Ways to Reduce Suffering

Summary. I give a list of ideas for things people might do to prevent suffering. These range from highly speculative, high-expected-payoff actions (e.g., preventing people from spending eternity in hell, or opposing the creation of infinitely many new universes in a laboratory) to more immediate, down-to-earth approaches (e.g., advocating efficiency in addressing Third-world diseases, or promoting use of humane insecticides). My current recommendation of the single best option for utilitarian donors is the Singularity Institute for Artificial Intelligence.

Introduction

To some extent, the current page is a summary of ideas that are explained in more detail elsewhere on this site. I'm not currently in a position to work on many of the below proposals, but I hope to tackle some of them in the future, when I'm older, wiser, and have more resources. At the moment, it seems like a safe bet for me to make and invest income that I can later use to fund projects like these (or else that I can live off of while working on them directly). Many readers will likely have suggestions or criticisms, which I welcome: <webmaster ["at"] utilitarian-essays.com>. Please also let me know if you might be interested in helping actually to carry out any of these ideas.

Infinite Suffering

Physics and cosmology

If we live in an infinite universe, there may exist infinitely many suffering organisms. Various levels of multiverses would contain various degrees of infinite suffering. Physicists' theories of quantum mechanics, time travel, inflationary cosmology, string theory, and the rest are thus highly relevant in assessing the amount of suffering in the universe.

Are there ways to exploit the laws of physics to prevent infinite (or even large finite) amounts of suffering? This seems an urgent question for research.

Hell

If hell exists, then billions of people may spend eternity experiencing some of the worst torments possible. It would be interesting to brainstorm the space of possible hell scenarios and weigh their relative probabilities. Part of this could include a systematic evaluation of the probabilities of various religions (as well as the probabilities of eternal hells conditional on those religions being true). This could be aided by a friendly artificial intelligence applying disciplined Bayesian reasoning, but even without AI, it ought to be possible to get some good back-of-the-envelope comparisons. I can imagine some religions coming out many orders of magnitude more probable than others, even just based on a few considerations. A systematic analysis of this type seems relatively straightforward, yet I haven't seen it done (except perhaps by a few authors who have decided in advance what answer they want to reach).

Lab universes

I assign nontrivial probability to humans (or their descendants) creating infinitely many new universes in a laboratory. Apart from causing infinitely many new replications of the Holocaust, this would create infinitely many animals that would suffer in the wild (as well as perhaps infinitely many people who could go to hell).

Ways to avert this scenario aren't obvious. Even if lab universes are possible, they'll probably require more advanced technology than humans are likely to have in the near future, so that immediate prohibitions wouldn't really help. Moreover, it's unclear whether publicizing the issue is even a good idea, since many people are fascinated by the idea of making new universes and might be inspired to support such efforts once they heard about them.

Infinity in general

Applying the framework of maximizing expected value to the case of infinite outcomes doesn't yet have a widely accepted foundation. Basic research on this problem might shed light on how to prioritize various scenarios involving infinite suffering. (Finally, a practical application of higher-level infinite-set theory!) For instance, should we allow for the possibility that even a single moment in hell could be infinitely worse than, say, countably many moments of torture on earth? On the other hand, it remains to be seen whether this type of research will ultimately have much impact. For the most part, any infinity-prioritization scheme tends to boil down mainly to an expression of one's pre-existing intuitions.

An Institute for Research on Infinite Suffering

But in the mean time, there remains the task of systematically investigating ideas related to infinite suffering, evaluating their plausibility, and suggesting avenues for further study. In addition, it would be helpful simply to raise the issue of infinite suffering into people's consciousness as a legitimate topic of concern and discourse. I would welcome something like an Institute for Research on Infinite Suffering that could begin to tackle these challenges. Plus, there's already lots of great work being done by very smart individuals--in philosophy, physics, mathematics, and other fields. Just a little nudging as far as choice of research topics could produce fruitful results. I think SIAI is perhaps the best organization right now working on these problems, and I recommend that readers consider donating there.

Finite Suffering--Animals

Wild animals

There are enormous numbers of animals in the wild enduring terrible pain, distress, hunger, and illness. It seems plausible that there are cost-effective ways of relieving some of this suffering. (Any ideas?) If so, animal-welfare organizations could help far more animals than they currently do--even just counting the higher-level species like mammals that they tend to focus on most--by shifting resources toward such projects.

It seems that many people are not significantly troubled by the suffering of wild animals, and those who are often see the problem as just one of those unhappy facts about the world. Perhaps there would be value in publicizing (say, through a film) the seriousness of the problem and the idea that humans might actually be able to improve the situation.

