Hey Chana, 

This post is very useful among other reasons to see how thoughtful the work came about, what went into it, why you think it worked, and what metrics and data you look at.

Thanks for getting it out the door when you did. It meant I actually had time to read it.

At ACE we have this Better for Animals resource that reviews evidence for various interventions. It also covers documentaries briefly. Pro-animal docs' lessons might not be particularly applicable to AI safety ones, but maybe it gets you thinking about those CTAs.

Because that's the main question I have after reading this post, what are you evaluating for? 

It is very impressive what you all did and I envy/admire the results. You made a beautiful product that people respect and many, many people consumed. But that does not equal changing the path we're on. Or maybe it does if awareness is the thing that you think shifts our future.

I'm not suggesting you didn't achieve your purpose, just that it's not crystal clear from this retrospective what that was and how you measure getting there. Maybe in the next evaluation report, you can share the purpose and outcomes you set for the video project, or the theory of change.

If it's primarily awareness that you're going for than the currently listed data points seem great. If you want public discussion, than optimizing for those comments might be the thing to try and the CTA can be something around asking people to leave in the comments (or when they share the video) about what worries/questions they have. If you need people to talk to their representatives, then your CTA and content maybe need another slight adjustment to prime them for that. (IDK if public discussion or political activism are useful steps on your path to impact. Again, I know embarrassingly little about this cause area.)

I admit that I'm curious in large part because I hope to learn from it for my own work. There are so many animal welfare documentaries and quite a few voices (and some funders) saying we need more of them based on their personal experience or intuition. But we don't have clear evidence that that these are a cost-effective method to help animals.

Now, the goals and context for AI safety are very different from those for animal welfare so it might work and work differently for you. But if I better understand what you're trying to do, then I can more easily see what the animal advocacy community should (not) copy.

Separate but somewhat related: how do you think about the use of videos as a path to change against different AI timelines? Do short timelines mean that time-consuming, high-production value video are less or more useful than some other alternative, e.g. high volume of low-cost, quick posts at strategic spaces. This is not my area of expertise so I'm honestly just wondering. (Again partially so I can learn from it to help animals better.)

Anyway, here's that resource and thanks again for sharing these insights:

https://docs.google.com/document/d/1O0ylEEQJMQMTBlHDHcNZwvTgifi7TNyd6GpabC_4VT0/edit?usp=drivesdk

One more caveat: some of the studies we looked at might be based on a time when media was consumed very differently.

My household's giving this year: ACE Recommended Charity Fund, ACE Movement Grants, ACE, Strong Minds, Friendship Bench, GiveWell Top Charities, Wild Animal Initiative.

Some random warm-glow giving related to friends and family and some smaller EA gifts.

I think I understand the worries and discomfort people feel about this approach. But I’m not sure how fruitful it is for all of us to have a vibes-based conversation about the possible merits of this campaign. It already exists. It might end up being good, it might end up being bad. We can make it better. If you think some of the risks taken and assumptions made by FarmKind are unaddressed, let’s talk about how we can mitigate those. Let’s also figure out how we can support FarmKind do what they intend to do for animals. And most importantly, let’s make sure we learn from this campaign.

How can we learn from this experiment?

• Trying new approaches in this complex and relatively new space is great if you thoughtfully measure if it works or not. Measurement and evaluation are especially important because there are backfire risks and because this is a deeply underfunded cause area, so we cannot afford to be careless.
• It can be easy to falsely attribute successes and failures. So, what are some indicators that this might demand pivoting / repeating? I’d love to hear from FarmKind, The Mission Motor, behavioral scientists, and ACE researchers who worked on the Better for Animals resource what they think would give us valuable insights.
• What is the bar for money raised that would make this worth it? What is the cost of FarmKind’s Veganuary campaign, what else could have been done with those funds, how much money is raised through their platform specifically in response to this campaign?
• Can we assess spillover effects?
• Are there some PhD students out here who are willing to work with FarmKind to figure out some RCTs to learn some stuff? E.g. how long do people donate, do they change their diet, what do they think of factory farming, what were their priors, etc.

