I think there is tons of uncertainty in utilitarianism, so I wanted to document some of the questions here. Feel free to add more questions or answers, and I'll try to keep this OP updated semi-systematically. Perhaps the utilitarian thing to do would be to spend some time bringing about answers to these questions?

The goal of this thread is not to start a discussion of every single problem of utilitarianism all in one (cramped) place, but to direct people to work on answering these questions, or create a repository for future research. Ideal conversations here would be meta-level (talking about the list, updating the list, telling me this is a dumb idea) or directing people to existing resources or conversations on each question in other threads.

Note that a blank answer to a question doesn't mean that there is no answer, or even that I don't have an answer, though it definitely could. And I probably forgot something, or many things.

This was last updated by me on July 14, 2012.

~

Problems in Meta-Ethics

Meta-ethics deals with questions like what the words "ought", "right", "good", etc. refer to and what morality is.

1.) Is there a meta-ethical reason to privelege utilitarianism over other moral systems? If so, what would it be? If not, is there a meta-ethical reason to dismiss utilitarianism completely?

I think this question has been resolved with good confidence that utilitarianism is just one moral system among many that can be used to judge action as morally right or wrong, and that other systems -- as long as they're internally coherent -- are no less valid from a meta-ethical standpoint. Thus, the adoption of utilitarianism over another system would be a matter of personal preference, even if it would be the utilitarian thing to do, and utilitarian-morally wrong not to do so.

Also, threats from non-cognitivism and anti-realism don't seem to undermine there being a fact of the matter as to whether something meets the standards of utilitarianism, so that would make utilitarianism an open option, and thus no reason to dismiss it either.

~

2.) If the adoption of utilitarianism is a matter of personal preference, why might someone want to adopt it?

The only reason I can think of is that some people just have utilitarian values as terminal values.

~

3.) If utilitarianism is largely a choice, how do we as utilitarians get along with those of other meta-ethical persuasions?

~

Problems in Normative Ethics

Normative ethics deals with how a given moral framework is actually implemented to measure action; how you figure out what is morally right given a stipulated definition.

5.) What matters when it comes to happiness -- total happiness or average happiness?

~

4.) How should preferences or happiness be aggregated?

~

6.) What does happiness look like or consist of? What does more happiness look like? What kind of happiness does utilitarianism care about?

~

7.) How is happiness measured?

~

Utilitarian Boogeymen

These problems straddle normative ethics and applied ethics, but they're classics, so it's important to keep them in mind all in one place.

8.) What do we make of the Utility Monster?

~

9.) What do we make of the Repugnant Conclusion?

~

Problems in Applied Ethics

Applied ethics refers to what actions are actually measured to be morally optimal; what the utilitarian thing to do actually is.

10.) What are some optimal actions for improving the world? What are some utilitarianism best practices?

Shifting toward veganism and shifting toward donating as much as your income to effective charities as possible seem to be commonsense targets right now.

~

11.) How does one keep a utilitarian motivation going?

Joining communities like GivingWhatWeCan, 80000Hours, or this forum seems to help. As is getting more utilitarian friends.

~

12.) What can be done about existential risk? Is reducing existential risk even desirable?

~

13.) What can be done about wild animal suffering?

~

14.) What are some effective utilitarian donation targets?

~

15.) What would a utilitarian utopia look like? Would it involve wireheading? Utilitronium?

~

Problems in Meta-Applied Ethics

Meta-applied ethics is the word I'm giving to the questions that need to be answered in order to better answer questions about applied ethics.

15.) Which currently existing life is sentient enough to suffer in ways relevant to a utilitarian?

Definitely humans, very likely other vertebrates, and fairly likely fish. Currently, investigation is most open on insects. See this essay at Vegan Outrech for more information.

~

16.) Do nonhuman animals living in the wild live net-negative lives?

~

17.) Would a future with humans be better than a future without humans?

~

18.) What should we make of Pascalian arguments of a shot at really high utility at a really low probability, especially when the precise level of utility or probability is unknown?

