Thanks for the kind words. :)

things like the dark tetrad traits (narcissism, machiavellianism, psychopathy, sadism) are adaptive even on a group level

Yup. And how adaptive they are depends on the distribution of other agent types. For example, against a population of pure pacifists, Dark Tetrad traits may be pretty effective. In a population of agents who cooperate with one another but punish rule-breakers, Dark Tetrad traits are probably less adaptive. Hopefully present-day society is somewhat close to the latter case, although human reproduction isn't very constrained by material resources in the developed world, so I'm unsure how much society punishing some Dark Tetrad people affects their reproductive fitness. Also, some narcissists "succeed" greatly in our society (Donald Trump, Elon Musk, and too many others to list).

I take your point that darkness and hate can lead to love/reduction in hatred

I mainly mentioned those points in case someone would quibble with the absoluteness of King's quote. As they say, "all generalizations are false". In our present world, where we generally don't kill or tyrannize everyone who disagrees with us, I think it's generally more effective to move more in King's direction than our primate instincts incline us to. If you can't massacre your enemies, you have to find a way to make peace with them.

Thanks for the post! There's a lot of deep food for thought in it. I agree it's nice to know that you're not alone in having these kinds of feelings.

reading an article by Brian Tomasik one night[...]. It was the most painful experience of my life.

Sorry about that! Several people have had strong adverse reactions to my discussions of suffering. On the whole I think it's better for people to be exposed to such ideas, although some particular people may be more debilitated than motivated by thinking about extreme suffering.

I notice a trend for news and Google results to be increasingly censored of violent and gory content. For example, in 2021 there was a news story about Darrell E. Brooks Jr. driving an SUV into a crowd of people, and the footage of the incident -- shown from far away -- was blurred and had no audio. Viewers couldn't really see what happened at all. Such censorship is plausibly good for viewers' mental health, and it's very likely good for advertisers' brand safety. But it's plausibly bad for the victims of violence if it reduces motivation to address their suffering.

"Darkness cannot drive out darkness: only light can do that. Hate cannot drive out hate: only love can do that."

That's a great quote. It's applicable to so many misguided and harmful actions in the world today, from Israel's flattening of entire neighborhoods in Gaza to the vitriolic rhetoric used by some parts of the SJW community. I suspect the world would be more peaceful and advocates would be more effective if they embraced King's "love your enemies" approach more on the margin. That said, there are some cases where darkness and hate lead to love or at least a reduction in hate:

• Stockholm syndrome, abusive relationships, loyalty to oppressive dictators, worship of gods who threaten hell, etc. The general phenomenon is that a powerful alpha male is asserting supremacy by ruthlessness, and in that case, it may be more adaptive for primate brains to accept defeat and love Big Brother rather than continuing to fight him.
• If censorship is severe enough, it may be possible to reduce the exposure of future generations to certain hateful ideas. Also, if you murder enough of your enemies that their population becomes small, you can decrease the number of them who remain to hate you.

I haven't looked into sheep and goats specifically, but I imagine their wild-animal impacts would be fairly similar as for cattle. Unfortunately they're smaller, so there's more suffering and death per kg than for cattle, but they're still much better than chicken/fish/etc.

Dairy is another lower-impact option, and I guess a lot of Hindus are ok with dairy.

there's no sense in which asking dumb questions can plausibly have very significant downsides for the world (other than opportunity costs)

I think the opportunity costs are the key issue. :) There's a reason that companies use FAQs and automated phone systems to reduce the number of customer-support calls they have. There have been several times in my life when I've asked questions to someone who was sort of busy, and it was clear the person was annoyed.

At one of my previous employers (not an EA organization), I asked a lot of questions during meetings, which apparently other people didn't like, because it was distracting. During one meeting, people didn't even bother to answer my questions. A few weeks later, my boss told me that he overheard someone saying: "Don't invite Brian to this meeting; he'll slow us down with too many questions." I was accustomed to a school environment in which teachers would always say "There's no such thing as a dumb question", and I didn't realize that people outside of school may not feel the same way.

