Here’s a question to ponder:
|
Question 1: If forced to choose, which of these nightmare scenarios would you prefer?
Scenario A: An evil alien flips a coin. If it comes up heads, he destroys all human life; otherwise he goes home. Scenario B: The same evil alien flips 7 billion coins, one for each person on earth. He destroys anyone whose coin comes up heads. |
I’ll tell you in a minute why I ask, but first let’s consider arguments in each direction:
Argument 1. In scenario A, I have a 50-50 chance of death, and a 50-50 chance of continuing my current life. In scenario B, I have a 50-50 chance of death, and a 50-50 chance of a life in which half my loved ones are gone. Surely I should take A.
Argument 2. In scenario A, there’s a 50-50 chance that all future generations will be destroyed-in-advance. In scenario B, even if the coin comes up heads, people will continue to be born, and in the very long run, the evil alien will be a forgotten memory. Surely I should take B.
Your answer to this question, I think, is likely to reveal a lot about how much you think we owe to future generations. If you think we owe them nothing, then Argument 1 is definitive. If you think we owe them the same respect we owe our contemporaries, then Argument 2 is definitive. If you think we owe them something in between, you might waver.
Now that sort of question might strike you as nothing more than Sunday-afternoon dorm room fare, but I don’t believe it can be dismissed so easily. It is pretty much impossible to take a coherent stand on issues ranging from Social Security reform to environmental conservation without first deciding how much we are obligated to care about future generations. A lot of people seem to think those issues are worth debating, which pretty much forces us to face up to the fundamental issues.
On the other hand, come to think of it, I suppose a person might prefer Scenario B to Scenario A for reasons that have nothing to do with future generations — namely the desire to be remembered. Is that something we care about? To focus on that issue, here’s another question:
|
Question 2: Suppose you’re happily married. If forced to choose, which of these nightmare scenarios would you prefer?
Scenario A: An evil alien flips a coin. If it comes up heads, he kills you and your spouse; otherwise he goes home. Scenario B: The same evil alien flips a coin. If it comes up heads, he kills just you; if it comes up tails, he kills just your spouse. |
Here again we have:
Argument 1. In scenario A, I have a 50-50 chance of death, and a 50-50 chance of continuing my current life. In scenario B, I have a 50-50 chance of death, and a 50-50 chance of a life in which my beloved spouse is gone. Surely I should take A.
Argument 2. In scenario A, there’s a 50-50 chance that my spouse and I will both be dead and unremembered (or at least unremembered by anyone who knows us as intimately as we know each other). In scenario B, however, we each get to live on, at least in the memory of a loved one. Surely I should take B.
Your answer to Question 2 will tell me something about how much being remembered matters to you, which will help me interpret your answer to Question 1.
The original version of this post included several more followup questions. But I suspect these two are grist enough for a day or two of discussions. I’ll post the followups after those responses start to peter out.
Related post here.
1) B
I don’t care about being remembered, but I’d hate to think that my decision is responsible for the complete extinction of the human race.
2) A
I don’t want either of us to have to suffer the loss of a spouse.
1) B, because life will go on. The argument you gave for A is just selfish.
Coase theorem would dictate that I could arrange for something in between with the Alien.
Something more mutually beneficial.
Some parents show a preference for 2B when they fly separately. The idea is that a plane crash will not make their child an orphan if the parents are not on the same plane.
(Your numbering is confusing. I suggest you change Arguments 1 and 2 to Arguments A and B.)
Interesting “dorm room fare” as you put it, but I don’t think one’s answers to these questions would give much guidance as to how to think about Social Security etc., because in these riddles, we’re weighing how much we owe it to future generations to bring them into *existence*.
For debates about Social Security and related issues, we know that the future generations will exist, and the question is what quality of life we owe it to them to provide.
Before having a kid, I might ask how much I am obligated to bring a hypothetical human into existence. After the kid is born, I might ask how much I am obligated to provide for them. I submit that these two questions can be answered completely independently, and there might not even be a correlation between the answers across a large sample of respondents.
Q1. From an evolutionary perspective, more of your genes are more likely to persist in scenario B. Total extinction of humanity removes all chance of your genes propagating. Removal of 50% of humanity does not prevent the genes you share with family (and all humanity) from propagating. You should prefer B.