This problem is well expressed in [url="http://lesswrong.com/lw/kd/pascals_mugging_tiny_probabilities_of_vast/]"Pascal's Mugging: Tiny Probabilities of Vast Utilities[/url]. Another piece of wisdom aiming at a solution to this problem is "Why We Can't Take Expected Value Estimates Literally (Even When They're Unbiased).

5.) What matters when it comes to happiness -- total happiness or average happiness?

Total happiness. I like to test my intuitions about such questions with thought experiments. Here is one: You are at the end of your life. You can gain either one absolutely brilliant additional day, or 10 trillion worthwhile but only moderately good additional days (assuming no externalities in either case). Here, I would choose the 10 trillion, irregardless of the utility distribution in the rest of the universe.

8.) What do we make of the Utility Monster?

We recognize that it is an unrealistic concept, but conceptually bite the bullet. Thought experiment: An immortal person is caught in a magic trap that will torture him for 10 septillion years. For every second of pain we accept instead, we can shorten the duration of his pain by 1 year. The right thing to do, imo, would be to absorb all his pain. Conversely, if he could gain 1 year of pleasure for each second of pleasure we forgo, the right thing to do would be to forgo all our pleasure.

9.) What do we make of the Repugnant Conclusion?

We acknowledge that mere existence is not the same thing as utility, and that lives even barely worth living require a recognizable amount of good conditions. We acknowledge that bad is stronger than good and that the repugnant conclusion only applies if the quantity of additional net-positive life actually outweighs the quality reduction in the lesser populated scenarios. We then conclude that taking these conditions seriously, a scenario in which they are met is not in any way repugnant, but actually desirable.

13.) What can be done about wild animal suffering?

Some relevant links by Robert Wiblin, Robert Wiblin again, David Pearce.

I've argued that we should embrace the repugnant conclusion, and the reasons we think that this conclusion is repugnant are:
1. A cognitive bias called: Scope Insensitivity
2. Problems with estimating the zero utility level.
as I've argued here. I'm pretty satisfied with that solution to be honest, compared with many of these other problems which seem truly baffling.

The problem I'm keenest to address is:
1. Which careers have the highest incomes in each country.

This is a clearly solvable problem which I believe will lead to a solution of the greater problem: which careers are most useful.

RyanCarey wrote:I've argued that we should embrace the repugnant conclusion, and the reasons we think that this conclusion is repugnant are:
1. A cognitive bias called: Scope Insensitivity
2. Problems with estimating the zero utility level.
as I've argued here. I'm pretty satisfied with that solution to be honest, compared with many of these other problems which seem truly baffling.

I agree. The repugnant conclusion only seems weird because fundamentally humans don't understand numbers bigger than about four, and so we have no conceptual idea of what it meas to send the number of humans living marginally positive lives to infinity. Intuitively we just think of infinity as a big number, kind of like 100,000, or 18.

The problem I'm keenest to address is:
1. Which careers have the highest incomes in each country.

This is a clearly solvable problem which I believe will lead to a solution of the greater problem: which careers are most useful.

Interesting question; that being said, if you're defining useful as utilitarian I'm not sure how well income will represent it: many people make a lot of money betting on inefficiencies in the market (i.e. financial sector), and the free market income=utility premise also breaks down when some beings (animals) have stakes in outcomes of events but aren't allowed to act freely.

Hi Hutch,
thanks for prompting me to clarify my position. I do mean useful in the utilitarian sense. I'm not relying on the idea that income equals utility, or ignoring the gap between a free market stable state and an optimal state. Rather, I think that more income means more donations, and I think that donating is one of the most promising ways to do good. Personally, I'm confident that for most people, being a 'professional philanthropist' (earning a large sum of money to donate) is the life decision that can have the greatest net happiness benefit, probably by donating to outreach / organised effective philanthropy / effective philanthropy education. I think that knowing which careers earn the most per year, or per hour, is key for many young utilitarians.