The situation might be better among altruists. I think one reason people at that organization didn't want to answer my questions was because they had no career incentive to do so, since they were evaluated based on what they individually produced, not based on helping coworkers. That said, lack of time can still apply in EA contexts. I often fail to reply to people who ask me questions, not because I think the questions are dumb but just because I'm slow and lazy and get asked questions frequently.

Thanks! It's worth noting that the rainforest and Cerrado numbers in that piece are very rough guesses based on limited and noisy data. As one friend of mine would say, I basically pulled those numbers out of my posterior (...distribution). :) Also, even if that comparison is accurate, it's just for one region of the world; it may not apply to the difference between, e.g., temperate forests and grasslands. All of that said, my impression is that crop fields do tend to have fewer mammals and birds than wild grassland or forest. For birds, see the screenshot of a table in this section.

Great post! In addition to biases that increase antagonism, there are also biases that reduce antagonism. For example, the fact that most EAs see each other as friends can blind us to the fact that we may in fact be quite opposed on some important questions. Plausibly this is a good thing, because friendship is a form of cooperation that tends to work in the real world. But I think friendship does make us less likely to notice or worry about large value differences.

As an example, it's plausible to me that the EA movement overall somewhat increases expected suffering in the far future, though there's huge uncertainty about that. Because EAs tend to be friends with one another and admire each other's intellectual contributions, most negative-utilitarian EAs don't worry much about this fact and don't seem to, e.g., try to avoid promoting EA to new people out of concern doing so may be net bad. It's much easier to just get along with your friends and not rock the boat, especially when people with values opposed to yours are the "cool kids" in EA. Overall, I think this friendliness is good, and it would be worse if EAs with different values spent more time trying to fight each other. I myself don't worry much about helping the EA movement, in part because it seems more cooperative to not worry about it too much. But I think it's sensible to at least check every once in a while that you're not massively harming your own values or being taken advantage of.

I think a lot of this comes down to one's personality. If you're extremely agreeable and conflict-averse, you probably shouldn't update even more in that direction from Magnus's article. Meanwhile, if you tend to get into fights a lot, you probably should lower your temperature, as Magnus suggests.

Thanks! Good to know. If you're just buying eyeballs, then there's roughly unlimited room for more funding (unless you were to get a lot bigger), so presumably there'd be less reason for funging dynamics. (And I assume you don't receive much or any money from big EA animal donors anyway.)

I'm honored that you're honored. :) Thanks for the work you do and for your answer here!

there are certain large grantors that I have been told prefer to fund nonprofits that already have raised at least a certain amount from other sources

Are those EA grantors? Or maybe you prefer not to say.

That makes sense about how more donors helps with fundraising. I wonder if that's more true for a startup charity that has to demonstrate its legitimacy, while for a larger and more established charity, maybe it could go the other way?

Makes sense about ex ante vs ex post. :)

Are you more optimistic that various different kinds of reflection would tend to yield a fair amount of convergence? Or that our descendants will in fact undertake reflection on human values to a significant degree?

Makes sense. :) There are at least two different reasons why one might discourage taking more than one's fair share:

1. Epistemic: As you said, there may be "collective wisdom" that an individual donor is missing.
2. Game theoretic: If multiple donors who have different values compete in a game of chicken, this could be worse for all of them than if they can agree to cooperate.

Point #1 may be a reason to not try to outcompete others purely for its own sake. However, reason #2 depends on whether other donors are in fact playing chicken and whether it's feasible to achieve cooperation. If you genuinely have different values from other donors, you should try to do the best you can by your own values, which could include taking advantage of opportunities to donate less than your "fair" share.

It's easy to feel warm fuzzies toward being "fair", but we can imagine scenarios where those fuzzies don't apply. For example, imagine that the USA and Russia are both contributing development aid to an international organization, and with any funds left over, Russia will buy attack drones from Iran. If there's an opportunity to get Russia to contribute more than its "fair share" to the development aid, leaving less money left over for drones, the USA should try to do that.

Maybe being the kind of person who would never even consider aiming to gain some advantage for one's own values is more effective at making cooperation actually happen, but being such a person could also lead to getting exploited. It seems non-obvious how exactly to best ensure that each party gives its fair share, especially when there are so many different possible donors to keep track of, and we have no way of knowing how much each entity would have contributed on its own.