Q2. Evolutionary perspective again. This probably depends on the ages of any children you might have, and your own age. Each scenario gives you 50% chance of survival. If you are old-ish with young children, then survival of one parent would maximise survival chances of those children, hence your genes. You should prefer B.
If you are young and childless, then maintaining your spouse will probably maximise your reproductive success. You should prefer A.
Personally, my choice was B and B. I am old-ish with a young-ish child.
1) B
2) B
“Future generations” never entered into my analysis, though. My heuristic for choosing was something along the lines of “select the scenario that eliminates the worst possible outcome.”
Harold:
Q1. From an evolutionary perspective, more of your genes are more likely to persist in scenario B. Total extinction of humanity removes all chance of your genes propagating. Removal of 50% of humanity does not prevent the genes you share with family (and all humanity) from propagating. You should prefer B.
My preferences are shaped by evolution only insofar as my ancestors faced situations in which those preferences were relevant. I’m not sure that applies in this case.
I’m with James from Pittsburgh, some life is better than no life.
Of course, a better solution would be to call in Doctor Who to take out the evil alien. :-)
hmmmmmmmmmm…..
Scenario A: An evil politician flips a coin. If it comes up heads, he destroys Social Security; otherwise he goes home.
Scenario B: The same evil politician flips 300 million coins, one for each US Citizen. He destroys Social Security for anyone whose coin comes up heads.
Sometimes death is the easier choice.
In the happy marriage scenario you miss an important consideration. I might believe my spouse would be miserable without me and wish to spare him or her that. This confounds your desired inference.
1B) Smallest odds of total annilihation of the human races
2B) Smallest odds of my kids having 0 parents
T.
JamesFromPittsburgh:
1) B
2) B
“Future generations” never entered into my analysis, though. My heuristic for choosing was something along the lines of “select the scenario that eliminates the worst possible outcome.”
Worst possible outcome for whom? Question 1 is designed such that any current individual, on a purely selfish level, would prefer A to B. It seems that choosing B must be a result of concern for others—and not other present individuals, because their selfish preference would also be A, thus you are doing them no favor by choosing B.
You may not have explicitly thought “I need to think about the future generations!” when pondering this scenario, but it seems to me that a (possibly subconscious) concern for future generations is the only rational justification for choosing B over A.
1) B
The value of the last remaining thousand humans is greater than that of any other thousand humans, ceteris paribus. Thus utility as a function of destroyed lives is not linear.
Alternately worded: Choosing A exposes the “final thousand” to a 50% chance of death. Choosing B almost certainly guarantees their survival.
2) A
I’m not sure about this one, though. The considerations above don’t seem to apply. Against Steven’s pro-B argument, the idea of figurative immortality appeals to me as much as it did to Woody Allen.
http://en.wikiquote.org/wiki/Woody_Allen
I’d suggest that this question does not, in fact, reveal a lot about how much we think we owe to future generations.
The more the effect of the choice, the less the obligation, and vice versa. If you’re obliged to defer to someone even for insignificant things, that’s a large obligation, and if you need not defer to them except under the most extreme circumstances, that’s a small obligation. It’s hard to get circumstances more extreme than this.
Of course, how large the obligation is is also affected by how much you have to give up, and you can argue that scenario A means you have to give up more, on the average (although that can be questioned, since the argument assumes that you don’t value the lives of your loved ones independently of whether you are around to be with them).
Still, that means that this generalization doesn’t work when deciding whether it means you’re obligated in real life situations such as Social Security. Social Security is a much smaller benefit to future generations than avoiding complete annihilation, so just because you have to defer to future generations if they gain a large benefit (=small obligation) doesn’t mean you must to grant them a small benefit (=large obligation).
I think I share with some of the previous commenters a dislike for the framing of the questions. I don’t think “obligation to future generations” is a useful framework for much personal decision-making, though it can be a useful framework for social collective decision-making.
In making personal decisions I weigh personal factors such as my childrens’ futures more highly. When weighing in on social/policy matters I try to think in more general terms, even though my personal preferences may not generalize 1:1.
In neither case do I think about it in terms I would refer to as “obligation.” I feel, and I infer that other commenters feel, that you are leading toward a statement of the form “If you choose option X then you must prefer policy Y and therefore advocating Z is illogical/inconsistent/erroneous.” Therefore I find myself resistant to the framing, without knowing a priori what you think it logically entails.
