In a universe where the first person you meet on Tuesday always tries to kill you, you would quickly start stabbing them first. In other worlds, you would not.
That doesn't follow from your logic. There could be multiple functions of maximal expectd utility. Or more fundamentally, how you sum over possible words reflects your prior anthropic biases (which worlds you think are most likely), which is sadly a completely arbitrary choice.
I took "all possible worlds that contain me" to mean all worlds where history went the same until my birth. Any world where significant things went differently would have led a different sperm to create a different person than him. That is, they should be reasonably similar but can still include diverse outcomes from for example nuclear war where Pr0methean is living in post-apocalyptic fallout to a USSR-US alliance leading to a fascist authoritarian government in your country to choice. I did in fact assume that worlds more similar to our current one would make up the majority [or at least the plurality] in that case. Was I wrong to assume that? Edit: thinking about it now, the plurality was post-hoc rationalisation, so ignore it. On a side note, how do I do strikethrough text?
Retract -- circle with an line through it.
What do you mean by circle with a line through it? Is that some sort of code for what buttons to press?
There should be a button with that appearance in the lower right-hand corner of your comments, which brings up a tooltip labeled "retract" when you mouse over it. Using it will strikethrough the entire text of your post, which 'round these parts is shorthand for "I, the author, no longer endorse this comment". Using it for a second time will delete your post, unless there are responses to it. There isn't any way to strikethrough portions of a post with LW's markup. Or at least I wasn't able to find one the last time I looked into this. The usual Markdown syntax is disabled here, probably to reserve the look for the retract option.
The causality is unlikely. There was never strikethrough syntax here and the retract option was not conceived until years after the creation of the forum (and syntax choices).
Ah, thank you. I hadn't noticed that

One way to train this: in my number theory class, there was a type of problem called a PODASIP. This stood for Prove Or Disprove And Salvage If Possible. The instructor would give us a theorem to prove, without telling us if it was true or false. If it was true, we were to prove it. If it was false, then we had to disprove it and then come up with the "most general" theorem similar to it (e.g. prove it for Zp after coming up with a counterexample in Zm).

This trained us to be on the lookout for problems with the theorem, but then seeing the "least convenient possible world" in which it was true.


I voted up on your post, Yvain, as you've presented some really good ideas here. Although it may seem like I'm totally missing your point by my response to your 3 scenarios, I assure you that I am well aware that my responses are of the "dodging the question" type which you are advocating against. I simply cannot resist to explore these 3 scenarios on their own.

Pascal's Wager

In all 3 scenarios, I would ask Omega further questions. But these being "least convenient world" scenarios, I suspect it'd be all "Sorry, can't answer that" and then fly away. And I'd call it a big jerk.

For Pascal Wager's specific scenario, I'd probably ask Omega "Really? Either God doesn't exist or everything the Catholics say is correct? Even the self-contradicting stuff?" And of course, he'd decline to answer and fly away.

So then I'd be stuck trying to decide whether God doesn't exist, or logic is incorrect (i.e. reality can be logically self inconsistent). I'm tempted to adopt Catholicism (for the same reason I would one-box on Newcomb: I want the rewards), but I'm not sure how my brain could handle a non-logical reality. So I really don't know what would happen ... (read more)

The point is that in the least convenient world for you, Omega would say whatever it is that you would need to hear to not slip away. I don't know what that is. Nobody but you do. If it is about eternal damnation for you, then you've hopefully found your holy grail, and as some other poster pointed out, why this is the holy grail for you can be quite interesting to dig into as well. The point raised, as I see it, is just to make your stance on Pascal's wager contend against the strongest possible ideas.
The least convenient world is one where Omega answers his objections. The least convenient possible world is one where Omega answers his objections in a way that's actually possible. And it may not be possible for Omega to answer some objections.
This is a very good point, and I believe I'll point it out to my rather fundamentalist sibling when next we talk about this: if I really, truly believed that every non-Christian was doomed to eternal damnation, you can bet I'd be an evangelist! Extreme Altruism I definitely don't value those billions of lives more than my own happiness, or more than the happiness of those I know and love. However, I would seriously consider giving all of my wealth if Omega assured me that me and mine would be able to continue to be reasonably happy after doing so, even if it meant severe lifestyle changes.
If I really, truly believed that every non-Christian was doomed to eternal damnation, I'd donate to a charity that distributes condoms to people in Africa. The key here is to minimize the number of non-Christians, not to make more people Christian.

Let's try something different.

  • Puts on the reviewer's hat.

The Yvain's post presented a new method for dealing with the stopsign problem in reasoning about questions of morality. The stopsign problem consists in following an invalid excuse to avoid thinking about the issue at hand, instead of doing something constructive about resolving the issue.

The method presented by Yvain consists in putting in place the universal countermeasure against the stopsign excuses: whenever a stopsign comes up, you move the discussed moral issue to a different, hypothetical setting, where the stopsign no longer applies. The only valid excuse in this setting is that you shouldn't do something, which also resolves the moral question.

However, the moral questions should be concerned with reality, not with fantasy. Whenever a hypothetical setting is brought in the discussion of morality, it should be understood as a theoretical device for reasoning about the underlying moral judgment applicable to the real world. There is a danger in fallaciously generalizing the moral conclusion from fictional evidence, both because there might be factors in the fictional setting that change your decision and which you ... (read more)

I do agree. I think in many ways reality already is "the least convenient possible world" and the clearsightedness of thought experiments doesn't match the muddiness of the world.

One difficulty with the least convenient possible world is where that least convenience is a significant change in the makeup of the human brain. For example, I don't trust myself to make a decision about killing a traveler with sufficient moral abstraction from the day-to-day concerns of being a human. I don't trust what I would become if I did kill a human. Or, if that's insufficient, fill in a lack of trust in the decisionmaking in general for the moment. (Another example would be the ability to trust Omega in his responses)

Because once that's a significant issue in the subject , then the least convenient possible world you're asking me to imagine doesn't include me -- it includes some variant of me whose reactions I can predict, but not really access. Porting them back to me is also nontrivial.

It is an interesting thought experiment, though.

So I asked him, "In the least convenient possible world, the one where everyone was genetically compatible with everyone else and this objection was invalid, what would you do?"

Obviously, you wait for one of the sick patients to die, and use that person's organs to save the others, letting the healthy traveler go on his way. ;)

But that isn't the least convenient possible world - the least convenient one is actually the one in which the traveler is compatible with all the sick people, but the sick people are not compatible with each other.

Actually, you don't even need to add that additional complexity to make the world sufficiently inconvenient.

If the rest of the patients are sufficiently sick, their organs may not really be suitable for use as transplants, right?

There's another benefit: you remove a motivation to lie to yourself. If you think that a contingent fact will get you out of a hard choice, you might believe it. But you probably won't if it doesn't get you out of the hard choice anyway.

On the other hand, if you think that a contingent fact will get you out of a hard choice, perhaps you will be more likely to find legitimate contingent facts.

Would you become a Catholic in this world? Or are you willing to admit that maybe your rejection of Pascal's Wager has less to do with a hypothesized pro-atheism God, and more to do with a belief that it's wrong to abandon your intellectual integrity on the off chance that a crazy deity is playing a perverted game of blind poker with your eternal soul?

