Steven Rouk

• Steven Rouk 4 Aug 2025 16:40 UTC

  1 point

  0 ∶ 0

  on: Road to AnimalHarmBench

  This is very cool work and I would love to see it get more attention, and potentially get it or something like it incorporated into future AI model safety reports.

  A couple things:

  1. Does your approach roughly mirror the assessment approaches that companies currently take for assessing potential harm to humans? I imagine that it will be easier to make the case for a benchmark like this if it’s as similar as possible to existing best-in-class benchmarks. (You may have done this.)

  2. I would recommend improving the graphs significantly. When I look at model performance, safety assessments, etc., 90% of the “rapid takeaway” value is in the graphs and having them be extremely easy to understand and very well designed. I cannot, at a glance determine what the graphs above are about other than “here are some models with some scores higher than others”. (Is score good? Bad?) I’d recommend that the graph should communicate 95% of what you want people to know so it can be rapidly shared.

  Thanks for your work! Very important stuff.

AI Anal­y­sis of US H.R.1 (“Big Beau­tiful Bill”) Im­pacts on Farmed Animals

• Steven Rouk 22 May 2025 20:55 UTC

  2 points

  0 ∶ 0

  in reply to: Jim Buhler’s comment on: Sen­tien­tia, Rev­er­some­las, and the Small An­i­mal Re­place­ment Problem

  Yeah, I think (a) many people don’t think about it, (b) some people do think about it and try to adjust accordingly (which they may succeed or fail at), and (c) some people acknowledge it but think of it as a temporary setback related to moral circle expansion.

  I’m curious if you’re hoping to shift people’s thinking about strategy in any specific direction here, due to bringing this up? Are there specific changes you’d like to see advocates make, or some definition of “success” for improving advocacy strategies based on this awareness?

  Might be interesting to do specific quantitative analysis and projections of various scenarios, as well, although of course that’s its own research project.

• Steven Rouk 21 May 2025 20:18 UTC

  3 points

  0 ∶ 0

  on: Sen­tien­tia, Rev­er­some­las, and the Small An­i­mal Re­place­ment Problem

  Hey Jim! Thanks for pointing me here via your comment on my post about Adding Nuance to the Small Animal Replacement Problem: Moral Circle Expansion.

  I don’t have much to say, other than thanks for continuing to think about this issue. I’ll add a couple of (hot take) thoughts that might be relevant, based on my current read of the movement:

  1. It seems to me that many animal advocates these days are aware of SARP and actively tailor their strategies to avoid it where possible. For example, being very careful to advocate for plant-based defaults at institutions rather than just for “reduced emissions” food swaps, when doing environmentally-focused campaigns (since “reduced emissions” could technically just swap beef for chicken, decreasing emissions but increasing total animals killed /​ days of suffering).

  2. More organizations are working on small animal welfare than ever before, such as the large amount of money and energy that has gone into fish welfare and chicken welfare in recent years, and even in previously completely-ignored areas (like shrimp welfare, insect welfare, etc.).

  So I think many of the main actors in the advocacy space are trying to adjust their strategies to account for SARP, regardless of whether or not it’s only a temporary setback from moral circle expansion. But I’m not sure how the percentages shake out, of who is accounting for it and who isn’t.

  I think your Reversomelas scenario is very interesting and worthy of more analysis.

Con­nect For An­i­mals 2025 Strate­gic Plan

• Steven Rouk 15 Nov 2024 21:08 UTC

  3 points

  0 ∶ 0

  on: How would your pro­ject use ex­tra fund­ing? (Marginal Fund­ing Week 2024)

  Here’s our Marginal Funding Week post for Connect For Animals. I’m Steven Rouk, the Executive Director. :)

How Con­nect For An­i­mals Would Use Marginal Funding

• Steven Rouk 12 Oct 2024 22:44 UTC

  4 points

  3 ∶ 0

  on: Some rea­sons for not pri­ori­tis­ing an­i­mal welfare very strongly

  Thank you for writing this post; I thought that some of the points you mentioned here were among the strongest arguments I’ve seen during Debate Week for why one might not prioritize animal welfare as highly. (Note: I run an animal advocacy organization focused on ending factory farming and voted “strongly agree” for spending the $100m on animal welfare.)

