Thank you for the words of support! So nice to hear from someone who doesn't empathize, aha. I'm delighted.

I hope to shed light... So that environmentalists see their own position clearly, with all the giant gaps that have been glossed over. I hope to demonstrate rigor to folks who think environmental positions are all just confused & wishful thinking. I hope people who care about nature feel seen, and want to get involved in building environmentalism with EA principles. I want EAs to see it's okay to care about nature and things that are meaningful to them, even if they aren't the most effective or the most universal.

And most of all, for my own sake, I want to interrogate myself and reflect the real world!

(I posted my next follow-up, and would be happy if you read it.)

I know what you mean! I am trying to provide a critical approach and would like to hear what you think of how it turned out.

Should forecasting receive more or less EA funding?

Mildly surprised how much is going into it. Mildly surprised how little is coming out of it. Still have high expectations for the return on improving decision-making. I'm a low information voter. Voting without factoring in a large increase in funds in the near future which would slam everything to the right.

I just came back from my first retreat and (due to unlucky events) had a pretty negative outcome. This gave me back some hope that there is something actually good out there for me yet. Thank you. Quality timing.

Total number is richness. 

Buncha pedantics: The problem with that is it's easy to add more "weedy" species to increase biodiversity per local area, even as you decrease global biodiversity. The real number is global extant species but we aren't perfectly coordinated to know what this specific place should do best relative to what everywhere else is doing. Aaand you can't totally re-establish density with reduced genetic diversity if it gets bad enough, but you are correct. Its the irreversible damage that is most concerning. 

Thanks for noticing that! I'll fix it.

Its been a week and I've spent some time trying to get a fuller picture of the insect welfare space. Thanks to Toby for urging me to do so.

Some things I knew previously but are worth repeating:

• Evolution isn't welfare optimizing.
• Insects have different life histories than other animals. They are in bit of a different class.
• Avoidance of danger is urgent so insects (probably) experience intense impulses. (aka pain and panic, even if they are otherwise very simple impulses)
• Insects might be too simple for hedonic adaptation (normalizing/numbing).
• Insects might experience each moment (of pain) as a discrete event. This might make painful experiences less bad (a single moment of shock), or it might make them worse (full blast isolated meaningless torture) depending on how you interpret it.
• Disease.
• Losing a leg would be proportionally bad to how much it reduces their survival odds.
• Most of the wildlife I see and interact with is pretty happy. But most of the wildlife I see are birds/mammals/large bodied/adult/healthy/k-strategists.
• It is very suspicious to me that we think that animals we understand the least are the ones we think have the most horrifying lives.

Some things I learned:

• Like 10% of insects die in minutes or hours of existing. Before they have time to have positive valence experiences.
• Another 10% die a bit later without getting to experience much and having probably bad experiences the whole time. (don't take these numbers too seriously, they keep changing as I try to figure out what is going on)
• Newly hatched larvae have tiny energy reserves relative to their metabolic rate, so starvation might be quicker and less devastating for insects than our default picture of starvation.
• Like 20%??? of insects get consumed from the inside by parasites for long periods of time. The parasites keep them alive and eat their non-essential organs first. This seems really bad.
• Insects might be more prone to fitness critical (life and death) events than other animals.
• R-selection might produce different welfare profiles.
• Most predation is quick. Seconds or minutes of dying. Extreme, but short. Like 50% of insect deaths are this way.

Some other thoughts:

• R-strategists might have different values than us. But I'm not sure I want to act on moralities that "pain doesn't matter" even if the uplifted insects might want that. Its a separate question than "what does our own morality say is right."
• Insects might not be morally important. This is a separate question from suffering and sentience.
• Generally, the insects with more evidence for memory, sentience, creativity, intelligence, etc are also not the ones that have the worst looking lives. (?)
• I have a strong instinct to think that the hedonic baseline of species is set by their lifehistory. If so, many species would experience substantial mortality and stress as normal. "Routine" portions of insect life I think would be closer to neutral baseline, even though the niche looks harsh.
• Bob Fisher pointed out that the goalposts keep moving, and that makes me very uncomfortable.

If we could press a button and fix all this I would be much more likely to think that insect lives are net negative. Which does not bode well for my objectivity here. For example, I am much more likely to think "disease is bad, really bad, we should do something about that for the insects" than I am to think "r-strategists have innately horrible-shaped lives. It's really bad. We should do something about that." I think this reveals something.

So right now I have updated to believe that some species probably do have truly miserable welfare profiles, bad enough to be seriously consider ending their way of existence. I'm not sure how many, probably few? (I already believed that some parasites might be overall bad and we should probably end their way of existence despite their amazing intricate biology) I probably think this circumstance is a lot rarer than other insect welfare supporters. I still anchor pretty strongly to expecting insect lives to be neutral as a default, despite frequent suffering events. I probably ought to create some serious goalposts and stick to them. I still have other objections along the lines of "the insects would have different values" and "its really suspicious that we think their lives are the worst when they are the most hard to sympathize with and we understand the least about them" and "we would not apply this moral system to ourselves" and "similar/small lives matter a lot less" and "suffering is less bad than weighted, enjoyment is more readily available and better than weighted." I am pretty strongly in support of intervening in nature to make it better, and I don't think the precautionary principal holds up under scrutiny. 

In general I am confused whether I should be thinking about suffering as a proportion of all insects, proportion of time within a species average, proportion within a life, total number of individuals, or just total bad moments. 

I'm really excited to see this. I'm optimistic that ecologists and conservation professionals will want to help with this. They WILL be excited to have an ally that cares about non-charismatic species. They are begging for more attention to the smaller foundational participants in the ecosystem. The field is also gradually becoming more pro-intervention. I am also optimistic that AI will contribute, and want to emphasize that a simple but important step is deploying far more field data loggers. The sooner you collect data, the sooner you have years of data to work with. 

I did not find this patronizing! I recognize that this is a charged topic and really appreciate you moving the discussion to higher quality format/content. Don't overthink too much, I am one person and being a little neurotic. You cleared the air. 

 I've probably missed quite a lot. I'm aware of this article about how animals die and the signs of distress very small animals exhibit. Gonna take your advice and try to get up to speed to see what I've missed.

The post you recommended seems like a very nice outline of the premises in (wild) animal welfare and problem with being able to tell how good an animal's life is, and how that measures up in the grand scheme of things. It is unfortunately dense and long and from 10 years ago, but it make me see this is well-trodden ground and left me wanting to know what the more recent developments are. 

Edit: I want to add that the tone of this comment felt dismissive to me when I first read it. I think I'm sensitive to being considered shallow for disagreeing so the word choices of "if you are new" and "this is a nice first engagement...to engage more deeply...is a great place to start" set me off. I've read a bit. I have heard some counter arguments. I'm not completely uninformed. I'm unconvinced. (So far.)

Among EAs that are focused on preventing suffering in the non-human world, United States based.

This is not what most conservation scientists and "animal lovers" tend to espouse. Those groups also have severe biases in what they pay attention to.

Net negative lives is a bit of a weird concept to grapple with, but I wouldn't dismiss it immediately (despite how it sounds at first encounter). They've definitely thought about it. The problem with net positive lives is that you then might want to fill the world with slight net positive lives and get to a Malthusian state. Its not simple to reason about what a net positive/neutral/negative life is, nor how good/bad it is that lives that never come to be, nor how the world should be shaped in response to these things.