Flipping a coin takes about a second. 7 billion seconds is over 200 years. I will take B. Perhaps the alien won’t get to me before I die naturally.
I’m sure B (question 1) would provide a hell of natural experiment
After reading question 1 I thought this was going to be another plug for how much better everyone is when there are more people.
Perhaps one could introduce a gradated experiment to help the point along:
Scenario 1: The evil alien flips 7 billion coins and for each one that comes up heads, he introduces a genetic disease which will kill the nth generational descendents of the unlucky person (the unlucky person is not informed of their being unlucky) .
Scenario 2: The evil alien waits n generations, flips the billions of coins (we don’t know what the future’s population will be) and kills off those for whom the coin comes up heads.
I think we miss corollary effects in the first scenario. Sure, option B leaves 50% of humans alive, but the effects on society of a loss of half the population would be, at least in the short term, devastating. Entire businesses, governments and economies would collapse. Many families would be left with no adult support at all.
Option B gives 100% chance of a terrible result. Option A gives 50% chance of terrible result that no one would be around to suffer from and 50% chance of status quo. I choose A.
Though I like Neil’s thought process.
Todd,
I see your point, but I think concern for only the present generation is sufficient to choose B. If we were to assume that the evil alien sterilizes the entire population of the planet before offering the choice, future generations couldn’t enter into the analysis at all. I’d still think B would be the better choice as it guarantees that some people will get to keep living.
To add to the fun. In question 1, what fraction of world income in perpetuity are we willing to pay to the alien not to flip the coin(s) in both parts? Does it make a difference whether we act cooperatively or make voluntary contributions independently?
None,
Whatever deal we make will be reneged upon. Blackmailers always return for more once you’ve demonstrated willingness to pay.
1) waver
2) B
I am married with young children, I wouldn’t want my children to risk loosing both of us.
If I pick 1,A, I’m not responsible for destroying the world, the alien is.
Am I at more risk of death with my wife if I fly on a plane with her as opposed to drive in a car with her or cross the street with her? Parents who fly separate, but don’t drive separate because they don’t want to orphan their children are, in my opinion, behaving irrationally.
Eliezer,
You are complicit in the destruction of humanity as you could have chosen B which would have (statistically) assured survival.
1B, no question.
I’ve never feared annihilation by aliens. But discussions about the efficiency of paper money – and electronic transactions – has long caused us numismatists to fear for the annihilation of coins. So the idea that aliens would gather seven billion coins warms my heart.
I agree with TjD.
@nobody.really: Then we probably ought to go to a bi-metallic standard to avert a specie shortage. Of course I haven’t considered that the original demand will be halved …
@Kirk, Humanity has been through this before. Most recently with the Black Death, where about 50% of Europe’s population (and probably Asia’s too) died. It was a horrible experience, I’m sure. And yet I suspect that most of the people who survived the plague preferred survival to death.
1A – I want to live and I want my wife and son to live with me. That is my #1 priority. 1B is likely going to thwart that desired outcome. Also, I don’t want half the world’s population destroyed, unless I can guarantee that they are all Obama supporters (-;
Choosing 1B guarantees unspeakable horror. If we are all gone, there is no one left to weep. 1A is my choice, and it is not even a close decision. Odds are pretty good that we all are unaffected.
2A – My wife, my son, and I are a team. We want to sink or swim together. 2B guarantees horrible pain for my son. I don’t want him to feel that pain. 2A is my choice, and it is not even close.
“we owe to future generations”
Owe is the wrong word. We may care about the future generations without feeling that we owe anything to them.
These are interesting scenarios.
Q1. Let k be the population, then the P(X) = 0.5, in scenario A, where P(X)=2 ^ (-k) in scenario B, where X is the event where k people die. From this perspective, it seems clear, you should choose scenario B over A.
I think Q2 is even more interesting.
Q2. Let n=2k, where k is the number of couples. Further, define the event m’ as the event that m people die. For scenario A, P(2i’) is non-zero for all values of i less than or equal to k (and zero for all values 2i-1), including everyone lives and everyone dies. In scenario B, P(m’) is 1 when m = k and 0 when m != k . Thus in scenario A, it’s possible for all people to live and for all people to die, where as in scenario B, the number of people who die is guaranteed to be exactly k. I don’t think it’s clear whether you should choose A or B for any value of k.