I don't think I would be able to bring myself to worship honestly a God who bestowed upon us the ability to reason and then rewarded us for not using it.

Would you want to, if you could? If so, given the stakes, you should try damn hard to make yourself able to.
4Robert Miles13y
I don't follow your reasoning. Because God made us able to do a particular thing, we shouldn't be rewarded for choosing not to do that thing? A quick word substitution illustrates my issue: "I don't think I would be able to bring myself to worship honestly a God who bestowed upon us the ability to murder and then rewarded us for not using it."
I certainly wouldn't like such a God. He'd be better than a God who bestowed upon us the ability to murder and then rewarded us for using it, but what kind of God would bestow upon us the ability to murder?
My statement does not generalize in that way, and was not intended to do so.
It does. It just doesn't if you accept the premise that intelligence is, in and of itself, good (and murder is not). I accept that premise, of course, and your assertion that it was not intended to be generalized as such. But still, within the framework of this hipothetical world, that simply cannot be true. In fact, it cannot be relevant. It is not a moral question at all; more of an utility vs principles thing. In the original Pascal's Wager, as I recall, false (outwardly) adoration does score you points. I seem to recall him saying that "at least you wouldn't be corrupting the youths" and "you may become convinced by habit", at least. So yeah, I would try my best (and probably fail) to aquire that reward, if it was shown to be worth it. On the other hand, on such a world, it probably would not. Heaven for the weather, Hell for the company, et al.

The problem with the 'god shaped hole' situation (and questions of happiness in general) is that if something doesn't make you happy NOW, it becomes very difficult to believe that it will make you happy LATER.

For example, say some Soma-drug was invented that, once taken, would make you blissfully happy for the rest of your life. Would you take it? Our immediate reaction is to say 'no', probably because we don't like the idea of 'fake', chemically-induced happiness. In other words, because the idea doesn't make us happy now, we don't really believe it will ... (read more)

7Swimmer963 (Miranda Dixon-Luinenburg) 13y
I try my best to value other peoples' happiness equal to my own. If taking a happiness-inducing pill was likely to make me a kinder, more generous, more productive person, I would choose to take it (with some misgivings related to it seeming like 'cheating' and 'not good for character-building') but if it were to make me less kind/generous/productive, I would have much stronger misgivings.
I would definitely take the Soma, and don't see why anyone wouldn't. Odd, the differences between what people find acceptable. Is anyone else with me in desiring chemically-induced happiness as much as any other? (Well, all happiness is chemically-induced, when you get right down to it, so I assume there are no qualitative differences.)
I wouldn't take it. I desire to help others, and it gives me pleasure to do so, it makes me suffer to harm others, and I desire not to do so. Being perpetually in a state of extreme pleasure would make this pleasure/suffering irrelevant, and might lead me to behave less in line with my desires. So, being perpetually in a state of extreme pleasure seems like a bad idea to me.
0Swimmer963 (Miranda Dixon-Luinenburg) 13y
I agree with you completely. I can understand why others might not agree with me, but for me, pleasure isn't so much a goal as a result of accomplishing my goals.
2Peter Wildeford13y
I'm reminded of Yudkowsky's Not For the Sake of Happiness Alone.
I think one of the points underrepresented in these "Not For the Sake of XXX Alone" posts is how people would respond to a least convenient world possible in which they would be forced to make sharp trade-offs between competing values. For instance, I value diversity, a kind of narrative depth to raw experiences. But if I had to choose either sustainable, chemically induced unsophisticated pleasure or else diverse pain and misery with narrative depth, I'd almost certainly choose the pleasure. This is relevant to FAI and CEV, I think. If the success probability of simple, pleasure-generating FAI is higher than more sophisticated (and difficult) "Not For the Sake of XXX Alone"-respecting FAI, it might be better opting for the pleasure-generating version.
Agreed. I also think people tend to underestimate the goodness of pure bliss: I have experienced such a state, and I'm here to tell you, the concerns about XXX become very much more minor than you would expect. They don't disappear - if you like painting, you'll still want to paint - but you suddenly understand how minor the pleasure painting gives you really is, in comparison. Or at least that's how I felt, anyway.
He makes good points, but note that there's nothing saying you couldn't take Soma and participate in the joy of scientific discovery (or whatever).
0Peter Wildeford13y
The argument wasn't that you need the joy of scientific discovery; it was that scientific discovery is important to us for reasons entirely apart from joy. You would never want a Soma substitute for scientific discovery, because that wouldn't involve... you know... actual scientific discovery. Additionally, another different take on this is Yvain's Are Wireheads Happy?.
This is just wire-heading isn't it? At least, that is what you should search for if you want to hear what people on this site tend to think about this sort of idea. I am not certain of my own view of it. I tend to think I'd wire-head at first, but then some implications I find on more reflection make me unsure.
Same here. That is, I know I'd wirehead - I don't see any bothersome implications with that idea alone. However, if you add in something like "once you wirehead you are immobile and cannot do anything else", then I become more unsure.
It does not matter if you are immobilized. Once you are wire-heading there is no reason you would ever stop since you've already got peak pleasure/joy. I think this effectively immobilizes you. There is no problem that could come to you that wouldn't be best solved by more wire-heading, except for a threat to the wire-heading itself.
I think you're simply assuming that we're motivated primarily by happiness in that case. Valuing the truth doesn't suddenly make me happy when someone announces to me, and I verify, that my entire family has been eaten by wombats. If I didn't value the truth at all, I might be able to ignore reality and persist in my erroneous belief that my family is alive and wombats are as cute and cuddly as I have always believed. But I don't try to do that, and I don't regret my decision or any inability to maintain erroneous beliefs. A soma drug offends my sensibilities on some level. It violates my moral value of "don't mess around with my brain except through standard sensory experiences or with my explicit and informed consent" (and no brainwashing: no concerted efforts to modify my opinions or attitudes or behaviors except through normal human interactions, like arguing and talking and long walks on the beach). I value at least some of these moral sensibilities higher than my current or future happiness. That's why I would choose against the soma. Not because I doubt its efficacy.

