I couldn’t track down comparative data on whose bednets seem to be the most responsible—presumably it’s more whoever distributes the most nets in areas close to major fisheries, rather than % of appropriate use.
I also don’t know whether the marginal bednet will increase this kind of fishing much (there might already be a glut of bednets).
But these are all important questions that I don’t think GiveWell or AMF have ever taken seriously.
Agreed. Being able to identify effective interventions that support or protect democracy in certain contexts doesn’t necessarily seem like a bad idea.
The challenge with the AfD is that they seem to be the victims of behaviour that could be considered antidemocratic: lawmakers are considering banning the party, and the state has put the party under surveillance. This would be unconstitutional in many countries. I think there could be legitimate arguments that “protecting democracy” could sometimes involve defending groups like the AfD, as well as defending democracy from them.
I’d prefer for a politically neutral pro-democracy organisation to have the courage to defend a party like the AfD when their democratic freedoms are under threat, while also ensuring that they are stopped from damaging the democratic system. But, because this could be very messy, and have significant PR risks, I think EA-aligned groups just shouldn’t be taking that risk either way.
Thanks for this, I’m trying to at least get a couple of numbers to get in the right ballpark to help people calibrate. Epistemic status: uncertain, a couple of hours of research. Takeaway: Probably billions of fish killed.
To get a higher end of the estimate of total fish killed as a result of mosquito net fishing, the appendix of the paper you link to gives an example of a single large-scale fishery in Madagascar where 68.7 million fish were killed in 2018-2019 by 75 families.
Unless I’m misreading, this looks like 80-90% of the total fish numbers were caught through mosquito net trawler (27.7% by weight, but much smaller fish on average, so a higher number).
This other paper in Mozambique provides another example—they two fishers catch around 35kg of (mostly tiny, < 20 mm, so 1-10g?) fish a day, so I guess around 10,000 fish a day? Over 2 million a year (so similar ballpark to Madagascar).
These are the locations at which mosquito net fishing had been observed in this study (data mostly from 2015), with a rapidly rising trajectory.
I can’t seem to find any better data on the scale of fishing at all of these locations, so I have no idea whether it’s closer to 50 or 10000 locations where mosquito nets are used for fishing to a similar extent to those in these articles, but 25 million fish * 50 locations would be 1.25 billion fish.
I think it’s safe to assume that at least 1 billion fish, possibly orders of magnitude higher, are killed yearly through mosquito net fishing, most of whom wouldn’t be killed if it weren’t for the distribution of mosquito nets (due to the small mesh sizes).
Of course, to make a reasonable welfare calculation, you shouldn’t just look at the number of animals killed and how painful their deaths must have been. You also have to consider how long these fish would have lived for counterfactually, the quality of their lives and suffering involved in their counterfactual deaths, as well as broader ecosystem effects etc.
But @Dylan Matthews’s claim that “There is little research on what fishing with these nets actually does to fish or people — but also little reason to think the magnitudes of these effects are remotely near the number of lives saved by nets” strikes me as indefensible.
Causing billions of fish to experience a painful death is not just a rounding error, and it may well be worth reevaluating whether bednets have had a positive net impact or not.
I agree with your general case, and I’m interested in the role that genetics can play in improving educational and socio-economic outcomes across the world. In the case of a world where biological intelligence remains relevant (not my default scenario, but plausible), this will become an increasingly interesting question.
However, I’m unconvinced that an EA should want to invest in any of the suggested donation interventions at the minute—they seem to be examples where existing research and market incentives would probably be sufficient. I’m not sure that more charitable support would have a strong counterfactual impact at the margins. (Note: I know very little about funding for genetics research—it seems expensive and already quite well-funded, but please correct me if I’m wrong here).
In terms of whether we should promote/ talk about it more, I think EA has limited “controversy points” that should be used sparingly for high-impact cause areas or interventions. I don’t feel that improving NIQ through genetic interventions scores well on the “EV vs. controversy” trade-off. There are other genetic enhancement interventions (e.g. reducing extreme suffering, in humans or farmed animals), that seem to give you more EV for less controversy.
Also, if we do make this case, I think that mentioning Lynn/ Vanhanen is probably unwise, and that you could make the case equally well without the more controversial figures/ references.
Finally, I’d like to see a plausible pathway or theory of change for a more explicitly EA-framed case for genetic enhancement. For example, we expect this technology to develop anyway, but people with an EA framing could:
1. Promote the use of embryo screening to avert strongly negative cognitive outcomes
2. If this technology is proven to be cost-effective in rich countries, remove barriers to rolling it out in countries where it could have a greater counterfactual impact
At risk of compounding the hypocrisy here, criticizing a comment for being abrasive and arrogant while also saying: “Your ideas are neither insightful nor thoughtful, just Google/ ChatGpt it” might be showing a lack of self-awareness...