I reflexively choose 1B – OF COURSE we’d act to preserve the future of the human race. The idea that anyone would stop to consider their own parochial concerns when the fate of humanity rests in the balance strikes me as beyond contemptible; it’s simply absurd. This is patently obvious because …
… well, upon reflection, I don’t know. At a visceral level, the question simply prompts a home-team booster reaction in me. I feel prepared to sacrifice for this abstract notion called “the future of the human race.”
In contrast, a colleague immediately draws the contrary conclusion. I’m constantly surprised to realize how often we draw opposite conclusions. I sense my colleague is anxious and loss-averse, placing little value of the prospects for future gain but highly sensitive to signs of social decay. She’s also quite cynical. She supports redistributionist policies designed to help her fellow citizens (she’s nationalistic in a compulsory-mutual-aid-society way), but is generally skeptical of calls to sacrifice for anything more abstract. And the needs of hypothetical future people qualify as “more abstract” than the needs of today’s people. While she’s not overtly hostile to environmental policies, she generally concludes that they reflect “elitist” concerns unrelated to the more pressing needs of her fellow citizens.
Then again, I could imagine some people might conclude that the fate of the Earth depends upon some predator – aliens, viruses, whatever – arising to check the uncontrolled growth in the population of humans. People who love whales, for example, might well actively court an option to cull the human herd – even to the point of extinction.
Harold said ‘. From an evolutionary perspective, more of your genes are more likely to persist in scenario B. Total extinction of humanity removes all chance of your genes propagating. Removal of 50% of humanity does not prevent the genes you share with family (and all humanity) from propagating. You should prefer B.’
Steve replied ‘My preferences are shaped by evolution only insofar as my ancestors faced situations in which those preferences were relevant. I’m not sure that applies in this case.’
Harold’s solution seems natural to me. Steve’s response raises the ‘extended phenotype’ question which broadens the original question, from ‘obligation to future generations’ to the entire biosphere. Dawkin’s extended phenotype principle suggests that genes which are adaptive are ones which are ‘selfish’ across bodies and species.
Another point relates to the use of the word ‘obligation’ rather than ‘preference’ re. transmitting a benefit to future generations. The word ‘obligation’ means we have entered the realm of deontics. But we wouldn’t need deontics, there would be no evolutionarily stable scenario where a species which evolved by natural selection would expend energy on the topic unless there had been genetic canalization of pehnotypal plasticity during some bottleneck. The one thing we can be sure off is that we all have ancestors who went through pretty much the same bottleneck so if it is the case that evo devo arguements are relevant in this context then it must also be the case that best there is some mixed strategy which is optimal= in other words the menu of choice is both unknown and severely constrained.
It’s interesting to read the various responses, especially to see other perspectives. For question 2, I chose A without hesitation. I’m happily married but with no kids. However, hearing from the people with young kids, I could very easily see why someone would choose B even though it is horrible.
For question 1, I would have to choose B. I couldn’t live with being responsible for the extinction of humanity (and yes, I know by definition I wouldn’t “live with” that decision as we would all be dead. Still, that’s small consolation).
“And yet I suspect that most of the people who survived the plague preferred survival to death.” -Al V.
Yes, but you don’t really know how the dead people feel about death. Maybe death is only bad for the survivors who have to cope with the loss?
If all of humanity is wiped out, then who has lost? the souls that haven’t been born yet?
My point about evolution is that the obligation is not to future generations so much as to our own genes. Evolution selects creatures that maximise propagation of their genes. I think this is why the extinction of humanity feels so bad to many people – it eliminates their genes.
Our genes program our thinking. Surviving genes are the ones that program our thinking to avoid extinction.
Whilst our ancestors did not face malevolent aliens flipping coins, they did face choices where altruistic responses would maximise gene survival, at risk to personal survival. Thus our instincts are quite strongly to avoid extinction – or total destruction of our tribe.