RyanCarey wrote:Hi Hutch,
thanks for prompting me to clarify my position. I do mean useful in the utilitarian sense. I'm not relying on the idea that income equals utility, or ignoring the gap between a free market stable state and an optimal state. Rather, I think that more income means more donations, and I think that donating is one of the most promising ways to do good. Personally, I'm confident that for most people, being a 'professional philanthropist' (earning a large sum of money to donate) is the life decision that can have the greatest net happiness benefit, probably by donating to outreach / organised effective philanthropy / effective philanthropy education. I think that knowing which careers earn the most per year, or per hour, is key for many young utilitarians.

Oh, whoops, that makes a lot more sense. Then yes, I agree, though of course you'd have to control for competitiveness in the professions.

Also, I think that the effect of word of mouth transmition of utilitarian ideas is non-trivial; e.g. for each other rich person you convince to donate their money to utilitarian causes you've doubled your impact.

peterhurford wrote:4.) How should preferences or happiness be aggregated?

I just wrote up a blog post about average vs. aggregate utility; feel free to comment on it either here or there.

Hutch, I know how important 'word of mouth' transmission is, for instance, I wrote about it in the following article!
http://80000hours.org/blog/21-professional-philanthropy-vs-professional-influencing

RyanCarey wrote:Hutch, I know how important 'word of mouth' transmission is, for instance, I wrote about it in the following article!
http://80000hours.org/blog/21-professional-philanthropy-vs-professional-influencing

Ooh, cool article.

2.) If the adoption of utilitarianism is a matter of personal preference, why might someone want to adopt it?

Utilitarianism is practical, it also seems to be the only ethical system I know of that isn't short-sighted/based in wishful thinking.

3.) If utilitarianism is largely a choice, how do we as utilitarians get along with those of other meta-ethical persuasions?

We try to separate people that are hurt by people of other meta-ethical persuasions against their will and ignore the rest. If no one is suffering we don't need to concern ourselves with it.

5.) What matters when it comes to happiness -- total happiness or average happiness?

I think the priority should be to reduce pain. So in that sense its neither. I think securing a minimal happiness for an arbitrary percent of the population should be met first. Happy people don't really need to be happier. I don't like turning this into a math or number game because that doesn't make sense to me, but as a metaphor its like, bringing a person from 1 happiness to 2, is more important than a person from 100 to 110 because the first person had 100% increase in happiness and the second only 10%. It's also more important bringing some one from -1 to 1 than some one from 1 to 10.

I'd need a more real example to answer better.

4.) How should preferences or happiness be aggregated?

I didn't understand this one.

6.) What does happiness look like or consist of? What does more happiness look like? What kind of happiness does utilitarianism care about?

What does it consist of? Literally? Chemicals in the brain. Metaphorically? Same old stuff every ones said, it depends on the person etc. I'd think utilitarianism cares about any form of happiness that doesn't create more problems.

7.) How is happiness measured?

Several different ways. Just ask people "How happy are you on a scale from 1 to 10." or look at suicide rates etc.

8.) What do we make of the Utility Monster?

I answered this for #5 and addressed it somewhat in the other thread I posted. Again you'd need to give me a more specific situation to answer better.

9.) What do we make of the Repugnant Conclusion?

Again, my answer for #5 addresses this a bit. Not just that but I also think that utilitarians have a duty to preserve the capability for future generation to be happy. As I said in my other post I think its a given humanity will die out as some point (10 years, 100 years, 1,000 years, 1,000,000,000,000,000 years, you get the point, I don't know when). But for the sake of argument lets assume its forever, 100,000,000 people living at a time for ∞ years creates the same happiness as 10,000,000,000 at a time for ∞ years.

I didn't know what the Repugnant Conclusion was till you posed this so I read this.

It says:

Is there a moral obligation to have children?

Is it suggesting we are a moral obligation to have children because we are utilitarians to increase the number of people to be happy?

If any thing its the opposite. We have a moral obligation to not overpopulate the planet so that others can be happier. If humanity wasn't approaching the double digit billions maybe, but the way things are now no.

I've made the personal choice that I will not have children of my own, instead I will adopt. I think that overpopulation is the biggest thing preventing happiness and causing pain, it should be one of the biggest issues to utilitarians.

10.) What are some optimal actions for improving the world? What are some utilitarianism best practices?