Insects

There's some chance that insects can feel pain, and--considering that the world supports 10^18 of them--the expected value of their suffering is enormous.

While perhaps not the most cost-effective option, one way to relieve expected insect suffering would be to promote the use of humane insecticides (i.e., insecticides that cause death more quickly and gently than conventional versions). If nothing else, a project like this would attract media attention while simultaneously demonstrating that insect suffering is not a completely intractable issue.

Of course, it's currently unknown whether, or to what extent, insects can feel pain. The topic has been examined by a few studies, but I would guess that the marginal benefit of research on the question remains large.

Animal-welfare research

There is lots of excellent academic research by biologists, veterinarians, and animal-welfare scientists on questions of crucial importance: Which animals can feel pain? What are signs and causes of distress? How can pain be managed and prevented? The benefits of such work are twofold. First, the information itself is important for determining how much different animals suffer in various circumstances so that activists can best target their resources to areas of greatest need. Second, the framework of animal-welfare researchers is highly utilitarian: Where is there lots of suffering, and how can we cost-effectively reduce it? The more people apply this mindset to laboratory and farm animals, perhaps the more they will also apply it to, say, wild animals and humans. Most people have few reservations taking a thoroughly hedonist approach to animal pain and pleasure; the more they do so, perhaps the more they will see that the same applies to Homo sapiens.

How can utilitarians best support such research? Should they fund promising animal-welfare scientists directly? Or lobby for more government and industry funding of research? Or maybe they should focus on putting the findings of the academic world into practice, or popularizing the material to a wider audience?

Indirect Approaches

Promoting cost-effectiveness

Even according to very limited, conventional metrics like US$/DALY, different Third-world public-health interventions vary by orders of magnitude in cost-effectiveness. The same is true to an even greater extent across charities generally. Since people are often less concerned with the efficiency of their donations than with simply feeling like they're doing something to help, there's wide room for those who are more conscious of the importance of cost-effectiveness to help others choose better charities.^[1]

A related approach is to target one's donations toward the subset of activities of charitable organizations that have highest expected returns. In particular, there may be cases in which fundraising has enormous expected value but charities are reluctant to devote too many resources toward it out of concern for the impression it will give donors regarding how their funds are being used. One could ask a charity to use all of one's donation on the dirty work jobs that ordinary donors who are more concerned about feeling good about what they've accomplished directly would be reluctant to support. (In going about this, it would be important to ensure that, by your donating more toward fundraising, you're not just allowing the charity to spend less of other donations on it. Rather, the point is to increase the total size of the fundraising budget. In addition, one should consider whether better fundraising for one charity simply means people will end up donating less to other charities.)

Of course, even utilitarian wannabes are not immune from the natural temptation to seek warm fuzzies over cost-effectiveness; we are still humans with tribal psychology, after all. Eliezer Yudkowsky proposes an interesting way to address this: Purchase Fuzzies and Utilons Separately:

the main lesson is that all three of these things -- warm fuzzies, status, and expected utilons -- can be bought far more efficiently when you buy separately, optimizing for only one thing at a time. Writing a check for $10,000,000 to a breast-cancer charity -- while far more laudable than spending the same $10,000,000 on, I don't know, parties or something -- won't give you the concentrated euphoria of being present in person when you turn a single human's life around, probably not anywhere close. [...] And if you threw away all concern for warm fuzzies and status, there are probably at least a thousand underserved existing charities that could produce orders of magnitude more utilons with ten million dollars. Trying to optimize for all three criteria in one go only ensures that none of them end up optimized very well -- just vague pushes along all three dimensions.

Career choice

Eliezer's point about trying to optimize two criteria simultaneously potentially optimizing neither may ring true here as well: Rather than seeking some way to incorporate doing good into your professional career, or purchasing habits, or investments, you might do better to treat those as business matters, and save your time and money to work on charity separately.

Exploiting probability: covariance and risk premia

In addition to taking advantage of covariance and risk premia individually, one might encourage others (perhaps large foundations?) to try similar strategies.

[1] Readers might point out some hypocrisy on my part in this regard: If cost-effectiveness is so important, why are you bothering to think about finite amounts of animal suffering on one planet, when there are things that could potentially cause infinite harm? While one could argue that raising awareness about wild-animal suffering might have infinite expected benefit to the extent that it shapes the way humans and their descendants act (as regards lab universes, among other things), I acknowledge that this criticism is completely valid. (...Sigh.)

It's worth noting that many religions enjoin their adherents to perform acts of charity, which often means giving time and money toward relieving finite amounts of (usually human) suffering. However, the focus is often on one's neighbors and community, and the mindset is often decidedly non-maximizing.