I like the phrase "precedent utilitarianism". It sounds to utilitarians like you're joining their camp, while actually pointing out that you're taking a long-term view of utility, which they usually refuse to do. The important ingredient is paying attention to incentives, which is really the rational response to most questions about morality. Many choices which seem "fairer", "more just", or whose alternatives provoke a disgust response don't take the long-term view into account. If we go around sacrificing every lonely s... (read more)

Actually, we would all be more safe, because we'd be in less danger from organ failure. We are each more likely to be one of the "others nearby" than the "lonely stranger".
That would be true if they were hunting people down. As stated, people would become more resistant to going to hospitals, which would cause problems that way.
This is an exact instance of the point of the post. It is important to assume they are hunting people down, because that's the LCPW and the fact that this trolley problem incorporates using someone who shows up at the hospital is entirely an unnecessary contingent fact.
On what basis would you say it's the case that utilitarians usually refuse to take a long-term view of utility?
When I've argued with people who called themselves utilitarian, they seemed to want to make trade-offs among immediately visible options. I'm not going to try to argue that I have population statistics, or know what the "proper" definition of a utilitarian is. Do you believe that some other terminology or behavior better characterizes those called "utilitarians"?
Well, in my experience people who self identify as utilitarians don't appear to be any more shortsighted in terms of real life moral quandaries than people who don't so self identify. I don't think it's the case that utilitarians tend to be shortsighted, just that people in general tend to be; if non-utilitarians tend to choose a less shortsighted action in a constructed moral dilemma, it's not usually due to consciously taking a long view. When I was in college, a professional philosopher once visited and gave a seminar, where she raised the traveler-at-a-hospital scenario as an argument against utilitarianism (simply on the basis that killing the traveler defies our moral intuitions.) I responded that realistically, given human nature, if doctors tended to do this, then because people aren't effective risk assessers, people would tend to avoid hospitals for fear of being harvested, to the point that the practice would probably be doing more harm than good. She had never heard or thought of this argument before, and found it a compelling reason not to harvest the traveler from a utilitarian point of view. So as a non utilitarian, it doesn't seem that she was any more likely to look at questions of utility from a long view, she was just more willing to let moral intuitions control her decision, which sometimes has the same effect.
And that is an advantage of traditional moral systems -- because they have been around for so long, they have had opportunities to be tried and tested in various ways. It won't give adherents a long-term view, but it can be a similar effect. Think of it as, "I don't have to think out the consequences of this because other people have thought through similar problems over a thousand years, and came up with a rule that says I should do X." One would be foolish to totally disregard traditional morality simply because of it's occasional clash with the modern world. It would be like disregarding a "traditional" gene made by "stupid blind arbitrary evolution" because we think we have a better one made by a smarter system -- it might be a good idea to compare anyways.
I tend to agree, but it depends on how something was tested. In "Darwinian Agriculture", I argue that testing by ability to persist is weaker than testing by competition against alternatives. Trees compete against each other, but forests don't. Societies often compete and their moral systems probably affect competitive success, but things are complicated by migration between societies, population growth (moral systems that work for bands of relatives may not work as well for modern nations), technological change (cooking pork), etc.
That would make a great movie! Lonely Stranger Jason Statham wakes up and realises all his family and friends have been killed by a tornado while he survives through luck and general masculine superiority. Beset upon on all sides by scalpel and tranquiliser wielding doctors he must constantly slaughter all the nearby sick people just to keep himself alive. Meanwhile, a sexy young biologist has been captured by a militant sect of religious Fundamentalists. Will Statham be able to break the imprisoned costar out in time to reveal her secret human organ cloning technology or will civilisation as we know it be destroyed by utilitarianism gone wrong?
What do you think of the definition of "Precedent Utilitarianism" used in the philosophy course module archived at ?

I would act differently in the least convenient world than I do in the world that I do live in.

Very good point, and crystalizes some of my thinking on some of the discussion on the tyrant/charity thing.

As far as the specific problems you posed...

For your souped up Pascal's Wager, I admit that one gives me pause. Taking into account the fact that Omega singled out one out of the space of all possible religions, etc etc... Well, the answer isn't obvious to me right now. This flavor would seem to not admit to any of the usual basic refutations of the wager. I think under these circumstances, assuming Omega wasn't open to answering any further question... (read more)

The souped up Pascal's Wager seems like the thousand door version of Monty Hall.

I am trying to imagine the least convenient possible world (LCPW) for the LCPW method.

Perhaps it is the world in which there is precisely one possible world. All 'possible' worlds turn out to be impossible on closer scrutiny. Omega reveals that talking about a counterfactual possible world is as incoherent as talking about a square triangle. There is exactly one way to have a world with anyone in it whatsoever, and we're in it.

Yes! I can't believe I don't see this repeated in one form or another more often. Fallacies are a bit like prions in that they tend to force a cascade of fallacies to derive from them, and one of my favorite debate tactics is the thought experiment, "Let's assume your entire premise is true. How might this contradict your position?"

Usually the list is longer than my own arguments.

The least convenient world is one where there's no traveler and the doctor debates whether to harvest organs from another villager. I figure that if it's okay to kill the traveler for organs, then it should be okay to kill a villager. Similarly, if it's against general principle to kill a villager for organs, then it shouldn't be okay to kill the traveler. Perhaps someone can come up with a clever argument why the life of a villager is worth intrinsically more than the life of the traveler, but let's keep things simple for now.

So, let us suppose that N sic... (read more)

The perverse incentive to become alcoholic or obese can be easily countered with a simple rule - a person chosen in the lottery is sacrificed no matter what, even if he doesn't actually have viable organs. To be truly effective, the system needs to consider the fact that some people are exceptional and can contribute to saving lives much more effectively than by scrapping and harvesting for spare parts. Hence, there should actually be an offer to anyone who loses the lottery, either pay $X or be harvested. A further optimization is a monetary compensation to (the inheritors of) people who are selected, proportional to the value of the harvested organs. This reduces the overall individual risk, and gives people a reason to stay healthy even more than normally. All of this is in the LCPW, of course. In the real world, I'm not sure there is enough demand for organs that the system would be effective in scale. Also, note that a key piece of the original dilemma is that the traveler has no family - in this case, the cost of sacrifice is trivial compared to someone who has people that care about him.
I think China used to have a similar system, except that instead of lottery they just picked prisoners from the death row.
2Marion Z.1y
That seems entirely reasonable, insofar as the death penalty is at all. I don't think we should be going around executing people, but if we're going to then we might as well save a few lives by doing it
We have two such systems today, except 1. We call it "taxes". 2. People die on an overall statistical basis (because people who are poorer die sooner, and paying taxes makes them poorer) rather than by loss of organs so it is hard to point to an individual death caused by taking things from one person to give them to someone who is more needy. For the second system, 1. We call it "a justice system". 2. The harm to innocent people is again statistical--because all justice systems are imperfect, they will convict X innocent people, and we've decided that harming X innocent people is an acceptable price to pay to convict more guilty people and protect the populace from criminals.

This might be better placed somewhere else, but I just thought I'd comment on Pascal's Wager here. To me both the convenient and inconvenient resolutions of Pascal's Wager given above are quite unsatisfactory.

To me, the resolution of this wager comes from the concept of sets of measure zero. The set of possible realities in which belief in any given God is infinitely beneficial is an infinite set, but it is nonetheless like Cantor Dust in the space of possible explanations of reality. The existence of sets of measure zero explains why it is reasonable to a... (read more)

Although I understand and appreciate your approach the particular examples do not represent particularly good ones:

1: Pascal's Wager:

For an atheist the least convenient possible world is one where testable, reproducible scientific evidence strongly suggests the existence of some "super-natural" (clearly no-longer super-natural) being that we might ascribe the moniker of God to. In such a world any "principled atheist" would believe what the verifiable scientific evidence support as probably true. "Atheists" who did not do th... (read more)

"I believe that God’s existence or non-existence can not be rigorously proven."

Cannot be proven by us, with our limits on detection, or cannot be proven in principle?

Because if it's the latter, you're saying that the concept of 'God' has no meaning.