  Specifically, I think your comments about solvability/​tractability are very important to keep in mind. This idea becomes clearer for me (and probably others) when considering wild animal welfare and invertebrate welfare, as you brought up. Those are areas where my gut reaction is something like “wow those are really tough, maybe we should just focus right now on animal suffering that humans are actively causing (like factory farming) since that is more tractable.” Not that I think my gut reaction is correct—just that I can see this being an important consideration to keep in mind.

  The thought experiment of traveling back in time is very interesting, as well. As one specific personal anecdote, the idea of “the correct era for advocacy” comes up for me when wondering if artificial intelligence may one day be able to help us with wild animal welfare at scale. If AI were to end up being crucial for certain efforts (like wild animal welfare), then I could see how that consideration might change one’s prioritization of what to work on right now. (Note: I’m familiar with wild animal welfare/​suffering work, but am certainly not an expert.) But of course, society would need to want to use AI for wild animal welfare work, and so I think in this specific instance it might just shift your resources potentially away from direct work and more towards field building and convincing people of the importance of the cause. (Which may be exactly what some wild animal welfare advocates are doing these days.)

  I’m not sure if these considerations would necessarily change my vote of how to prioritize the funding in this specific $100m question, partially because I have taken many of these ideas into account already, but I think these are the some of the right questions to be asking, and I haven’t seen these ideas discussed as much. (I’ve seen moral value comparison stuff discussed way more, how much do you value chickens vs. humans, etc.) I think questions of present tractability become much more important when dealing with overall funding allocations, such as the question of should a much larger percentage of our funding be going towards animal welfare, rather than a marginal $100m (which we can more confidently allocate to animal welfare).

  Also, I have notes in my comment here about some organizations that I think could effectively use quite a bit more funding, and some new initiatives (like plant-based defaults in institutional food settings, like the work done by Greener By Default) that seem possibly very impactful but haven’t scaled up (primarily due to lack of funding, I would guess).

• Steven Rouk 10 Oct 2024 18:42 UTC

  1 point

  0 ∶ 0

  in reply to: Toby Tremlett🔹’s comment on: Dis­cus­sion thread: An­i­mal Welfare vs. Global Health De­bate Week

  I added a couple notes about tractability in my comment here, if anything in there is new information to you: https://​​forum.effectivealtruism.org/​​posts/​​nrC5v6ZSaMEgSyxTn/​​discussion-thread-animal-welfare-vs-global-health-debate?commentId=oKbr42XMkwu8bgCMt.

  I think there are a lot of impactful, established initiatives that could utilize extra funding, as well as some newer initiatives that are very promising that could be scaled up. Plus lots of work to be done in neglected regions.

• Steven Rouk 10 Oct 2024 16:24 UTC

  9 points

  3 ∶ 0

  on: Dis­cus­sion thread: An­i­mal Welfare vs. Global Health De­bate Week

  In general, I agree with the position that investing an additional $100m into animal welfare opportunities would be more impactful than global health opportunities even under views that use moral weights on the lower end of the scale for nonhumans, and potentially way more impactful if we use moral weights that grant nonhumans greater capacity for sentience (which I think we probably should).

  In short—I think the scale of animal suffering is much larger (even when only considering animal agriculture, and not wild animal welfare); animal welfare is much more neglected (even when only considering Open Phil’s grants, which probably skew towards animal welfare compared to average funding); and I’m less certain about the difference in tractability, but I think it’s reasonable to think that tractability could be comparable for both. If animal welfare work is less tractable for some reason, I think it’s unlikely for the reduced tractability to make up for the large difference in scale and neglectedness.

  Other pieces have made a much more in-depth case for this than I will do here, such as Open Phil Should Allocate Most Neartermist Funding to Animal Welfare, The Marginal $100m Would Be Far Better Spent on Animal Welfare Than Global Health, other pieces linked by MichaelStJules before debate week began, and the Rethink Priorities moral weights project. Below, I’ll just add a smattering of points that might be important that I don’t see discussed as frequently.