As I just said, I think working against overpopulation is very important.
Also:
- Starvation
- Unnecessary or selfish Wars
- Disease

It seems like a stupid list being its kinda common sense to prevent those things.

As for action against them donation, volunteering, talking about it. All good. As for if utilitarians should not have kids, that's a very tricky subject that would need its own thread. On the one hand overpopulation is a massive problem, on the other if utilitarians stop having kids there will be less genetic material attracted to utilitarianism, and with less utilitarians there in theory will be less good in the world.

11.) How does one keep a utilitarian motivation going?

I remember there are people with much worse lives. Just a general sense of empathy.

12.) What can be done about existential risk? Is reducing existential risk even desirable?

Don't be stupid, reckless or selfish/greedy. Against from my other thread. The continuation of a sentient utilitarian species is only trumped by the continuation of life.

Two things here.

1) I'm assuming you are asking this to know if the survival of humanity warrants how we treat other animals. If we can determine our pleasure is greater, than its justified (nothing new, but a hard thing to do and too open to bias).
2) What's the likely-hood enough of humanity will be utilitarians (or other philosophies) to prevent unhappiness?

I'd say yes. It's desirable.

13.) What can be done about wild animal suffering?

What can be done, or what should be done? If you prevent a rabbit from being eaten by a fox you save the rabbit from suffering but cause the fox to suffer. I think humanity needs to get its own shit together first.

I'll reply to Problems in Meta-Applied Ethics tomorrow, its late. I didn't even proof-read (sorry) I'll do that tomorrow too.

Goodnight all.

Nap wrote:On the one hand overpopulation is a massive problem, on the other if utilitarians stop having kids there will be less genetic material attracted to utilitarianism, and with less utilitarians there in theory will be less good in the world.

Utilitarianism reproduces memetically, not genetically. There is no utilitarianism gene. If there are personality traits that correlate with acceptance of utilitarianism, and genes that correlate with those personality traits, the total correlation between genes and acceptance of utilitarianism may exist, but be very weak. You can of course pass on memes to your biological or adopted children's brains, but you cannot control that they stay dominant there once they become powerful adults. You would do more good to hand an attractive utilitarianism pamphlet out to many existing adolescents.

Hedonic Treader wrote:Utilitarianism reproduces memetically, not genetically. There is no utilitarianism gene. If there are personality traits that correlate with acceptance of utilitarianism, and genes that correlate with those personality traits, the total correlation between genes and acceptance of utilitarianism may exist, but be very weak. You can of course pass on memes to your biological or adopted children's brains, but you cannot control that they stay dominant there once they become powerful adults. You would do more good to hand an attractive utilitarianism pamphlet out to many existing adolescents.

If all utilitarians stopped having kids I think over time the likely-hood of more utilitarians emerging would have to go down. There would be one driving force for this natural selection and that is that utilitarians don't have kids. Without any or at least very little overlap into other traits, by that I mean the only thing determining if some one had kids or not is if they are utilitarian, I think this is much too strong of a mechanistic to be ignored for long.

We can buy ourselves some time though, because the larger a population is with frequent gene trading the harder it is for natural selection to affect the population, thus this evolution away from utilitarianism would take longer.

We have a population of over 7 billion and there are no natural borders we can't frequently cross to isolate groups of population. This means any natural selection affecting us would take considerable time to kick-in.

It's like adding yellow pain to blue paint. If there is a drop of yellow added at a time to a drop of blue, it turns to green after just 1 drop, and to green yellow after just 2. But if you are adding 1 drop of yellow paint to 7 billion drops of blue it might take a while.

I suspect that if you donate your sperm, and then use the money to pay some charity that distributes condoms in Africa, you'd end up decreasing overpopulation on the net.

DanielLC wrote:I suspect that if you donate your sperm, and then use the money to pay some charity that distributes condoms in Africa, you'd end up decreasing overpopulation on the net.

Ya. I like this idea.

You're not really contributing to population because the woman wanting the sperm would be making the decision to have a child with or without your involvement, you'd be helping to decrease the population with donations, and you wouldn't be taking away utilitarian attracted DNA from the population (maybe not take away as much at least or maybe even contributing more).