Formalize this a bit: "I believe that X’s existence or non-existence can not be rigorously proven." Where X is of the set of beings imagined by or could be imagined by humans, e.g.: God, Gnomes, Zeus, Wotan, Vishnu, unicorns, leprechauns, Flying Spaghetti Monster, etc. Why is any one of the statements that result from such substitutions more meaningful than any other?
I think just because something cannot be proven (even in principle) does not necessarily imply that it is not true, let alone has no meaning. See Godel's Incompleteness Theorem, for example.
It is the latter (I’m an agnostic). However, I don’t see why the concept has no meaning. Would you say that axioms in math are meaningless?
It's possible to decide which axioms are in effect from the inside of a sufficiently complex mathematical system (such as this universe), however. For that matter, it would be possible to deduce the existence of a god, too; you just have to die. Granted, there are some issues with this, but nobody said deducing the axiom had to be convenient.
"It's possible to decide which axioms are in effect from the inside of a sufficiently complex mathematical system (such as this universe), however." I don't think I understand what you mean. "For that matter, it would be possible to deduce the existence of a god, too; you just have to die." When you meet a god, how can you be sure it's not a hallucination?
Assuming the entity in question is cooperative, try this: Ask it if P=NP is true, and for a proof for its answer to that in a form that you can easily understand. There's three possible outcomes: * It doesn't comply. Time to get suspicious about its claims to godhood. * It hands you a correct proof, beautifully elegant and easy to grasp. * It hands you a lump of nonsense, which your mind is too damaged to distinguish from a proof. If you get something that appears like an elegant proof, memorize it and recheck it every now and then. If your mind is sufficiently malfunctioning that it can't distinguish an elegant proof for P=NP from something that isn't, you may not be able to notice that from inside. There's still a chance whatever is afflicting you will get better over time; hence, do periodic rechecks, and pay particular attention to any nagging doubts about the proof you get while performing those. In the meantime, interpret the fact that you've gotten an apparent proof as significant evidence for the entity in question being real and very powerful.
Or: it says "This is undecidable in Zermelo-Fraenkel set theory plus the axiom of choice". In the case of P=NP, I might believe it I would not believe a purported god if it said all 9 remaining Clay math prize problems are undecidable.
If it really is undecidable, God must be able to prove that. However, I think an easier way to establish whether something is just your hallucination or a real (divine) being is asking them about something you couldn't possibly know about and then check if it's true.
8Eliezer Yudkowsky15y
It says "There is no elegant proof". Next?
Ask again, with another famously unsolved math problem. Repeat until it stops saying that or you run out of problems you know. If you ran out, ask the entity to choose a famous math problem not yet solved by human mathematicians, explain the problem to you, and then give you the solution including an elegant proof. Next time you have internet access, check whether the problem in question is indeed famous and doesn't have a published solution. If the entity says "there are no famous unsolved math problems with elegant proofs", I would consider that significant empirical evidence that it isn't what it claims to be.
Depending on your definition of "elegant", there are probably no famous unsolved math problems with elegant proofs. For example, I would be surprised if any (current) famous unsolved math problems have proofs that could easily be understood by a lay audience.
It could give a formally checkable proof, that is far from being elegant, but your own simple proof checkers that you understand well can plough through a billion steps and verify the result.
"Would you say that axioms in math are meaningless?" They distinguish one hypothetical world from another. Furthermore, some of them can be empirically tested. At present, Euclidean geometry seems to be false and Riemannian to be true, and the only difference is a single axiom.
I think the words "true" and "false" have some connotation that you might not want to imply? Perhaps it would clearer to phrase this as "At present, it seems like the geometry of our universe is not Euclidean and that the geometry of our universe is Riemannian.".
They distinguish one hypothetical world from another. It's a subtle distinction, but I think it's more accurate and useful to say that the axioms define a mathematical universe, and that a mathematical universe cannot be true or false but only a better or poorer model of the physical universe.
Euclidean geometry isn't a theory about the world, and therefore cannot be falsified by evidence from the world. The primitives (e.g. "line" and "point") do not have unambiguous referents in the world. You can associate real-world things (e.g. patterns of graphite, or wooden rods) to those primitives, and to the extent that they satisfy the axioms, they will also satisfy the conclusions. Math is not physics.
"Math is not physics." It's made out of physics. I think perhaps you mean that math isn't about physics. To the degree that axioms aren't being used to talk about potential worlds, I would say that they're meaningless.
Riemannian geometry is not an axiomatic geometry in the same way that Euclidean geometry is, so it is not true that "the only difference is a single axiom." I think you are thinking of hyperbolic geometry. In any case, the geometry of spacetime according to the theory of general relativity is not any of these geometries, but it is instead a Lorentzian geometry. (I say "a" because the words "Riemannian" and "Lorentzian" both refer to classes of geometries rather than a single geometry -- for example, Euclidean geometry and hyperbolic geometry are both examples of Riemannian geometries.)
First i've heard of this, super interesting. Hmm. So what is the correct way to highlight the differences while still maintaining the historical angle? Continue w/ Riemannian geometry? Or just say what you have said, Lorentzian.
Special relativity is good enough for most purposes, which means that (a time slice of) the real universe is very nearly Euclidean. So if you are going to explain the geometry of the universe to someone, you might as well just say "very nearly Euclidean, except near objects with very high gravity such as stars and black holes". I don't think it's helpful to compare with Euclid's postulates, they reflect a very different way of thinking about geometry than modern differential geometry.
"They distinguish one hypothetical world from another." Just like different religions. "Furthermore, some of them can be empirically tested. " Empirical tests do not prove a proposition, but increase the odds of its being correct (just like "miracles" would raise the odds in favor of religion).

0: Should we kill the miraculously-compatible traveler and distribute his organs?

My answer is based on a principle that I'm surprised no one else seems to use (then again, I rarely listen to answers to the Fat Man/Train problem): ask the f**king traveler!

Explain to the traveler that he has the opportunity to save ten lives at the cost of his own. First they'll take a kidney and a lung, then he'll get some time to say goodbye to his loved ones while he gets to see the two people with the donated organs recover... and then when he's ready they'll take the re... (read more)

2Marion Z.1y
Only replying to a tiny slice of your post here, but the original (weak) Pascal's wager argument actually does say you should pretend to believe even if you secretly don't believe, for various fuzzy reasons such as societal influence, and that maybe God will see that you were trying, and that sheer repetition might make you believe a little bit eventually

My answers:

1.No, because their belief doesn't make any sense. It even has logical contradictions, which makes it "super impossible", meaning there's no possible world where it could be true (the omnipotence paradox proves that omnipotence is logically inconsistent; a god which is nearly omnipotent, nearly omniscient and nearly omnibenevolent wouldn't allow suffering, which, undoubtably, exists; "God wants to allow free will" isn't a valid defence, since there's a lot of suffering that isn't caused by other ... (read more)

either God does not exist or the Catholics are right about absolutely everything.

Then I would definitely and swiftly become an atheist, and I maintain that this is by far the most rational choice for everybody else as well. My prior belief in God not existing is relatively high (let's say 50/50), but my prior belief in all of Catholicism being the absolute truth is pretty much nil. And if you're using anything vaguely resembling consistent priors, it has to near-nil for you too, because the beliefs of Catholicism are just so incredibly specific. They na... (read more)

I find this method to be intellectually dangerous.