How can we mitigate possible harms?

• Risk: discussion remains focused on individual diet change, not ending factory farming
  • Can FarmKind, now that they have the attention, redirect their messaging and no longer talk about diets but instead about the horrors of factory farming?
  • Can both vegans/Veganuary and FarmKind state that what they care about is a more hospitable world for all and that industrial agriculture is the enemy.
• Risk: moral circle expansion is slowed
  • Can Toni and FK and participants come out saying something like this, "Don’t get us wrong, we are all actually bleeding hearts, we do care about animals, we don’t think eating animals the way society does now is necessary, natural, or normal, but we are just being pragmatic. We think being vegan is good, but preaching veganism is not."?
  • Can they direct some of the funds they raise to high-impact interventions that do things like education programs aimed at fostering compassion and empathy for animals, anti-speciesist policy work, actions promoting moral consideration of animals in public discourse, etc.?
• Risk: time is wasted on infighting
  • Can both Veganuary and FarmKind state that what they agree on and care about is a more hospitable world for all and that industrial agriculture is the enemy?
  • Can Toni flip-flop some more, and in February say, “You know what, I was wrong. It’s not either/or; it should be both or can be a little bit of each.”?
  • Can FarmKind share the metrics and results of their campaign and show up in vegan spaces like R/vegan to explain their approach and solicit feedback?
  • I think AVA is planning to host a discussion about this at their Summit in Canada in May.
• Risk: fewer people reduce their animal consumption or do it later
  • Meat producers can use this in their propaganda; can we use AI to find the conversations about this that misrepresent the arguments and counteract them?
  • Can Toni and other former vegans come out and say something like, “Actually, after having hung out with all of these meat eaters and learning more about where their food comes from and having seen what it does to their bodies, I think it’s actually kinda gross/disgusting/unsympathetic. I’m happy they donate, but for their sake, I hope they eventually put their mouths where their money is.”
  • What would happen if Veganuary went on offense with aggressive angles like:
    • “We applaud that FarmKind offers all the weak-willed meat-addicts out there a compassion cheat code against animal cruelty. We do hope that the people who listen to Toni and FarmKind’s advice 1) also talk with their doctors and nutritionists and 2) learn about the hidden truths about factory farming.”
    • “We agree that there are multiple roads that lead to Rome, and the super-highway is one where we both do no harm and reduce harm as much as possible. So, we actually already recommend to people who participate in Veganuary also donate to high-impact pro-animal charities. Yes, we are even more holier-than-thou than you thought. We hope vegans put their money where their mouths are. And we hope that offsetters eventually put their mouths where their money is, for animals’ sake and their own.”
    • "How do you know someone is a meat eater? They will tell you. (And they're more likely to need GLP-1.)"

    • “If you’re not one of these privileged people who can buy humanely raised meat and donate money, remember that beans are healthy, cheap, and cruelty free.” 

      (I don’t particularly endorse any of these messages, but I could see people pulling up a chair and a popcorn bucket to watch this while being exposed to different arguments based on the same premise, that farming cruelty is bad.)

How can we increase the likelihood of success?

• Opportunities to increase donation conversion
  • Is there a possibility for a follow-up press release by FarmKind or a pitch with testimonials of carnists who have made donations?
  • What would happen if FarmKind dares vegans and Veganuary supporters to donate? Can they do a donation contest with Veganuary? ACE can probably set up a fundraising page for vegans if Veganuary doesn’t want to do it on the FarmKind site. (Happy to credit FarmKind for those donations, but I’d like them to go where they are likely to do the most good.)
  • Can Toni share where she donates to?
  • Can we leverage the comment sections to encourage people to share where they donate and include donate links?
• Opportunities for awareness increase
  • Can Toni talk about how Veganuary doesn’t talk about animals enough and too much about health and climate, and how the big problem is factory farming?
  • Can FarmKind include and promote people in their pitches who also started reducing their meat intake after learning more about factory farming?
  • Can FarmKind or Toni talk about small-bodied animals and their Shrimpact work? What if Toni says, “Sure, maybe it’s okay if some of these people want to eat some red meat and offset their donations, as long as they don’t start eating chicken or salmon, or eggs.”