But agreed that this post is probably not the best place for an argument on the feasibility of a pause, especially as the post mentions that David M will make the case for how a pause would work as part of the debate. If your concerns aren’t addressed there, Gerald, that post will probably be a better place to discuss.
My argument questions the ideas of lives saved, DALYs and QALYs as metrics—just like using lives saved as a metric, QALYs generally implicitly assume that death is worse than a very bad life, no matter the levels of mental suffering, pain, and physical debilitation.
I’m probably criticising GiveWell’s methods as much as the post- their methodology assumes that the value of saving lives/ averting deaths is positive.
I generally agree more with HLI’s ‘WELLBY’ approach, as long as negative WELLBYs are taken seriously.
As I said on my final para, I do see global health interventions as probably being net positive, despite potentially saving more net negative lives, so my argument definitely wasn’t to “defund GiveWell”. It was more that “saving lives” is a bad metric and bad thing to feel good about.
My cherry picking of negative phenomena was in response to the cherry picking of the original post. I think boring/ useless school (I didn’t quote anything but… most African rural schools are boring and useless...), unpleasant labour, hunger/ stunting and poor mental health are very relevant variables, as they define a lot of the waking hours of the poorest people in the world.
FGM and child marriage are probably less representative of general welfare—I was responding to the “first kiss” idea in the post.
I chose Burkina Faso at random. For central African countries I might have stressed sexual violence, which seems to be lower in Burkina Faso.
Thanks for responding.
I accept your point about life satisfaction vs happiness measures not being equivalent. But if GiveWell recipients think that their life is significantly closer to the worst possible life than the best possible life, this still makes my point pretty well. Doesn’t seem obvious to judge the net welfare of someone who is, say, 3⁄10 for life satisfaction and ‘rather happy’. I haven’t seen good studies on GiveWell recipients’ happiness or moment-to-moment well being (using ESM etc.), or other ways of measuring what we care about, but I would appreciate better info on that.
My (implicit) estimates for child marriage, stunting and mental illness should be adjusted for the fact that average GiveWell charity recipients in Burkina Faso have worse lives than the average citizen, but I acknowledge my language was imprecise. Stunting might plausibly cross the 50% threshold in that category, but might be under. The median marriage age for Burkinabe girls is 17, and is probably lower in the GiveWell pop. Some orgs define child marriage as <18.
Mental illness thresholds seem to vary a lot, but this https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0164790 article is a good example of how bad mental health is for ‘ultra-poor’ kids in Burkina Faso. My thinking would be that 20-30% of the kids in this study have lives clearly on the net-negative side, which I think would be unlikely to be outweighed by the more neutral/ positive lives. Don’t know exactly how this would match with a typical GiveWell population.
“To answer your comment “you have to work out whether you think this life you’ve saved is more likely or not to be net positive. ”—We have worked it out, and the answer YES, a resounding yes”
I consider this just obviously false. I just don’t believe that you/ global health people have disproven negative-leaning utilitarian or suffering-focused ethical stances. You might have come to a tentative conclusion based on a specific ethical framework, limited evidence and personal intuitions (as I have).
I’d say that there’s probably a fairly fundamental uncertainty about whether any lives are net positive. There’s definitely not a consensus within the EA community or elsewhere. It depends on stuff like suffering happiness assymetry and the extent to which you think pain and pleasure are logarithmic (https://qri.org/blog/log-scales).
Most of us will acknowledge that at least some lives are net negative, some extremely so, and that these lives are far more likely to be saved by GiveWell charities. I suspect any attempt to model exactly where to draw the line will be very sensitive to subtle differences in assumptions, but my current model leans towards the average GiveWell life being net negative in the medium term, for the reasons I’ve mentioned.
In terms of language, I think “great care and dignity” are suitable for most contexts, but I think that it’s important that the EA forum is a safe space for blunt language on this topic.
I’m increasingly convinced that EA needs to distance ourselves from this framing of “saving lives=good”. And we need to avoid the satisfying illusion that giving to global health charities is saving a life “just like our own”. (Particularly disagreeing with @NickLaing’s comments here). If you’ve decided to save the lives of the ultra-poor, you should be able to bite the bullet and admit you’re doing that.
We all like the idea of saving a kid who’s “playing with their friends in the schoolyard, maybe spending time with their grandma, or maybe just kicking a football, alone”, and ” celebrating her birthday” and the “first kiss”.