We do not live in the LCPW, and constantly considering ethical problems as if we do is a mind-killer. It trains the brain to stop looking for creative solutions to intractable real world problems and instead focus on rigid abstract solutions to conceptual problems.

I agree that there is a small modicum of value to considering the LCPW. Just like there's a small modicum of value to eating a pound of butter for dinner. It's just, there are a lot better ways to spend ones time. The proper response to "We... (read more)

I don't think I could disagree more. The point of ethical thought experiments like the sick villager problem is not to present a practical scenario to be solved; it's to illuminate seeming contradictions in our values. Yes, a lot of them have some holes -- where did the utility monster come from? Are there more of them with different preferences? Is it possible to make it happy by feeding it Tofumans? -- but spending a lot of time plugging them only distracts from the exercise. The reason they appear to be extreme corner cases is that only there does the contradiction show up without complications; but that doesn't mean they're not worth addressing on their own terms, the better to come up with a set of principles -- not a narrow heuristic -- that could be applied in real life. The LCPW, therefore, is a gentle reminder to think in those terms. Training yourself to think creatively has value, of course; if you're actually faced with sick villagers that you could save by chopping up healthy patients for their organs, you should by all means consider alternative solutions and have trained yourself for doing so. But this thought experiment isn't aimed at a rural surgeon with an itchy bonesaw; it's aimed at philosophers of ethics, or at armchair hobbyists of same. If you are such a person and you're faced with something like that as a hypothetical, and you can't prove that not only that hypothetical but no analogous situation will ever come up, engage with the apparent ethical contradiction or show that there isn't one; don't poke holes in the practical aspects of the hypothetical and congratulate yourself for it. That's like a physicist saying "well, I can't actually ride a beam of light, so obviously it's not worth thinking about". Though you could decide the whole genre isn't worth your time. That's fine too. Philosophy of ethics is a little abstract for most people's taste, including mine.
That's fair. I understand the value: it exposes the weakness of using overly rigid heuristics by presenting a situation where those heuristics feel wrong. And I agree that it's an evasion to nitpick the thought experiment in an attempt to avoid having to face the contradiction of your poorly-formed heuristics. My standard response to thought-experiment questions is: "I would do everything possible to have my cake and eat it too." In many cases, that response satisfies whoever asked the question. Immediately defaulting to the LCPW is putting words into the other person's mouth by assuming they wouldn't be satisfied with that ethical approach. Making the LCPW something "normal" seriously underestimates how world-bendingly different it would be from reality. If we truly lived in the LCPW, in most cases it would be such a different reality from the one we exist in that it would require a completely different set of ethics, and I just haven't really thought hard enough about it to generate a new system of ethics for each tailor-made LCPW. Incidentally I don't have a problem with the LCPW when it's actually realistic, as is the case with the "charity" example.
I agree that insisting on assuming the LCPW is a lousy strategic approach to most real-world situations, which (as you say) don't normally occur in the LCPW. And I agree that assuming it as an analytical stance towards hypothetical situations increases the chance that I'll adopt it as a strategic approach to a real-world situation, and therefore has some non-zero cost. That said, I would also say that a one-sided motivated "exploration" of a situation which happens to align with our pre-existing biases about that situation has some non-zero cost, for basically the same reasons. The OP starts out by objecting to that sort of motivated cognition, and proposes the LCPW strategy as a way of countering it. You object to the LCPW strategy, but seem to disregard the initial problem of motivated cognition completely. I suspect that in most cases, the impulse to challenge the assumptions of hypothetical situations does more harm than good, since motivated cognition is pretty pervasive among humans. But, sure, in the rare case of a thinker who really does "explore assumptions about a difficult problem" rather than simply evade them, I agree that it's an exercise that does more harm than good. If you are such a thinker and primarily engage with such thinkers, that's awesome; whatever you're doing, keep it up!


Do you have a blog or home page with more material you've written? Failing that, is there another site (apart from OB) with contributions from you that might be interesting to LW readers?

2Scott Alexander15y
Thanks for your interest. My blog is of no interest to anyone but my immediate personal friends, but I am working on a website. I'll let you know when it's up.
Hey Yvain. I found your blog a little while ago (I think it was from an interesting comment on Patri's LiveJournal, or maybe he linked to you). I disagree that your blog isn't interesting to people that aren't immediate friends (for example, I found your arguments about boycotts and children's rights to be interesting and persuasive). I respect that you seem to not want to link to it here, so I won't. But I urge you to change your mind!
Ha, this was just enough information for my google-fu to finally succeed. Yvain, I have a feeling that between your articles here, your travels through Outer Mongolia, and your apparent all-around awesomosity, EY has some stiff competition for cult leader.
0Scott Alexander15y
Thank you, Michael, for not linking to it here, and thank you, Badger, for the kind words. Although I'm not going to accept any comparisons to EY until I've come up with and implemented at least one feasible plan to save the world.

with regards to the third question: what if I believe that any resources given simply allow the population to expand and hence cause more suffering than letting people die?

If you don't really believe that, and it's just your excuse for not giving away lots of money, you should say loud and clear "I don't believe I'm morally obligated to reduce suffering if it inconveniences me too much." And then you've learned something useful about yourself.

But if you do really believe that, and you otherwise accept John's argument, you should say explicitly, "I accept I'm morally obligated to reduce suffering as much as possible, even at the cost of great inconvenience to myself. However, I am worried because of the contingent fact that giving people more resources will lead to more population, causing more suffering."

And if you really do believe that and think it through, you'll end up spending almost all your income on condoms for third world countries.

Is this not just an alternative way of describing a red herring argument? If not, I would be interested to see what nuance I'm missing.

I find this classically in the abortion discussion. Pro-abortionists will bring up valid-at-face-value concerns regarding rape and incest. But if you grant that victims of rape/incest can retain full access to abortions, the pro-abortionist will not suddenly agree with criminalisation of abortion in the non-rape/incest group. Why? Because the rape/incest point was a red herring argument

This is a good argument against Pascal's Wager, but it isn't the least convenient possible world. The least convenient possible world is the one where Omega, the completely trustworthy superintelligence who is always right, informs you that God definitely doesn't value intellectual integrity that much. In fact (Omega tells you) either God does not exist or the Catholics are right about absolutely everything.

Would you become a Catholic in this world? Or are you willing to admit that maybe your rejection of Pascal's Wager has less to do with a hypothesized p

... (read more)

I apologize for banging on about the railroad question, but I think the way you phrased it does an excellent job of illustrating (and has helped me isolate) why I've always vaguely uncomfortable with Utilitarianism. There is a sharp moral contrast which the question doesn't innately recognize between the patients entering into a voluntary lottery, and the forced-sacrifice of the wandering traveller.

Unbridled Utilitarianism, taken to the extreme, would mandate some form of forced Socialism. I think it was you who commented on OvercomingBias, that one of t... (read more)

Unbridled Utilitarianism, taken to the extreme, would mandate some form of forced Socialism.

So maybe some form of forced socialism is right. But you don't seem interested in considering that possibility. Why not?