I’m left wondering why the alien uses a coin with head and tails in the first place. I mean, doesn’t he (she? it? qrx??) have his (qrxs?) own random number generator? The fact that qrx uses such a coin means qrxs psyche is inextricably bound up with human culture. Clearly then, this is not the last time some poor soul on this planet will be faced with this dilemma or a similar one qrx devises. Qrx’ll be back! Probably, in fact, qrx’s (Qrx’n?) been here before!
Anyway, having shown qrxself so bound to human culture, I’ll steal one of qrxs old spaceships from Area 51, load it with an atomic bomb and a fancy computer virus on a Macbook, infiltrate qrxs Demographic Ordinance Ordering Mothership, introduce the virus, leave the bomb behind, and fly out before the whole thing explodes. Qrxs highly advanced technology will be no match for my cunning, daredevilry and carefully scripted plot holes! Ha! Steal our coins? We’ll show you! We shall prevail!
Mike H, I think the reason qrx uses a coin so that that we will clearly understand the question. We all know what a coin toss is.
In the unlikely instance that anyone can really imagine both what such a scenario would actually feel like and what he or she might actually choose, the answers given would have no relationship to how he or she behaves in the actual world we live in (one where there is not the slightest chance that an evil alien threatens humanity with mass homocide, especialy a threat that includes a bizzare choice of who gets killed or when). How we think about our world and our place in it depends on what our world is- the scenario described is not our world. I know economists are fond of using these imaginary scenarios stripped of variables in order to get at the “fundemental issues” but there must be a name for scenarios so different from our lived lives that our reaction to them proves (or even suggests) nothing.
Steve, I understand you are just trying to come up with a fun hypothetical to motivate discussion, but it somewhat disturbs me how you framed the two arguments in Question 1 (one for A and one for B). What if we’re just thinking of currently alive people, and someone is risk averse when it comes to the death of other human beings?
In other words, it seemed kind of weird to me that you framed both arguments in terms of, “How does the alien killing billions of people affect *me*?”
Oops sorry I was unclear. More specifically, this is what I’m trying to say: Argument 1 in Question 1 strikes me as an inconceivably selfish perspective, even if we stipulate that everybody is sterile and there will be no future generations.
So I’m asking, do you agree with me, and concede you were just trying to come up with some argument for Scenario A?
@Al, sure the live ones are glad to be alive, and most would probably choose B in retrospect, but ALL the dead ones would absolutely choose A in retrospect.
I agree with Bob.
Would you expect answers to change if there were no parent/child relationships, and there would be no “future generations”?
Also, there are many environmentalists who would choose B in favor of the status quo.
As for Social Security reform, that has nothing to do with what we owe future generations, it only has to do with what we hope we can convince future generations they owe to us. You can’t save the things you will need and want to consume.
Bob Murphy (and everyone else who has asked me specific questions): I’m on the road, no time to respond in detail, but I will revisit these next week. Meanwhile, carry on without me!
@Eliezer : foo! there goes my plan!
How do the answers change if we say :
“an evil alien tosses a coin. If it comes up heads, he utterly destroys all humans. If it comes up tails, he leaves us alone, and utterly destroys all members of the dominant lifeform on a remarkably earthlike planet we hadn’t yet discovered, who have a technological and cultural sophistication remarkably similar to ours, and even look a bit like us.”
Mike H,
I think that’s an interesting twist. And to complete the decision, do you choose the scenario you’ve laid out or do you choose the plan to have a coin flipped for each individual on each planet. So you can be responsible for the extinction of one of the species or killing half of both species (probably anyway, since there is still a non-zero probability of everyone living, as well as everyone dying in the second scenario).
I think I’d still choose the second choice, not because I don’t want the other race to go extinct, but because I don’t want the human race to go extinct.
Scenario 1: B, hands down. I think Keith nailed it. The value of the people remaining will be higher than the value of the people dying. From a deontological perspective they seem a wash.
In Scenario 2, my first instinct was to try to convince the alien to just kill me and go home, but I suppose if he was really evil, he would then kill my spouse and go home.
@Tom Burk: If the alien would outsource its coin-flipping to Diebold, you could get your wish.
I choose A both times. If you believe in the many worlds interpretation of quantum mechanics, humanity survives either way. Would it really bother anyone if it only survives in half of the future worlds…
Even without that, if I’m dead, I really couldn’t care less if an evil alien glasses the rest of the world too. Rather have a chance of a normal life if I survive.