There are probably more and more productive ways to help FarmKind and Veganuary and the whole EAA movement in this endeavor. Please share your ideas. Also, what will you do this January, donate, go vegan, or both?

Three final thoughts that I didn’t really know where to put:

1. If we think AI can soonish solve some of the big alt-protein questions (taste, scaling, price, etc.), then we will still need people to stop thinking they need animal products. If we think public discussions will affect alignment, then we need pro-animal messaging to be out there. I’m wondering if this means that hard-to-measure interventions toward increased prevalence of anti-speciesist values might have become more important than I thought they were. On the other hand, if we think AI will solve factory farming, maybe in the meantime we need to focus as much of our time as possible on increasing the welfare of animals who are farmed until then, and that’s more likely done through welfare campaigns than promoting veganism. Either way, we should probably be careful in how we talk about vegans and bring animals up more often, even in meat reduction work. However, I’m very uncertain about all of this and curious what you think.
2. What could this offsetting approach to donating mean for effective giving? Is there a way to leverage this work to get people to make GWWC pledges or to get offsetters to think about how they use their donations in general. FarmKind wasn’t successful in becoming the Giving Multiplier for animals and pivoted to offsetting, but maybe they can still direct offsetters to the Giving Multiplier?
3. I work fulltime in animal advocacy. I don’t think that gives me an excuse to eat animals. I am vegan. I don’t think that absolves me from donating to effective charities to reduce as much harm as possible. It’s a privilege that I can do all three of these things. In this world, few people can. It seems good to encourage people to do everything they can, while also understanding that might be limited. So, let’s help people help more animals as best as they can. We need to understand better what works and work together to make that happen.

Edit: This is my personal take and not Animal Charity Evaluators' opinion.

My top three charities in this election are all animal charities. With Animal Charity Evaluators at the top. Sure, that's in part because I work there. And likely influenced by this donation election happening over a US holiday and during a season when even more animals are abused and slaughtered than usual. That's mentally quite taxing and this is a way for me to deal with that.

But I do not just vote this way, I donate this way. A significant part of my salary goes back to ACE.

ACE influenced at least $12.3 million in total donations in our last fiscal year. That includes $6.4 million in counterfactual gifts. That's a direct consequence of our efforts and vital to reduce the suffering of billions of individuals. I see this as evidence of ACE's potential; I think we can do much, much more to engage people who dislike animal abuse to help farmed and wild animals effectively. To do that, the organization needs to be able to invest a bit more in growth.

Last year, for every dollar ACE spent on our charity evaluations and recommendations, we generated $6.05 in donations for highly effective animal charities that wouldn’t have been donated to otherwise. (While a strong multiplier is encouraging, what ultimately matters for animals is the total amount of additional funding directed to effective work—$6.4 million in counterfactual donations. A hypothetical organization could have a multiplier of 100x but only influence $10,000; we’d rather have a lower multiplier and move millions more dollars to where they’ll help animals most. So, don't be too (un)impressed by that number.)

Because of ACE, there are animal lovers who now donate to more effective charities, people who now help more animals. Some of them would otherwise not have given to animal welfare at all. That means more piglets, squabs, chicks, calves, and shrimplets have a chance at a decent life. 

You can read more about ACE's meta-fundraising impact and how it was calculated here: 

Announcing our Latest Influenced-Giving Metrics - Animal Charity Evaluators

Having followed this project for a few months now, I've been excited about its progression. Besides the aims listed here, there seems potential for secondary benefits in the industry.

Hi, would you help me interpret your writing style so I can better understand? No (fast) answer expected at all.