  Neglectedness: Global health seems likely to continue getting more funding than animal welfare due to society prioritizing human interests. Thus, animal welfare will probably continue being more neglected.

  Not only does global health get much more funding now than animal welfare, but I imagine this will continue to be the case going forward because humans and human institutions seem to strongly prioritize human welfare over nonhuman welfare.

  If that’s true, it would mean that animal welfare is not only much more neglected now, but on the whole it will probably remain that way for many years, if not indefinitely. Thus, marginal dollars will probably be more impactful when allocated to animal welfare.

  Scale: Nonhumans may account for the majority of moral value now and indefinitely into the future.

  Not only do nonhumans seem to account for the majority of all moral value now, but it seems like they probably always will, unless something drastic happens to change that. One implication of this is explored in the piece Net global welfare may be negative and declining.

  I think it’s unlikely that humans are or will be the majority of moral value, given the history of the world so far and some reasonable assumptions about our trajectory into the future. (I think while possible, it’d be much more surprising if humans do end up ever becoming the majority of moral value.)

  If that’s true, then animal welfare efforts are important not only in the near-term but also in the long-term, since the moral importance of our decisions will mostly depend on the effects on nonhumans.

  Tractability: There seem to be many impactful animal welfare initiatives that could already effectively utilize much more funding, and many more projects could exist if funding were available.

  There are great organizations that exist now that I think could easily deploy an additional $1m+, such as The Humane League, the Good Food Institute, Mercy For Animals, etc.

  Additionally, there are small /​ mid-sized impactful organizations that could scale up their work with access to additional funding, such as Aquatic Life Institute, Shrimp Welfare Project, and New Roots Institute.

  There are also potentially impactful megaprojects for animals that others have written about (some of which could be promising).

  There are whole categories of work that are just getting started that could use a lot more funding to scale. For example, I think plant-based defaults work seems extremely promising, and that work is really just beginning and has not been executed at scale yet. Organizations working on this like Greener By Default are relatively new, but their work seems to already be having a large impact. I think one of the best examples of this is NYC Health + Hospitals switching to plant-based default meals, which over half of patients ended up choosing, leading to millions of plant-based meals being consumed instead of animal-based. Plant-based defaults as a strategy is very new and (in my opinion) could effectively utilize millions of dollars in the coming years. I think we’re still in the early stages of discovering how impactful this strategy could be.

  Others have pointed out that public policy /​ political advocacy is particularly neglected in the animal welfare space and could use more funding, and I think we’ve seen some successes here already with a relatively small amount of funding. (Animal Policy Careers is a new initiative focused on this space, one of the few.)

  There’s also the opportunity to grow the animal welfare movement in neglected regions, in particular Asia and Africa, and for neglected species, in particular aquatic animals.

  And, as the animal welfare space grows, there are new opportunities for meta /​ movement building organizations like Connect For Animals and Hive, and for capacity building organizations like Vegan Hacktivists and Animal Defense Partnership, and for additional global conferences like the AVA Summit series.

  From my experience as a nonprofit founder in the animal welfare space, many impactful organizations are very funding constrained, and many other impactful organizations do not get started due to lack of funding. Given the scale and neglectedness of animal welfare work, I think that a good number of initiatives would clear the bar for investment.

  (And if we needed more research and analysis of the impact of these organizations, then part of the $100m could go towards that.)

  Reason why global health may be a more effective use of $100m now than animal welfare: second-order effects for total future well-being.

  In general, as mentioned above, I do think there is a very strong case for spending an extra $100m on animal welfare than on global health.

  But, I think the strongest argument that might nudge me in the other direction is the idea that second-order effects for global health could result in higher overall future well-being, possibly due to better governance and collaboration by humans in the future that leads to overall reduction in nonhuman suffering as well. In this case, I still think the majority of impact in the long-term would probably come from positive effects on nonhumans.