But you don’t need to be a negative utilitarian to recognise that the kid whose life you’ve saved probably isn’t having a great life- mostly for the reasons you donated to that charity- it’s shitty being a poor person in the poorest countries in the world.
Let’s say you saved a life in Burkina Faso:
- If you saved a girl, she’ll probably be a victim of FGM, and get married as a child to an older man—that “first kiss” you mention might be as a 15-year-old girl with her 30-year-old husband
- If they don’t go to school, they’ll do hard and dangerous work in agriculture, fishing or worseas children.
- If they do go to school, they’ll probably spend their days in extreme boredom, getting left behind, and end their school years functionally illiterate and innumerate
- They’re likely to spend a lot of their childhood hungry—they will get ill often, with malaria, diarrhea, or other communicable diseases
- They will be likely to grow up stunted or wasted, and with diminished cognitive abilities
- All of the above tend not to be great for mental health, so they’re fairly likely to become depressed, anxious, or suffer from more serious mental issues
- When you ask them how happy they are on a life satisfaction/ happiness scale, they’ll give you around 4⁄10
Based on this reality, and your estimates about how the world is likely to improve in the coming few decades, you have to work out whether you think this life you’ve saved is more likely or not to be net positive.
I’m not saying that we shouldn’t give more money to global health charities- they improve lives and stop people getting horrible diseases. All else equal, fewer communicable diseases are better. But I’m disagreeing strongly with the framing of this piece.
Interesting—I definitely think this is valuable. I have two small recommendations for the survey:
- Specify in the sugary drinks question whether it only includes commercial, fizzy sugary drinks, or any drinks with sugar in (e.g. coffee with sugar, milkshakes, bubble tea, traditional sweet drinks etc.) As it is, you give examples of commercial fizzy drinks, but it’s a little ambiguous whether other sweet drinks might be included.
- Make it clear that you can choose percentages over 100% for the first two options (a life in prison, or without any pleasure is worse than death—many people are likely to believe this). I think that your example percentages (e.g. 0.1%, 1%, 10%, 20%, 30%, 100% etc) are anchoring people to a particularly low score.
Interesting. I think there are two related concepts here, which I’ll call individual modesty and communal modesty. Individual modesty, meaning that an individual would defer to the perceived experts (potentially within his community) and communal modesty, meaning that the community defers to the relevant external expert opinion. I think EAs tend to have fairly strong individual modesty, but occasionally our communal modesty lets us down.
With most issues that EAs are likely to have strong opinions on, here are a few of my observations:
1. Ethics: I’d guess that most individual EAs think they’re right about the fundamentals- that consequentialism is just better than the alternatives. I’m not sure whether this is more communal or individual immodesty.
2. Economics/ Poverty: I think EAs tend to defer to smart external economists who understand poverty better than core EAs, but are less modest when it comes to what we should prioritise based on expert understanding.
3. Effective Giving: Individuals tend to defer to a communal consensus. We’re the relevant experts here, I think.
4. General forecasting/ Future: Individuals tend to defer to a communal consensus. We think the relevant class is within our community, so we have low communal modesty.
5. Animals: We probably defer to our own intuitions more than we should. Or Brian Tomasik. If you’re anything like me, you think: “he’s probably right, but I don’t really want to think about it”.
6. Geopolitics: I think that we’re particularly bad at communal modesty here—I hear lots of bad memes (especially about China) that seem to be fairly badly informed. But it’s also difficult to work out the relevant expert reference class.
7. AI (doom): Individuals tend to defer to a communal consensus, but tend to lean towards core EA’s 3-20% rather than core-LW/Eliezer’s 99+%. People broadly within our community (EA/ rationalists) genuinely have thought about this issue more than anyone else, but I think there’s a debate whether we should defer to our pet experts or more establishment AI people.
I think there’s a range of things that could happen with lower-level AGI, with increasing levels of ‘fire-alarm-ness’ (1-4), but decreasing levels of likelihood. Here’s a list; my (very tentative) model would be that I expect lots of 1s and a few 2s within my default scenario, and this will be enough to slow down the process and make our trajectory slightly less dangerous.