I think Keith has it right. In the first nightmare there is a significant positive value to having (about) half survive. This is a value judgment, the same one that is made every day by old men who plant trees they won’t live to see grow. (Or by those who want the surviors to take revenge).
Less clear in the second. Jack and Jill together are happier than Jack seperately and Jill separately added up.
Uh oh, does anyone else think Steve has been abducted by aliens, and he’s asking the studio audience for help before giving his final answer?
1) How could you not choose B. What utility is there for you in there being 7 billion people. At a population of 3.5 billion, there are still more than plenty for you to enjoy network effects. It’s not like living people would be alone in their grief. Death would come to mean something other, and yeah it would be sad, but people are a resilient bunch.
2) You have life in 3/4 outcomes for B. Not being married, and perhaps not understanding ‘true love’ I take B.
I know you disagree with the sentiment, however, I can’t not express the sentiment that I believe we would be better off with half the population. I would be much happier in a state with 50% population than in the state with 100%. So my preference would be heavily towards B.
A better scenario, I believe, is: You’re on a desert with 20 people. You’re chances of survival go down a lot if you have <10 people. Which choice do you make? There is nothing at stake in the above scenario, even if there is supposed to be.
@Bob Murphy:
In other words, it seemed kind of weird to me that you framed both arguments in terms of, “How does the alien killing billions of people affect *me*?”
I agree. I think that acting as an agent for others I should consider their want but some do not http://www.thebigquestions.com/2012/03/22/another-nightmare/#comment-46760 responding to a point much like yours http://www.thebigquestions.com/2012/03/22/another-nightmare/#comment-46749
I know you are trying to get at a different point, but the implied moral baseline for your question is utilitarianism. There is often a place for utilitarianism in our world but I do view it only as a subset of other moral values. In my particular case, I would tell the “alien” to go F……k itself.
In the first question, scenario B is preferable because humanity is likely to survive. I consider survival more important than a temporary amount of unhappiness.
As to question 2, I’d go with scenario A unless there was a need for one of us to survive (we knew where the self-destruction switch was so we could prevent question 1 from happening, but only if one of us survives). Both surviving would normally be a much better outcome than the other three possible outcomes.
B…
for an alien to flip 7B coins sequentially (1 per sec), would take approximately 222 years without sleep, breaks, etc… Can that alien live that long?
Also in infinity (7B is close to that), the chances of heads and tails are 50-50.
I’ll take my chances…
Solution A makes the 50-50 probability way to risky…
Addendum:
Can the Earth benefit from the fertilizer provided by 3.5B people and all the other benefits..?
Q1. I would choose Scenario B because:
1. The probability of a significant number of humans surviving is greater in scenario B.
2. I like the logic of the poster above who said that the utility of the last 1000 humans surviving is greater than that if subsequent 1000s (though due to network effects the 1000 should be revised upward to a million or 10s of millions)
3. If the most likely outcome of scenario B is 3.5 B people, that takes us back to 1950-1970. That is still a pretty good era to live in.
4. I like humans, I enjoy being around other humans and I am sure other humans like other humans and like to be around other humans.
Q2. I would choose scenario B irrespective of the age of spouses, or whether they have children or not. In fact it is a very simple choice. My question to those who choose A, those who cannot imagine living without their spouses is how many of them would commit suicide if their spouse predeceases them?
My opinion of my own life, and of humanity in general, is such that it’s very, very easy for me to pick a pair of As, consider my utility function to only be based upon my own survival and not mind too much at all if we all go up in smoke at the same time. Anyone else who thinks this: you are not alone, and I for one think your view is absolutely OK.
Data points here: married, very much in love, no kids.
While your dilemmas are interesting, your rationales are thin. Do you really think the only answer to 2 is about rememberance? There’s a range of reasons for why people might choose either, but to prejudge it as “it tells me about how you want to be remembered” is putting the cart before the horse.
Similarly, question 1 is not just about future generations, it’s about all current human life as well. For it to be only about future generations, you’d have to say something like ‘sterilise’ rather than ‘kill’, and even then you run into complicating issues like who’s going to take care of the currently alive as they age.
The questions are interesting, but your interpretation of what they mean is highly off-kilter.