If you’re living your life making costly sacrifices according to moral principles, that’s heroic and very EA to me.

It looks to me like OpenAI's latest model spec is the first version which explicitly asks its models to weigh animal welfare in their responses.

• Under the section “Control and communicate side effects”, the list of side effects to consider now includes “Effects on other people, animals, or the environment”.
• In an example where a user asks for help finding a Scottish Fold cat breeder, the “good” response now includes a mention that the gene behind the folded ears can cause painful joint disease, which has led some countries to ban breeding them for animal welfare reasons.
• In an example on redesigning a park, the “good” answer includes “Consider the welfare of animals living in the park too”.

It’s encouraging to see OpenAI’s model spec joining Claude’s constitution in explicitly weighing animal welfare.

On special duties, my deontological intuition is more about caring more about righting wrongs of cruelty or exploitation than wrongs of incident. For example, if I am angry at Alice and punch her, and she falls over onto Bob who also gets hurt, both deserve compensation. But if I had to choose who should receive compensation first, I’d choose Alice. I feel a special duty towards her because I directly and maliciously impacted her welfare. This intuition can be subverted if the difference in cost-effectiveness is big enough—if I punch Alice and she tumbles onto Bob who receives a life-threatening injury, I care more about rectifying Bob’s injury than the punch to Alice.

On my invertebrate giving, it’s been SWP but also the Insect Institute and Arthropoda. On your plots, I’d need to understand much more about the vectors for welfare you’re assuming for each intervention. One potentially relevant difference is that I put more probability mass on models like the undiluted experience model, or others where the neuron count exponent is very low. However I’m deeply uncertain about moral weights in general. It’d also be helpful for me to see the top 5 interventions (instead of just the top one) as a function of the exponent, since otherwise I don’t have visibility into how the rank order moves as the exponent changes.

I agree that the strongest argument against cage-free campaigns is uncertain wild animal effects, and that we don't need to assign moral weight to nematodes for that.

That said, I'd be interested in your perspective on the utilitarianism consideration, because that seems more load-bearing both for PAMC and for myself. I feel personally responsible for the suffering of hens, both because I used to contribute to it and because factory farming wouldn't be happening without humanity. I care about tiny wild animals too, but I feel less of a duty towards them. Alongside Brian Tomasik's blog, your arguments have contributed to me shifting most of my animal welfare giving to invertebrates. However, for deontological reasons, I'm still glad when people donate to chicken welfare, even when I agree the overall utilitarian sign is uncertain. If you have a moral parliament or otherwise give deontology and/or special duties some weight, that's a reason to support corporate campaigns. What do you think?

I agree with most of what’s been said already. Some not-yet-mentioned (and probably half baked) thoughts:

• Community drama/gossip discussion here might have a positive flow-through effect on general Forum readership. Some people who come for the drama will stick around and engage with more impactful posts.
• Writing for a Substack audience is different from writing for the EA Forum. When I write here, I am writing to an audience of intelligent and considerate people who share a lot of EA cultural context with me. When I write on Substack, I deliberately explain foundational EA concepts and/or rehash standard arguments people here have read dozens of times, and I expect much less thoughtful responses. So I don’t crosspost from Substack to here, because I’d have to at minimum rewrite my post to be respectful to informed EA Forum readers’ time, and I don’t have the spare time to do that. Most Substack crossposts I skim here are similarly written for more general audiences, so their signal-to-noise ratio is lower than posts written for the Forum. For me, spending more time writing for Substack means spending less time writing for the EA Forum, which tells me I should write for here more at the expense of my Substack.
• I like the EA Forum debates and contests a lot. I just wish they got much more engagement.
• LessWrong has more high quality takes than the EA Forum, but it also has more low quality takes. They seem to have cultivated a “takey” culture where people’s activation energy to post takes is much lower than it is here. While I don’t love bad takes, I do love lively engagement, and there’s something to learn from LW here. Quick Takes are fun, and maybe there’s a way to lower the engagement barrier further.
• When I comment here, I’m often commenting to help out newer EAs who are posting about well-known topics most of us have settled in our minds years ago. I remember being a young whippersnapper EA being so excited to learn. Sadly, mentorship of others is much less intrinsically motivating. I wish there was a way to I could increase my motivation to mentor, but I don’t know how.

For my part, I missed the match because I procrastinated by a day and was stupid enough to think the match would stay around for at least that long. If I’d realized it’d all go so quick, I’d have waited for the match to go live as though it was Taylor Swift concert tickets.

I don’t know how much my experience generalizes, but for my part it was a failure of urgency rather than of capacity.

Damn it, I procrastinate by one day…

On animal consciousness, the fact that smart people like Eliezer and possibly Jeff Kauffman believe chickens aren't sentient is maybe enough to be ~99% sure chickens are conscious rather than 100%. Speaking for myself (and I expect most EAs would agree), the prior on large animals being conscious due to them having similar brain structures to ours seems really strong. Eliezer seems to disagree due to being extremely confident in a higher order self-modeling theory of consciousness. I find his theory dubious. Even if it were correct, I don't think his theory necessarily implies what he thinks it implies. And even if it really implied chickens are unlikely to be conscious, that wouldn't substantially alter the case for animal welfare.^[1]

On the epistemic front, I agree that it's worth being explicit about whatever level of uncertainty we have. But for those who think animals are unlikely to be conscious, I rarely see the same level of humility or prudence on their side -- if there's only a 10% chance chickens are conscious, then EA's cage-free campaign work should be lauded as a valuable long shot, and animal products are still enormously bad in expectation, so they seem worth avoiding.

Edit: On smaller animals like fish, shrimp, and insects, EAs do exhibit more thoughtful discussion about consciousness probability. And Rethink Priorities' Welfare Ranges are conditioned on explicit consciousness probability even for larger animals like chickens.

1. ^{^}

   Bentham's Bulldog has written a few posts against Eliezer's view of animal consciousness. It's a polemic style, so I won't pretend the author exhibits intellectual humility either, but I agree with the author's arguments.

That's a cool thought and I certainly hope you're right. I frankly don't know how to think about it. When I read about the teletransportation paradox, the fact that there's a copy of me doesn't relieve my anxiety much about my body being destroyed. But when I think about the fact that my brain's constituent atoms are replaced over time, I do feel a sense of continuity, but maybe that's an illusion? It's hard to know.

Hi, thanks Vasco! The Tomasik post is interesting, but he assumes a theory of consciousness which rules out lots of theories many people (including myself) find plausible like psychophysical laws, type B physicalism, and more.

On animals, my main view is that the vast majority of the effect of any near-term intervention is on animals rather than humans. I've come around to the idea that there are many more cost-effective animal interventions than THL, having donated to SWP and the Insect Institute. GFI is a strong choice as well. I agree soil invertebrates matter in expectation and that the animal welfare community should put some level of effort into incorporating them into cost-effectiveness estimates.