Forgive the vagueness, but these are the kind of things I have in mind:
1. Mild fire alarm:
- Hacking (prompt injections?) within current realms of possibility (but amped up a bit)
- Human manipulation within current realms of possibility (IRA disinformation *5)
- Visible, unexpected self-improvement/ escape (without severe harm)
- Any lethal autonomous weapon use (even if generally aligned) especially by rogue power
- Everyday tech (phones, vehicles, online platforms) doing crazy, but benign misaligned stuff
- Stock market manipulation causing important people to lose a lot of money
2. Moderate fire alarm:
- Hacking beyond current levels of possibility
- Extreme mass manipulation
- Collapsing financial or governance systems causing minor financial or political crisis
- Deadly use of autonomous AGI in weapons systems by rogue group (killing over 1000 people)
- Misaligned, but less deadly, use in weapons systems
- Unexpected self-improvement/ escape of a system causing multiple casualties/ other chaos
- Attempted (thwarted) acquisition of WMDs/ biological weapons
- Unsuccessful (but visible) attempts to seize political power
3. Major fire alarm:
- Successful attempts to seize political power
- Effective global mass manipulation
- Successful acquisition of WMDs, bioweapons
- Complete financial collapse—
Complete destruction of online systems- internet becomes unuseable etc.
- Misaligned, very deadly use in weapons systems
4. The fire alarm has been destroyed, so now it’s just some guy hitting a rock with a scorched fencepost:
- Actual triggering of nuclear/ bio conflict/ other genuine civilisational collapse scenario (destroying AI in the process)
Okay, I think your reference to infinite time periods isn’t particularly relevant here (seems to be a massive difference between 5 and 20 years), but I get your point that short timelines play an important role.
I guess the relevant factors that might be where we have different intuitions are:
How long will this post-agentic-AGI, pre-God-AGI phase last?
How chaotic/ dangerous will it be?
When bad stuff happens, how likely is it to seriously alter the situation? (e.g. pause in AI progress, massive increase in alignment research, major compute limitations, massive reduction on global scientific capacity etc.)
If you hold these assumptions robustly, the most direct answer would be to focus on the kind of beings who are likely to experience greater suffering by default, namely factory-farmed animals, and potentially some wild animals. You should focus on interventions (alternative proteins, vegan advocacy) that are likely to cause these animals not to come into existence, rather than welfarist approaches that improve the lives of animals but keep numbers relatively constant. This is a very popular approach, so you’d be welcome in this part of the EA space.
But you might want to slightly relax your assumptions slightly when considering practical work you could do. Assuming that reducing suffering is your ultimate goal, if the “best way not to suffer is not to live”, it doesn’t necessarily follow that the most effective way to reduce suffering (given limited resources) is stopping beings coming into existence.
For example, an intervention to help people in poor countries detect particularly painful congenital defects before birth and terminate these pregnancies might reduce suffering and satisfy your assumptions, but if it’s expensive, it might be more effective to reduce the suffering of existing people, for example, providing relatively cheap pain relief for people with late-stage cancer.
Or if you could cause x number of factory-farmed chickens to be raised in a free-range/ organic way for the same cost/ resources as stopping y number of factory-farmed chickens being born, there’s probably some number for x and y for which you’d choose the first option.
Rambling question here. What’s the standard response to the idea that very bad things are likely to happen with non-existential AGI before worse things happen with extinction-level AGI?
Eliezee dismissed this as unlikely “what, self-driving cars crashing into each other?”, and I read his “There is no fire alarm” piece, but I’m unconvinced.
For example, we can imagine a range of self-improving, kinda agentic AGIs, from some kind of crappy ChaosGPT let loose online, to a perfect God-level superintelligence optimising for something weird and alien, but perfectly able to function in, conceal itself in and manipulate human systems.
It seems intuitively more likely we’ll develop many of the crappy ones first (seems to already be happening). And that they’ll be dangerous.
I can imagine flawed, agentic, and superficially self-improving AI systems going crazy online, crashing financial systems, hacking military and biosecurity, taking a shot at mass manipulation, but ultimately failing to displace humanity, perhaps because they fail to operate in analog human systems, perhaps because they’re just not that good.
Optimistically, these crappy AIs might function as a warning shot/ fire alarm. Everyone gets terrified, realises we’re creating demons, and we’re in a different world with regards to AI alignment.
I sympathise, but… For 1), if your negative utilitarianism (NU) is a sincerely held, ‘psychologically normal’ belief, I think that you can be a very strong NU and still want to pursue totally ‘normal’ EA goals. For any brand of utilitarianism, greatly reducing or eradicating suffering is a valid and obviously normal goal. Assuming you don’t have the capability for ‘magic annihilation’, there are so many alternatives. Is there a worldview where ‘ending your own suffering’ is higher expected-value than ending factory farming or treating extreme cancer pain in Sub-Saharan Africa? Preventing S-risks is also a productive way for an NU to work on/ think about an EA topic.
I feel that strong negative utilitarianism (we should only consider disutility) is just a non-starter. It doesn’t match any of my moral intuitions.