You're not really contributing to population because the woman wanting the sperm would be making the decision to have a child with or without your involvement

Not necessarily. When you sell sperm, it lowers the market price of sperm, making it so more people will buy and fewer will sell. The total increase in children born will be PED/(PED-PES) times the number of children you sell enough sperm to make, where PED is price elasticity of demand, and PES is price elasticity of supply. Note that PED is normally negative, so it's |PED|/(|PED|+PES). This is similar to the effect eating meat has on the amount of animals killed, which comes out to -PES/(PED-PES) times the amount you refrain from eating, or PES/(|PED|+PES).

I have no idea what values PES or PED have in this case, so I can't really tell you much beyond that it comes out to somewhere between zero and one.

Nap wrote:

Hedonic Treader wrote:Utilitarianism reproduces memetically, not genetically. There is no utilitarianism gene. If there are personality traits that correlate with acceptance of utilitarianism, and genes that correlate with those personality traits, the total correlation between genes and acceptance of utilitarianism may exist, but be very weak. You can of course pass on memes to your biological or adopted children's brains, but you cannot control that they stay dominant there once they become powerful adults. You would do more good to hand an attractive utilitarianism pamphlet out to many existing adolescents.

If all utilitarians stopped having kids I think over time the likely-hood of more utilitarians emerging would have to go down. There would be one driving force for this natural selection and that is that utilitarians don't have kids. Without any or at least very little overlap into other traits, by that I mean the only thing determining if some one had kids or not is if they are utilitarian, I think this is much too strong of a mechanistic to be ignored for long.

We can buy ourselves some time though, because the larger a population is with frequent gene trading the harder it is for natural selection to affect the population, thus this evolution away from utilitarianism would take longer.

We have a population of over 7 billion and there are no natural borders we can't frequently cross to isolate groups of population. This means any natural selection affecting us would take considerable time to kick-in.

It's like adding yellow pain to blue paint. If there is a drop of yellow added at a time to a drop of blue, it turns to green after just 1 drop, and to green yellow after just 2. But if you are adding 1 drop of yellow paint to 7 billion drops of blue it might take a while.

The thing is that raising kids takes time, and if your goal is to grow the number of utilitarians it might be (and probably would be) more efficient to spend the time convincing people to become utilitarians than changing a baby's diaper.

Nap wrote:utilitarian attracted DNA

Is that really a real thing?

In order to not derail the Open Problems thread, we could discuss gene selection and reproductive choices in this thread.

I posted my reply to Hutch and DanielLC in my other thread. Here.

This is my reply to Problems in Meta-Applied Ethics.

15.) Which currently existing life is sentient enough to suffer in ways relevant to a utilitarian?

I think this question needs to be put into the context of our technological evolution. If we were living in the wild and we hunted to survive I'd think this would excuse our eating habits more than they do now.

If we had artificial or grown from plants meat (that was just like "the real deal"), there wouldn't be a good excuse at all left to eat real meat (not one I can think of at least).

16.) Do nonhuman animals living in the wild live net-negative lives?

"in the wild", "unnatural/natural" in common use these words might be fine, but in a philosophical setting they don't work well. What is unnatural? Nothing can be unnatural because every thing is part of nature. Would you call a beaver damn unnatural?

If you like to think like Peter Singer than you are thinking about interests, and animals in the wild do have interest and on occasion they are met.

I think its overall best to try and separate any damage we do to their interests and let natural selection run its course (I don't mean we have to stop interacting with all animals in the "wild" I'm just saying we should try to not spill oil everywhere for example and not concern our selves with things that are not our fault, for now at least). If a species starts to show signs of higher sentience than maybe.

17.) Would a future with humans be better than a future without humans?

Nah. This is wandering into the realm of B.S. in my opinion but if this helps you here is a quick (quick, maybe or maybe not good) solution.

Again thinking in terms of interests, its every living organisms interest to spread and last over time. Humanity is biological based earthling (earthling in this context meaning all life on earth not just human) life's best hope for both. We are life's best bet to spread to other planets (or even solar systems). This is in the interest of all life and so we are the best thing life could ask for right now.