  However, the track record of development in richer countries doesn’t necessarily make me think that a positive trajectory would automatically be the case (i.e. that human development will naturally lead to better overall well-being for nonhumans as well). In fact, increased wealth has seemed to go along with increased meat consumption, with that increased consumption being the fundamental cause of factory farming. Increased development also seems to correlate with increased animal exploitation in other areas, and greater impacts on wild animals (which still seem very difficult to fully account for). So I’m not convinced that human development will naturally lead to greater animal welfare; if current society serves as an example, it seems like the opposite may be true. Thus, heavy investment in animal welfare work still seems extremely necessary if we are going to turn our growing wealth and health into a society that positively affects nonhumans.

  What links here?

  • Steven Rouk's comment on Some rea­sons for not pri­ori­tis­ing an­i­mal welfare very strongly by Engin Arıkan (12 Oct 2024 22:44 UTC; 4 points)
  • Steven Rouk's comment on Dis­cus­sion thread: An­i­mal Welfare vs. Global Health De­bate Week by Toby Tremlett🔹 (10 Oct 2024 18:42 UTC; 1 point)

Three tricky bi­ases: hu­man bias, ex­is­tence bias, and happy bias

• Steven Rouk 1 Feb 2024 14:39 UTC

  1 point

  0 ∶ 0

  on: AMA: Emma Slaw­in­ski, the RSPCA’s Direc­tor of Policy, Preven­tion and Cam­paigns.

  Do you think about fish or insects at all, or are you choosing to remain focused on chickens for now?

• Steven Rouk 1 Feb 2024 14:37 UTC

  1 point

  0 ∶ 0

  on: AMA: Emma Slaw­in­ski, the RSPCA’s Direc­tor of Policy, Preven­tion and Cam­paigns.

  Thanks for doing this AMA, Emma!

  I’m curious as to how you think about prioritizing different types of interventions that might reduce the suffering of farmed animals in very different ways, for example:

  1. Working on reducing suffering through welfare reforms.

  2. Working on reducing numbers of animals through diet change policies (e.g. plant-based default food policies) or behavior change messaging (e.g. pro-veg messages).

  3. Working on reducing numbers of animals through increased alternative protein availability and consumption (e.g. plant-based meats, etc.).

  Since each of these is a plausible way to help farmed animals, yet they’re very different strategic approaches, how do you all think about allocating your time and energy?

Ad­ding Nuance to the Small An­i­mal Re­place­ment Prob­lem: Mo­ral Cir­cle Expansion

[Cause Ex­plo­ra­tion Prizes] Con­nec­tion can ac­cel­er­ate the end of fac­tory farming

• Steven Rouk 2 Mar 2022 19:14 UTC

  8 points

  0 ∶ 0

  in reply to: Charles Dillon 🔸’s comment on: EAA is over­in­vest­ing in cor­po­rate welfare reforms

  I’m late to the discussion, but I might add that I have a hypothesis that we have heavily underinvested in finding, connecting, and supporting existing supporters of farmed animal welfare. One symptom of this would be a seeming lack of diversity in the funding opportunities. Another symptom might be difficulty finding these opportunities, even if they do exist, due to lack of social network connectivity (i.e. there are no easy ways to find opportunities outside of our well-connected local social networks). Thus, perhaps one of the first things we should invest more heavily in is building up this connective infrastructure for the movement.

  Lastly, I think the definition of “good opportunity” varies wildly, and a more holistic understanding of risk and uncertainty would nudge us in the direction of valuing strategic and tactical diversity as an inherent good, above and beyond any kind of impact evaluation or estimation. Thus, at an extreme, if you had 100% of funding invested in CWRs, then nearly any non-CWR opportunity would be seen as a good opportunity due to increasing the diversity of approaches.

  Of course, we don’t have that extreme case of 100% investment in CWRs, but I think Kato’s point is that a more pluralistic movement (i.e. a more diversified one than we currently have) does probably lead to higher impact, which would expand our definition of good opportunities to include things we might otherwise pass on.

  I believe Harish Sethu gave an excellent talk at the AR Conference a few years back using an apples and oranges market analogy to demonstrate this same kind of idea.