But a weaker negative utilitarianism is a powerful and potentially valid position. These are my views:
Good things are actually good. Pleasure is usually good. Laughing/ smiling/ dancing/ sex/ rollercoasters/ MDMA trips are actually really good.
Bad things might be a bit worse than good things are good (at least our intuitions might be skewed). But a rudimentary thought experiment can calibrate your scales here—I don’t see how you can end up at a strong negative position.
There is a hedonic treadmill/ normalisation effect, making ‘normality’ into suffering, but this is not always the case—some hedonistic pleasure can give you a warm glow for a long time, and make you genuinely happier in the long term.
We probably have a weak bias towards pleasure and suffering in our lives being more balanced than they are, and towards believing that our lives are more positive than they actually are. E.g. someone can say that his life is 6⁄10 on a happiness scale, but he could be totally wrong, and his life could actually not be worth living.
This bias is unlikely to be so strong that even the happiest-seeming people actually have net-negative lives. So there are almost definitely some net-positive lives in the world.
Lots of human lives are net-negative. My median estimate would be 25% (around 5⁄10 on a happiness scale), but I have very wide error margins.
This is just a rounding error compared to animal lives, which are more likely to be net-negative, therefore the world is likely to be net-negative, even from a standard utilitarian position.
Even under very weak negative utilitarianism, the ‘benevolent world-exploder’ argument may be valid from a near-termist view, and is compatible with fear of s-risks. But, even to a moderate weak negative utilitarian, it can be countered with an optimistic long-termism where we may be able to both eradicate suffering and normalise increasingly wondrous states of pleasure.
Some random thoughts:
Whenever I hear someone (usually Will Macaskill) say: “If utilitarianism is correct...” I don’t hear it as a coherent sentence. I don’t think there is a fundamental ethical truth that utilitarianism is trying to get at. I’d still happily call myself a utilitarian, as I believe it’s the best, most coherent way of thinking about ethics- I even ‘bite the bullet’ of the repugnant conclusion. But utilitarianism doesn’t feel like the kind of thing that can be ‘true’ or ‘false’.
But I feel I agree with the ‘realist’ side of this paper more than the ‘anti-realist’. When I’m ‘doing’ ethics, I’d rather do it well and consistently. I tend to slightly disagree with Scott Alexander in that post.
But ethics, especially utilitarianism, is so demanding that you need a way to ‘get out of it’ - I do that by just ‘not doing ethics’ - I would be a ‘better person’ if I did ethics more consistently, but I just don’t—I also spend most of my time not even considering utilitarian ethics. I should but don’t need to be ethical. I’m bound by utilitarian ethics in a similar way that I’m bound by my own bodily desires, politeness norms, family ties, occasional propensity to turn into a dire wolf etc.
But I don’t think it works in the Scott Alexander/ lizard/ utopia case. If there’s a time to take ethics deadly seriously, it’s when you’ve got a zillion potentially net-positive lizard lives on the scales!
Yep, agreed (I haven’t read those books, but I broadly know the story). I wasn’t trying to imply that eugenics was the main cause of the one-child policy, but the two are definitely connected. Post-1CP, the state took a really active role in controlling how and when kids were born, compulsory sterilisation (mostly of females, despite vasectomies being safer) became normalised for ‘quality and quantity’ etc.
A stronger “eugenics taboo” could plausibly have limited the scope of the policy.
No, there is no way to be confident.
I think humanity is intellectually on a trajectory towards greater concern for non-human animals. But this is not a reliable argument. Trajectories can reverse or stall, and most of the world is likely to remain, at best, indifferent to and complicit in the increasing suffering of farmed animals for decades to come. We could easily “lock in” our (fairly horrific) modern norms.
But I think we should probably still lean towards preventing human extinction.
The main reason for this is the pursuit of convergent goals.
It’s just way harder to integrate pro-extinction actions into the other things that we care about and are trying to do as a movement.
We care about making people and animals healthier and happier, avoiding mass suffering events / pandemics / global conflict, improving global institutions, and pursuing moral progress. There are many actions that can improve these metrics—reducing pandemic risk, making AI safer, supporting global development, preventing great power conflict—which also tend to reduce extinction risk. But there are very few things we can do that improve these metrics while increasing x-risk.
Even if extinction itself would be positive expected value, trying to make humans go extinct is a bit all-or-nothing, and you probably won’t ever be presented with a choice where x-risk is the only variable at play. Most of the things you can do that increase human x-risk at the margins also probably increase the chance of other bad things happening. This means that there are very few actions that you could take with a view towards increasing x-risk that are positive expected value.
I know this is hardly a rousing argument to inspire you in your career in biorisk, but I think it should at least help you guard against taking a stronger pro-extinction view.