While Utilitarianism is excellent for considering consequences, I think it's a mistake to try and raise it as a moral principle.

Why not?

It seems like you have some pre-established moral principles which you are using in your arguments against utilitarianism. Right?

I don't see how you can compromise on these principles. Either each person has full ownership of themselves (so long as they don't infringe on others), or they have zero ownership.

To me it seems that most people making difficult moral decisions make complicated compromises between competing principles.

Utilitarianism itself requires the use of some pre-established moral principles.
Thought experiment: A dictator happens to own all the property on the planet. Until now, he has been giving everybody exactly enough food to survive. In a fit of rage/madness, he stops. You would support the death of all humans other than the dictator, rather than taking his property?
Good god, Aurini (2009) sounds quite pompous. I can't even deal with reading his entire comment. I've since drifted from Libertarian to full-fledged Reactionary. I will attempt to answer the question as such. Either the Dictator is God, and we're all damned anyway, so any question of 'Rights' is irrelevant - or the Dictator is Moral, in which case we will kill him by sodomizing him with a red-hot poker. He remains King, so long as he is a competent King (through our eyes, his police's eyes, et cetera). Supposition: the Dictator is God, but only a God - his peers see his mistreatment of us. Recognizing the fact that he is wasting Good Wheat, they murder him, and install a new Potus. Life returns to happiness - because happy citizens pay the most taxes (just ask Russia). EDIT: Neither of which is an "End of History" solution, mind you, but I'm just beginning to realize how intractable the problem is. Obviously the new Dictator God will be just as idiotic and faliable as the last - which is why, as nice as Monarchy might seem, it ultimately self-destructs into Democracy, and then Dictatorship (just ask Marc Antony).
Most LessWrong posters are still firmly in the Cathedral and may fail to appreciate the significance of this, for they can not imagine a world outside it. A sizeable minority though has been influenced by the teachings of Darth Moldbug and the other lords of the alternative right, showing them a surprising taste of the true power of intellectual reaction. Some such as I have embraced these teachings for now, since it seems the very complexity of the world around us demands it. For there are very difficult and old problems which despite protestation to the contrary remain unsolved. Yet sanity on them must be approached if humanity is to have any hope at all.
What an interesting way of dodging the question. What's this supposed to mean? First of all, the context of the hypothetical clearly indicates that the dictator is human. Second, what do you mean by 'Moral"? Also, why is it suddenly more acceptable to murder than to just take their property? There's these things called mental hospitals and prisons, you should look into them. Edit: Look, it might help to define the LCPW in this context: the question is whether it is ever moral to take someone's property from them, yeah? So focus on that by making the actual act of doing so really easy - he's an absent dictator, owning all property on Rinax from his penthouse on Earth. One day he gets buried in paperwork and forgets to sign the form releasing the next year's food allowance to his subjects. How long should they wait before they break open the shipping containers and steal his food?
Heh, I'm nothing if not Interesting. The quote is a typo, incidentally - I meant to write "morTal". As seductive as the concept is, I see no firmament underlying the basis of 'human rights' - without a godhead who frimly endorses them, I'm not sure what they mean, beyond self-evident utlitarianism... Oh Gods, have I become a Utlitarian? Possibly. It's hard to say; given that narrative of who is 'right' and who is 'wrong' is inevitably written by those who are on the firing squad, I tend not to like this question. I'm honestly not sure how to respond to your comment; what sort of reply would make sense? Let me ask you - was Darth Vader obviously the Good Guy, or was he a Villains whom you could Sympathize With?
"Giving something for 'free' is just another form of enslavement " Hmmmm, this actually really puzzles me - how do you handle inheritance? Presumably, it being my property, I ought to be free to delegate it as I wish in my will. But, equally, to the person receiving it, it's a 'freebie', a form of enslavement. What about the other gifts that come from a privileged upbringing (access to a good education, or even just good nutrtion, say)? Surely, as a 2 year old, you didn't do anything particularly special to be deserving of these things compared to our example African kids - indeed, I doubt there's anything a 2-year old could do to shift themselves from one circumstance to the other.
Or "mu". Ownership, self or otherwise, is the wrong frame entirely, for instace.
There are important differences between moral principles and government policies. Even if you accept the premise that the morally optimal course of action is X, it does not logically follow that the government should mandate X. For one thing, it may or may not be feasible to enforce such a law, or the costs of implementing it may outweigh the benefits. Furthermore, some moral philosophies (though not utilitarianism) place firm boundaries on what is and is not the proper role of government. I would be curious to know your true rejection of utilitarianism.
More generally, reaching the moral conclusion that agent A should do X (or even is obligated to do X), doesn't obviously entail that agents B, C, D should compel A to do X, nor punish A for failing to do X — nor even that B, C, D are permitted to do so.
0Swimmer963 (Miranda Dixon-Luinenburg) 13y
The Canadian government has socialist elements, and I wouldn't mind and would even choose to live in a society that had more. As far as I know, the only freedoms it takes away are those that infringe on other people...considering that human beings are social animals, many or most of our decisions do affect other people. (I have not researched this. Feel free to prove me wrong or enlighten me on other aspects.)
The problem for me - speaking as a Canadian - is that there's no choice about it. To be honest Canada's a pretty good place to live. Despite the personality-disordered weirdo we have running the place, it's relatively free; decent amounts of freedom of speech, stable currency, only moderate corruption in our police forces, and greater economic liberty than the US (that's right - Soviet Canuckistan is less government run than the US) - my biggest upsets are Gun Control, the state of Domestic Violence Law, and the 'Human Rights' Tribunals which censor speech critical of protected groups. The worst thing our Monster in Parliament is trying to do is enact the equivalent of the Patriot Act, ten years too late. The fundamental problem, though, is the lack of choice to begin with - immigration has huge barriers, and it's not like there's room for any more countries. We're all forced into the coutnry we live in, and I suspect that the real civilizing force is the decency of regular people, who manage despite the government. It's like the post office, fifty years ago - they delivered the mail, they were adequate, but they weren't performing anywhere near the level that was possible. Nobody complained (much) because they were accustomed to it. As soon as private delivery companies entered the scene.the post office had to shape up fast. If I had the choice, I might choose to enter a socialist collective of sorts - at the very least, I'd want to live in an incorporated city which took care of the roads and sewers. The same thing should go for countries; nobody forces me to live in Calgary and accept the local tax burden, it wouldn't be right. Similarly it isn't right to force people to pay taxes in a country, when they're deeply opposed to certain elements of government. Keep in mind, I'm not just complaining without a solution in mind; there are workable solutions that would pay for things such as national defence, while subjecting government to the integrity of the privat
Anectodal evidence: In France, the post office is much worse since they have competition.

I don't see any problem with acknowledging that in a world very different from this one my beliefs and actions would also be different. For example, I think the fact that there are and have been so many different religions with significantly different beliefs as to what God wants is evidence that none of them are correct. It follows that if there was just one religion with any significant number of adherents then that would be evidence (not proof) that that religion was in fact correct.

Maybe if Omega tells me it's Catholicism or nothing I'll become a Cath... (read more)



Do you have a blog or home page with more material, or is there another site (apart from OB) with contributions from you that might be interesting to LW readers?