18.) What should we make of Pascalian arguments of a shot at really high utility at a really low probability, especially when the precise level of utility or probability is unknown?

I try to have the view that we should only shoot for certain improvements. High-risk high-reward situations are not appealing to me.

If you are constantly gambling, a greater happiness is uncertain. If you can ensure a constant uptrend happiness is certain. This is due to my longsighted approach to most problems. I like to think about adverse long term affects a decision might have.

peterhurford wrote:
6.) What does happiness look like or consist of? What does more happiness look like? What kind of happiness does utilitarianism care about?

My attempts at high/low pleasure vs. hedonistic, act vs. rule, and classical vs. negative utilitarianism: http://measuringshadowsblog.blogspot.com/2012/07/utilitarianism-part-3-classical-act-one.html.

Again, feel free to comment here or there.

peterhurford wrote:Utilitarian Boogeymen

These problems straddle normative ethics and applied ethics, but they're classics, so it's important to keep them in mind all in one place.

8.) What do we make of the Utility Monster?

~

9.) What do we make of the Repugnant Conclusion?

My attempt at the Repugnant Conclusion and Utility Monster here.

I for one am pretty happy to accept the Sam is Great Conclusion.

I'm surprised though that when discussing the RC and UM alongside, you don’t point out how closely related they are – each is conceptually a reductio of rejection of the other. Think having loads of just-barely-happy people sounds horrible? Then you must support condensing them into fewer more-happy-people. One entity getting all the utility is unjust? Then you presumably prefer it if we divide it into numerous proportionately-less-happy entities.

It seems you might be able to win people over with a different reductio then – if you don’t like the RP because it spreads happiness too thinly and you don’t like the UM because it concentrates it so much, you must think there’s some sort of Goldilocks proportion, where we have just the right amount of entities with just the right amount of happiness.

I’m too tired to try and fill in the dots, but you could presumably them commit people who claimed that to some fairly odd conclusions eg about adding two more happy people to the crowd being better but less than twice as good as adding one, or even worse than adding none at all.

This is a great list of questions!

>> 12.) What can be done about existential risk? Is reducing existential risk even desirable?

These days I'm largely absorbed with the 'what can be done', since I'm persuaded that this is a top priority for us. There are many xrisks and many means of reducing them. The most effective things to do depend on what specific opportunities each of us have. I generally recommend learning about the technical details of the various risks and the organizations already working on them. My own organization, the Global Catastrophic Risk Institute, is developing an organization directory to help people understand what all is out there. I'm happy to talk to people with specific interests - shoot me an email if you'd like.

5.) What matters when it comes to happiness -- total happiness or average happiness?

One random thought about average util: Why should average utilitarians calculate the average only over the amount of sentient stuff (eg. number of beings), rather than the total amount of all stuff?

Let's say you have material resources x and can use them to create either a not-too-impressed mildly happy person or a non-sentient statue (let's say for the sake of illustration that it has no consciousness but looks like a person).

VS.

Avg util usually wouldn't count the statue at all, but wouldn't it be a more accurate reflection of the universe to count it as a zero utility pattern that takes measure, thus increasing the denominator?

If we calculate the avg utility as 'avg utility per existing stuff' rather than 'avg utility per sentient stuff', the decision output should be equivalent to total util.

Nonsentient matter doesn't actually feel its neutrality. If you were going to count it, why not count stuff that doesn't exist at all?

DanielLC wrote:Nonsentient matter doesn't actually feel its neutrality. If you were going to count it, why not count stuff that doesn't exist at all?

Good question. Maybe because it doesn't exist? An unconscious being doesn't feel its neutrality either. It seems weird to have a different util calculus outcome for being unconscious vs. neutrally conscious.

I think the best argument for average utilitarianism is antrhopics. If you are randomly chosen out of all the people in the universe, then you'd want to make it more likely you're a happy person. If there's twice as many, you're half as likely to be a given one. You don't have to worry about unconscious matter, because you can't be it, but you don't want to end up as neutrally conscious matter if you can do better.