Thanks for elaborating. I created this post with Claude about whether reducing small risks increases large ones. I plan to publish it in 7 days or so. I will let you know once I do.

Hi Tandena.

How much do these values dominate our decisions versus other values?

Here are some comparisons between the total welfare of animals, trees, and microorganisms.

Hi Jordan.

A near-term slowdown could prevent the warning shots that would make the public take AI risk more seriously.

Do you think this generally applies to other risks? For example, could decreasing the risk of hard aircraft landings increase the risk of catastrophic aircraft landings?

Thanks.

Corporate campaigns to reduce the suffering of farmed chickens have been estimated to spare numerous hens from cage confinement per dollar

I wonder if it would be better to say "around 10 hens" instead of "numerous hens", which is more ambiguous.

mean of 10, with a range between 3.6 and 22 hens per dollar

Nitpick. I think you meant 3.6 to 29. Laura Muresianu said there was "a typo in the methodology -- it should be 3.6 to 29 chickens per dollar spent, not 3.6 to 22".

Great to know.

In the coming weeks, we plan to address various concerns about philanthropic giving, like:

• Giving now vs. later
• Giving in the face of new AI trajectories
• Cause prioritization done wrong and how to prevent it
• How to avoid giving disasters
• And many more.

Will you address effects on non-target individuals?

I see no escape from the uncertainty about the effects on soil invertebrates if one wants to increase welfare while accounting for all animals. I am not aware of any intervention supported by impact-focussed funders with effects on soil invertebrates robustly smaller than those on the target beneficiaries. Not saving human lives, not decreasing the consumption of animal foods or ingredients, not replacing fast with slower growth chicken, not replacing layers in battery cages with ones in cage-free aviaries or barns, not replacing standard with bird-safe glass, not replacing rodenticide with fertility control bait, and not even electrically stunning farmed shrimps.

I just linkposted a comprehensive analysis of how electrically stunning farmed shrimps affects the living time of soil invertebrates. I continue to think it may change the welfare of soil invertebrates more or less than that of farmed shrimps.

Hello. Thanks. Assuming the stunners are subsidised has a small effect on the change in soil-arthropod-years per electrically stunned shrimp.

Who pays for the stunner. Scenario. The headline assumes the producer pays for it, as under a legal requirement to stun or unsubsidised adoption. Where SWP donates the stunner, as in its Humane Slaughter Initiative, only the running costs raise the producer's cost of stunning a shrimp: the mean cost falls by 32.2%, and every mean change in living time falls by about as much (33.0–34.1% in the simulation), with the probability of more living time unchanged. Full estimates are in the Scenarios section.

Here are the details.

I assume the stunners are not subsidised in the baseline scenario because I am expecting subsidised stunners to stun a minority of the shrimps electrically stunned in the future. @Aaron Boddy🔸, who was the CEO of the Shrimp Welfare Project (SWP) until July 2026, commented the following on 7 Octoner 2024.

We’ve had some good indications recently that HSI does contribute to “locking-in” industry adoption, with Tesco and Sainsbury’s recently publishing welfare policies, building on similar wins in the past (such as M&S and Albert Heijn).

This has always been the Theory of Change for the HSI project. Although we’re very excited by how cost-effective it is in its own right, ultimately we want to catalyse industry-wide adoption - deploying stunners to the early adopters in order to build towards a tipping point that achieves critical mass. In other words, over the next few years we want to take the HSI program from Growth to Scale.

Hi Marcus. Nice post.

Corporate campaigns to reduce the suffering of farmed chickens have been estimated to spare hundreds of hens from cage confinement per dollar, which is plausibly orders of magnitude more cost-effective, in suffering averted, than even the best global health interventions, if you’re willing to give chickens some moral consideration.

I think funding current work on corporate campaigns targeting layers helps a few hens per $, around 1 % of the cost-effectiveness mentioned above. Animal Charity Evaluators (ACE) estimated The Humane League's (THL's) accountability work targeting layers in 2024 helped 11 hens per $. My intuition having worked for the last 11 months at Anima International is that our marginal cost-effectiveness is 30 % of the cost-effectiveness of our whole spending. Applying this multiplier to ACE's estimate, I get 3.30 hens helped per $ (= 0.3*11). Relatedly, @abrahamrowe guessed in January 2025 that corporate campaigns for chicken welfare not funded by Coefficient Giving (CG), including ones run by organisations funded by CG, were as cost-effective as burning money. I was surprised by Abraham's claim at the time, but I have meanwhile come closer to his position.

I think we know for sure the probability that people who've been convicted of crimes will reoffend is far greater than for people who have never committed a crime. We don't have to ask what if.

I agree people who have been convicted of crimes are much more likely to commit crimes than people who do not (although I wonder how much this varies with the type of crime). However, say the probability of a Austin's alternative hire commiting a financial crime was 0.01 %, and that this is 0.1 % for Caroline, 10 times as high. I just made up these probabilities. It is not obvious to me that 0.1 % is too high, and that therefore Caroline should not have been hired. Is your view that the probability of Caroline commiting a financial crime at Manifund is obviously too high? I expect hiring managers to generally be more informed about the people they hire than me because they have much more context.

Hi Jim. Great post.

One reason to doubt DiGiovanni's unawareness argument (specifically, his empirical premise) is that it is unclear where the line is between A) my understanding of a decision problem is fine-grained enough for me to trust my best guess, and B) it is too coarse-grained (and I must suspend judgment). While DiGiovanni does not object to this and recognizes some version of the problem, he suggests that the line must come before “promoting the impartial good”.

I do not think there is a fact of the matter about the location of the line separating A) and B). I would say it depends on the operationalisation of the claim under assessment, which should be falsifiable. Relatedly, I agree with Brian Tomasik that:

It doesn't make sense to ask questions like, Does a computer program of a mind really instantiate consciousness? That question reifies consciousness as a thing that may or may not be produced by a program. Rather, the particles constituting the computer just move—and that's it. The question of whether a given physical operation is "conscious" is not a factual dispute but a definitional one: Do we want to define consciousness as including those sorts of physical operations?

In the context of consciousness, I think the right move is i) considering all systems have more than exactly 0 moral value instead of ii) separating systems into ones that have exactly 0 moral weight, and ones that have more than exactly 0 moral weight. In the context of cluelessness, I believe the right move is i) considering the expected value (EV) of any 2 actions is comparable more or less robustly instead of ii) separating actions into ones whose EV is comparable, and ones whose EV is not comparable.