Yvain, you frequently seem to have extra line breaks in your post, which I've been editing to fix. I'm leaving this post as is because I'm wondering if you can't even see them, in which case are you using an unusual browser or OS?

So I asked him, "In the least convenient possible world, the one where everyone was genetically compatible with everyone else and this objection was invalid, what would you do?"

That's a pretty damn convenient world. It's basically like saying "In a world where serious issue X isn't applicable, what would you do?" which might as well be the better question instead of beating around the bush.

Sorry if this was posted before.

The acceleratingfuture domain's registration has expired (referenced in the starting quote) (

I have a question related to the initial question about the lone traveler. When is it okay to initiate force against any individual who has not initiated force against anyone?

Bonus: Here's a (very anal) cop out you could use against the least convenient possible world suggestion: Such a world—as seen from the perspective of someone seeking a rational answer—has no rational answer for the question posed.

Or a slightly different flavor for those who are more concerned with being rational than with rationality: In such a world, I—who value rational answers above all other answers—will inevitably answer the question irrationally. :þ

I'm not sure what this means. There is a finite number of choices. Each of them has a specific utility. The one with the highest utility is the most rational. Are you saying that one or more choices has undefined utility?

In fact (Omega tells you) either God does not exist or the Catholics are right about absolutely everything.

That sounds like it would decrease my probability that God exists by several dozen orders of magnitude.

Yes, but the important part is that it would mean that you know God won't punish you for becoming a Catholic.
I should point out that - if for some reason we're taking absurdly low-probability hypotheses into account - the idea that religion will prevent us from using the Force to live forever seems more likely to me than any deity who could offer us eternity.
Generally you use the probability times the utility. It would seem reasonable to take absurdly low-probability hypotheses into account if the difference in utility is absurdly high. That being said, refusing to take into account probabilities below a given value regardless of utility is a perfectly acceptable answer. I can't assert that you take them into account any more than I can assert you're a utilitarian in the doctor example. I don't know if the Force counts as a religion, but even if it doesn't there are a few things that are not religions that would work. You are still missing the point, though. Lets say that Omega also gives an upper bound for the absolute value of utility you will have if catholicism isn't true.
I know you've seen the Pascal's Mugging problem - that's what I meant to refer to. An upper bound to utility elsewhere doesn't matter if P(Catholicism) gets a sufficient leverage penalty (and the same again for all stronger claims). Are you saying that according to Omega, Hansonian leverage penalties are unsalvageable and this upper bound is the solution? (On its face, the claim "Catholicism is true" does not logically rule out the Mugger's claim, but of course we could go further.) I'd be more skeptical about this than I would be if Omega told me P=NP and also self-modifying AI is impossible by Godel's Incompleteness. But of course if I accepted it, this would change the equation.

I think that traveler's problem may pose two questions instead of one. First of all - is that a right thing to do just once, and the second is if it's good enough to be a universal rule. We can counclude that's the same question, because using it once means we should use it every time when a situation is the same. But using it as a universal rule has an additional side effect - a world where you know you can be killed (depraved of all posessions, etc.) any moment to help some number of strangers is not such a nice place to live in, though sometimes it's po... (read more)

Would this be moral or not?

Of course it is, if you live in this hypothetical world. The fact that in real life things are rarely this clear, or the fact that in real life you will be jailed for doing this, or the fact that you'd feel guilty if you do this, or the fact that in real life you won't have the courage to do this, doesn't mean that it's wrong.

But in real life I'd hardly ever violate the libertarian rights because of all the reasons mentioned above.

The biological commentary is indeed accurate, but I question its relevance in the context of the question, which seems to be one in favor of a utilitarian ethical discourse without the biological considerations. It might be better to assume the biological factors involved are compatible, or assume all other factors are equal, and disregard the biology.

The first answer that comes to mind for most I'm sure is that 10 is greater than 1, and that such a sacrifice would return a net gain in lives saved. However, this question is complicated by what it is abou... (read more)

Why, if the sick people are so close biologically, can't we sentence one of them instead to help the rest?

But in the least convenient possible world, here comes Omega again and tells you that Charity X has been proven to do exactly what it claims: help the poor without any counterproductive effects.

You don't need the least convenient possible world and Omega for that; for non-excessively-large values of proven, this world and suffice. I'm surprised that in three years nobody pointed that out before.


I dont think this would be a moral example Debt Consolidation Loan

The least convenient possible world is the one where Omega, the completely trustworthy superintelligence who is always right, informs you that God definitely doesn't value intellectual integrity that much. In fact (Omega tells you) either God does not exist or the Catholics are right about absolutely everything.

The problem with this specific formulation is that fundamentalist Christian beliefs are inconsistent, and thus it is trivially follows from Omega's wording that God does not exist.

A better wording would be to postulate that Omega asserts the poss... (read more)

"first, do no harm"

It's remarkable that medical traditions predating transplants* already contain an injunction against butchering passers by for spare parts

*I thought this was part of the Hippocratic oath but apparently it's not

An injunction to do no harm is part of the Hippocratic oath, and the actual text has multiple translations, so I don't think it's too far-fetched to attribute "first, do no harm" to the oath.
Obligatory wikipedia link. On the other hand: This was from the article on first, do no harm.

3: Extreme Altruism.

I don't want to save starving Africans. In most circumstances I would not actively mass murder to cull overpopulation but I wouldn't judge myself immoral for doing so.


2: The God-Shaped Hole....Do you admit that even if believing something makes you happier, you still don't want to believe it unless it's true?

I would, and do, admit that I don't want to believe it unless it's true. I watch myself make that decision more than enough to be honest about it.

(I'll note that believing things that aren't true makes me miserable and stressed. My verbal beliefs go about interfering with my behavior and my aversion to hypocrisy frustrates me. I'm usually better off believing the truth and just going along with the lie. However, I've assumed that in the least convenient possible world my God shaped hole was repared to normal function.)


You are a doctor in a small rural hospital. You have ten patients, each of whom is dying for the lack of a separate organ; that is, one person needs a heart transplant, another needs a lung transplant, another needs a kidney transplant, and so on. A traveller walks into the hospital, mentioning how he has no family and no one knows that he's there. All of his organs seem healthy. You realize that by killing this traveller and distributing his organs among your patients, you could save ten lives.

I wouldn't kill him. It isn't worth the risk for me. I al... (read more)


1: Omega, the completely trustworthy superintelligence who is always right, informs you that God definitely doesn't value intellectual integrity that much. In fact (Omega tells you) either God does not exist or the Catholics are right about absolutely everything.

Would you become a Catholic in this world?

Yes, plus pay bribes/alms at whatever the going rate is for having doubts that have been updated to greater than 50% based on observations. Since faith is somewhat destinct than raw prediction I suspect God'd be cool with that.

*kill traveler to save patients problem

assuming that

-the above solutions (patient roulette) were not viable

-upon recieving their new organs, the patients would be restored to full functionality, the equal of or better utility generators than the traveler

then I would kill the traveler. however, if the traveler successfully defended himself, and turned the tables on me, I would use my dying breath to happily congratulate his self preservation instinct and wish him no further problems on the remainder of his journey. and of course Id have left instructions w ... (read more)

The problem is the "least convenient world" seems to involve a premise that would, in and of itself, be unverifiable.

