There are large fundamental differences between different forms of utilitarianism, but as far as I can tell, it always comes down to donating everything you can to charity. I'm sure the best charity would change with even the subtlest difference, but it wouldn't be obvious. What theories would be different to the point that you can tell that a different class of charities would be the best?

Some utilitarians believe that infinite utility is possible, and no matter how unlikely it is, you should work towards it. Others believe it is somehow not worth striving for. The first group is likely to care more about preventing existential dangers than the second, and thus a charity on nuclear disarmament would be much more likely to be the bast one.

This infinite utility idea is very reminiscent of Pascal's Wager, actually.

Of course, Pascal's Wager says that we devote ourselves to following religious teaching because if a god exists, we will be rewarded with infinite happiness. And a probability of infinite happiness, no matter how small, is superior to the small amount of happiness that might be obtainable by disobeying god.

I respond: but surely if a god existed, he would not present ancient superstition as fact. (e.g. flat world hypothesis). Nor would he suggest ancient ethical systems (disobey me and I will stone you). A more likely avenue to infinite utility would be to obey a more likely god: one who understands present and future science and rewards those who implement modern ethical systems such as utilitarianism.

That diversion took a little longer than I would have liked. On topic, I've never encountered an argument for achieving infinite utility through a means other than appeasing a god. Infinite utility seems impossible to me given that the world has limited matter and limited energy. But I suppose we don't really know yet how consciousness comes and it's impossible to be perfectly certain about a thing like that anyway. So perhaps I should give greater emphasis to preventing existential dangers?

Regarding your idea of donating everything you can to charity, I agree with your sentiment, but I don't think that donating everything I earn to charity is the best way to help people. For example, if I earn enough to raise five children, this allows them to earn money to donate a greater sum to charity. So surely I should invest some resources towards finding a partner? For various reasons, there may be more utility in raising biological children with a partner.

It isn't given that there is a limited amount of energy in the universe and entropy will always increase. What is given is evidence that points to this. There is some evidence that it's wrong. For example, the law of conservation of energy does not hold in general relativity. If we find either to be wrong, there is a significant chance of infinite utility.

If you raise children to help increase utility, would it count as charity?

Well that I can answer: who cares whether it counts as charity! It does not involve funds being distributed among a charitable organisation, but it can involve kindess and generosity and it does involve contributing resources.

You can frame it as a charity or as selfishness, but what matters is that raising kind children has immensely positive consequences.

Anyway, the point is the correct way to use your money varies little between theories. Having children will be about as good an idea no matter what kind of utilitarian you are.

RyanCarey wrote:This infinite utility idea is very reminiscent of Pascal's Wager, actually.

Of course, Pascal's Wager says that we devote ourselves to following religious teaching because if a god exists, we will be rewarded with infinite happiness. And a probability of infinite happiness, no matter how small, is superior to the small amount of happiness that might be obtainable by disobeying god.

I respond: but surely if a god existed, he would not present ancient superstition as fact. (e.g. flat world hypothesis). Nor would he suggest ancient ethical systems (disobey me and I will stone you). A more likely avenue to infinite utility would be to obey a more likely god: one who understands present and future science and rewards those who implement modern ethical systems such as utilitarianism.

Have you come across Alan Dawrst? He's a utilitarian who argues (among other things) that utilitarianism should lead us to accept Pascal's Wager. He argues specifically for Christianity, on the grounds that its hell and heaven are (supposedly) defined as the greatest infinitude of suffering/pleasure, and that more people accept it than any other religion.

His argument's key failing, IMO, is the assumption that someone believing something ipso facto increases the probability that that thing is true.

Alan's registered here, though he hasn't posted yet, so maybe he'll hop in to this discussion...

Certainly if God was omnibenevolent and omnipotent, which are both central to Christianity, utility would be positive infinite no matter what you do. As such, if you're going to use pascal's wager, it would have to be for a religion with an omnipotent, but not omnibenevolent god.

If there are aleph null QALYs of positive utility, and aleph null QALYs of negative utility, wouldn't they be necessarily equal? Of course, it's theoretically possible to have infinite positive utility and only finite negative utility. For example, if positive utility remained constant and negative utility decreased exponentially with time.

His argument's key failing, IMO, is the assumption that someone believing something ipso facto increases the probability that that thing is true.

I wouldn't say it's ipso facto. Rather, in most cases the fact that people believe X is Bayesian evidence for X, because "lots of people believe X" is a fact about the world that needs explaining, and the explanation "people believe X because they discovered that it's true" is relatively clean and compact. But there are certainly exceptions to this general principle. And there are definitely other factors that count against Christianity, including some Biblical suggestions of the concept of a flat earth as RyanCarey mentioned.

Alan Dawrst wrote:

His argument's key failing, IMO, is the assumption that someone believing something ipso facto increases the probability that that thing is true.

I wouldn't say it's ipso facto. Rather, in most cases the fact that people believe X is Bayesian evidence for X, because "lots of people believe X" is a fact about the world that needs explaining, and the explanation "people believe X because they discovered that it's true" is relatively clean and compact.

There is a difference between believing something causing it to be true and something being true causing people to believe it.

Alan Dawrst wrote:

Certainly if God was omnibenevolent and omnipotent, which are both central to Christianity, utility would be positive infinite no matter what you do. As such, if you're going to use pascal's wager, it would have to be for a religion with an omnipotent, but not omnibenevolent god.

I agree. God is almost certainly not a utilitarian; the problem of evil, for instance, makes that implausible. A sometimes angry and occasionally violent God like Yahweh of the Old Testament seems more likely. Moreover, even if we thought God was a utilitarian, it would be almost impossible to make a Pascalian hell-based argument for being in him because a utilitarian God wouldn't have a hell.

You misunderstood me. I was assuming God was Utilitarian. He could still cause negative utility, so long as he balances it out. If he is not a Utilitarian, Pascal's argument would sort of work. This would, however, leave you with the problem of believing isn't omniscient, isn't omnibenevolent, or both. Unfortunately, I have since realized that my method of doing infinite utility was flawed. Specifically, I was adding all the utility at a specific moment together, then taking the total of them as a transfinite cardinal number. As such, ignore my argument. If infinite utility works, you could probably increase it after it's infinite.

Alan Dawrst wrote:Yes, that's (formally) true. Similarly, if there are aleph-null positive QALYs, then any finite (or even aleph-null-sized infinite) increase in QALYs "makes no difference". Of course, Cantor's cardinal numbers aren't the only way to represent infinity. Nick Bostrom's excellent "Infinite Ethics" discusses a number of options, including hyperreal numbers:

http://www.nickbostrom.com/ethics/infinite.pdf

I'd really like to see more research on this topic.

I am an Eternalist, and all other forms of infinite utility violate Eternalism. Changing the order of events can change the total utility. Of course, this is another topic for another thread, which I coincidentally posted around the same time as this one.

I suspect if you did hyperreal numbers, the ultrafilter used would matter, but I never learned to integrate with hyperreals so I don't know. I'd use limits, which would lead to universes where neither is better than the other, and they're not equal either, but that isn't really worse than which is better depending on which ultrafilter you use. I think that would be the only difference between those two methods, so they're pretty much equivalent.