To me, this post reads as written with a lot of confidence and urgency, giving a sense of certainty. 

Is this a stylistic choice to create engagement (which could either mean increased readership or a discussion aimed at learning)? Or is this because you are completely convinced? Or time pressure or something else? 

Thanks for explaining!

Interesting approach to earning to give. I only see farmed animals in your pictures though, a cow and a chicken, no companion animals. Regardless, or maybe because of that, I suspect ACE's Recommended Charity Fund is possibly the best donation opportunity for your clients.

Hey Kuhan, I really liked this. Thanks for writing it. It led me to think a bit about how this applies to animal welfare.

What I really like about this, is how your thought experiment encourages altruists to think from the perspective of those they’re trying to help. That principle doesn’t just help humanize EV, it can also help with creating willingness to help individuals regardless of the cause of their suffering. An animal living in a fire zone probably doesn’t care if you’re helping them because humans are to blame or if nature is.

One of the difficulties in animal welfare (but maybe also in other cause areas I don't understand as well) is how uncertain probabilities are in many interventions, not just of success but also of potential backlash or other negative outcomes. 

The other challenge when I try to apply this for animal welfare is that thinking from the individual’s perspective doesn’t necessarily result in the highest EV intervention. A hen suffering in a cage now might prefer a really low chance of a sanctuary rescue over a corporate welfare campaign that likely affects more chickens because the latter will only affect future birds. (And this thought-experiment hen could be super kind and utilitarian, but at a certain point I expect excruciating pain will be the main decision-making factor.) You can make an honest and strong values argument about how we must be willing to weigh the welfare of current and future hens equally, but some of the rhetorical power is lost.

An animal example (which Claude helped me come up with) that I think could potentially work for animal grantmakers is to imagine you’ll be born as a salmon into aquaculture who knows where. A funder can either:

• Fund a program that will definitely improve stunning at one facility, affecting 50,000 salmons
• Fund R&D that has a 20% chance of getting stunning technology adopted across the whole industry, affecting many millions of salmons

I’d prefer not to be born as a salmon at all—and if I were, I might rather die before reaching the smolt phase—but if I knew I’d make it to slaughter age I would hope, depending on the details and the certainty of that 20%, the grantmaker would fund the R&D.

Thanks for writing this Abraham! 

Charity evaluators should think about their impact as partially just “moving money around,” not counterfactual donations.

In simple terms, I see Animal Charity Evaluators doing two things related to this topic:

1. Get donors to say no to good grants so they can say yes to great ones (the moving money around on the margin stuff)
2. Creating impact from being the only reason good work happens (the counterfactual stuff)

As an evaluator that aims to help people help more animals, I currently think this approach will accelerate the journey to a better future. And the faster we get to that kinder world, the more suffering is reduced. It's been tricky to assess though when working on the first becomes opportunity cost to the second to a degree that we're doing less good than we otherwise would. 

Examples of work under one:

• Increasing the quality and rigor of evaluations.
• Strengthening grantmaking decisions and distribution size criteria.

Examples of work under two: 

• Influencing conventional animal donors to dedicate a portion of their giving portfolio to effective animal advocacy.
• Getting conservation-minded animal grantmakers to incorporate wild animal welfare considerations.
• Encouraging traditional grantmakers to incorporate effective giving principles within their cause area.

Though you can only get so far with certain traditional funders in terms of scale (some people just really love pretty predators or donkeys), recently I've seen many are actually craving ways that allow them to do more within their field and to clarify their decision-making internally and to their grantees.

An additional benefit of interacting more with conventional donors is that besides their giving behavior, their moral circle might expand. Which seems particularly relevant for animal advocacy.

(BTW not just GiveWell is trying to calculate their counterfactual, ACE is making an attempt too. The report of the most recently completed fiscal year will actually be posted in the next few weeks, the last one can be read here. But yeah, we found it's pretty difficult to do.)

One more thought: I think your points also apply to the recruitment of board members.