The best example is the pascals wager issue - Omega tells me with absolute certainty that It's either a specific version of God (Not, for instance Odin, but Catholicism), or no God.

But I'm not willing to believe in an omniscient deity called God, taking it back a step and saying "But we know it's either or, because the omniscient de . . . errr . . . Omega tells you so" is just redefining an omniscient deity.

Well, if I do... (read more)

Yes, to make it work, you may have to imagine yourself in an unreachable epistemic state. I don't see why this is a problem, though.
No, to make it work you have to assume that you believe in omniscience in order to clarify whether you believe in omniscience, a classic 'begging the question' scenario.
You're right that the existence of Omega is information relevant to the existence of other omniscient beings, but in the least convenient world Omega tells you that it is not the Catholic version God, and you still need to decide if that being exists. (And you really do have to decide that specific question because eternal damnation is in the payoff matrix.) Omniscience is almost a side issue.
Not if omniscience is A) a necessary prerequisite to the existence of a deity, and B) by definition unverifiable to an entity that is not itself omniscient. Without being omniscient myself, I can only adjudge the accuracy of Omega's predictions based in the accuracy of it's known predictions versus the accuracy of my own. Unfortunately, the mere fact that I am not omniscient means I cannot, with 100% accuracy, know the accuracy of Omega's decisions, because I am aware of the concepts of selection bias, and furthermore may not be capable of actually evaluating the accuracy of all Omega's predictions. I can take this further, but fundamentally, to be able to verify Omega's omniscience, I actually have to be omniscient . Otherwise I can only adjudge that Omega's ability to predict the future is greater, statistically, than my own, to some degree 'x', with a probable error on my part 'y', said error which may or may not place Omega's accuracy equal to or greater than 100%. Omega may in fact be omniscient, but that fact is itself unverifiable, and any philosophical problem that assumes A) I am rational, but not omniscient B) Omega is omniscient, and C) I accept B as true has a fundamental contradiction. By definition, I cannot be both rational and accept that Omega is Omniscient. At best I can only accept that Omega has, so far as I know, a flawless track record, because that is all I can observe. Unfortunately, I think this seemingly small difference between "Omniscient" and "Has been correct to the limit of my ability to observe" makes a fairly massive difference in what the logical outcome of "Omega" style problems is. Jonnan
The whole idea of an unreachable epistemic state seems to be tripping you up. In the least convenient world, you know that Omega is omniscient, and the fact that you cannot verify that knowledge doesn't trouble you.
Argument #1 works in the least convenient imaginable world, in my opinion. However, the OP concerns the least convenient possible world. The existence of an omnicient Omega seems to be possible in only the same sense as the existence of a deity; i.e., no-one has proven it to be impossible. The ability to hypothesize the existence of Omega doesn't imply that its existence is actually possible.
It's been more than two and a half years, dude! OK, here goes. I made a misstep by involving Omega in my least convenient world scenario at all. But I was right to try to redirect attention away from omniscience -- it just doesn't matter how you get to the epistemic state of discounting all possibilities other than Catholicism or atheism. All you need to grant is that it's possible for your brain to be in that state. Did knowledge from Omega put you there? Did you suffer an organic brain injury? Did your social context influence the possibilities you were willing to consider? Were you kidnapped and brainwashed? Who cares? It's irrelevant -- the presence of eternal damnation in the payoff matrix makes it so. However you got there, you must now face Pascal's Wager head on. How will you answer?
sorry - I was led there by a recent thread. Given the epistemic state of recognizing only those two possibilites, I suppose I would cop out as follows. I would examine the minimum requirements of being a Catholic, and determine whether this would require me to do anything I find morally repugnant. If not, I would comply with the Catholic minimum requirements, while not rejecting either possibility. In other words, I would be an agnostic. (I don't think Catholicism requires a complete absence of doubt.)
You could two-box. If you get the million and the thousand you prove that he's not omniscient. All that's required is that you make the choice he did not predict.

“Do you head down to the nearest church for a baptism? Or do you admit that even if believing something makes you happier, you still don't want to believe it unless it's true?”

I believe that God’s existence or non-existence can not be rigorously proven. Likewise there is no rigorous protocol for estimating the chances. Therefore we are forced to rely on our internal heuristics which are extremely sensitive to our personal preferences for the desired answer. Consequently, people, who would be happier believing in God, mostly likely already do so. The same p... (read more)

I think Pascal's Wager and the God-Shaped Hole should get more play.

To your Pascal's Wager statement

Perhaps God values intellectual integrity so highly that He is prepared to reward honest atheists, but will punish anyone who practices a religion he does not truly believe simply for personal gain.

I don't think what you say is incommensurable with the Catholic position that what is most important to the Omega is that we pursue the best thing we know i.e. intellectual integrity along with charity. But perhaps I am wrong. You might know more about this th... (read more)

I think the GSH is largely that our whole way of thinking, our terminology, our philosophy, our science evolved in theistic societies. Taking god out of it leaves a lot of former linkages dangling in the air, probably we learn to link them up sooner or later but it requires revising a surprisingly large amount. For example, a godless universe has no laws of nature, just models of nature that happen to be predictive. For example, there isn't really such a thing as progress because there cannot be a goal in history in the godless universe. There is social change, and it is up to you to judge if it is good. For example, there are no implicit promises in the godless universe, we could do everything "right" and still get extinct. This is non-intuitive deep on the bones level, our whole cultural history teaches that if you we make a good enough effort some mysterious force will pat our backs, give a B+ for effort and will pick up the rest of the slack: because this is what our parents and teachers did to us. Just look at common action movies, they are about heroes trying hard, and almost failing, then getting almost miraculously lucky. Deus ex machina. The GSH becomes very intense when you start raising children. For example it would mean not giving praise for effort, in fact, sometimes punishing good solutions to demonstrate how in the real world you can do things right and still fail. This would be really cruel and probably we don't want to do it. Most education tends to imply what it teaches is certain truth, laws of nature etc. so things get hard from here.
There's a nice exposition of roughly this idea over at Yvain's / Scott Alexander's blog.
To Hollander: When we create models, they are models of something other than your own mind's processes. Or are you a coherence theorist/ epistemological anarchist? I think that some models (of progress, of biology, of morality) are more true aka, less wrong. Their predictive power comes from the near-miraculous fact that the symbols we use for math and science can be manipulated and after the manipulation still work in the world! I am always in awe at this natural wonder. Logic, Nature, Beautiful. gjm: Thanks for that link! It's really good, as is the previous post on his blog. I underestimate how metaphysically-light most atheisms are. Since I still believe in a knockout-fundamental-goodness in the universe that we model with morality, I might be more in Scott's camp.

In good ol days there was concept of whose problem something is. It's those people's problem that their organs have failed, and it is traveller's problem that he need to be quite careful because of demand for his organs (why he's not a resident, btw? The idea is that he will have zero utility to village when he leaves?). Society would normally side with traveller for the simple reason that if people start solving their problems at other people's expense like this those with most guns and most money will end up taking organs from other people to stay alive ... (read more)