But "what does meat taste like, and how do we replicate it?" is much narrower and much more tractable — and a place where we think the field can be far more rigorous than it has been.

I agree these are important questions! I wonder what you think of the idea that maybe we can create tastes that are superior to any meat, perhaps unrecognizable tastes that no humans have tried, that are irresistible?

My doubt was on the epistemics, and specifically on the estimation of welfare gain by an intervention.

Re: Benatar's view. He holds the view that the continuation of a life accrues harm. At the same time, he indeed also holds that it is overall better (or more like, less bad) for people's lives to continue once they have started, because death is even more significant harm.

I can't say how many anti-natalists are utilitarians of any sort, or the reverse. I am pretty sure many negative utilitarians think that the continuation of any sentient life is net negative.

Going back to Benatar's view and applying it to our subject matter. He would likely claim that:

1. Continuation of the lives of third-world children is a harm in itself, both because of the expected negative welfare, and also for some other non-utilitarian reasons.
2. Nonetheless, letting them die is still overall bad, because dying is an even greater harm. 
   
    

Thank you very much for the post! I have read some comments (and except for Cynthia's, mostly incompletely). I want to leave a comment that is meant to be a reply to some comments, but also possibly the post itself:

Some comments in the discussion, and perhaps the post implicitly, seem to treat global health as a point of high certainty — Lewis described it as "the closest to total certainty of positive impact of any areas." But I think that "certainty" is partly an artefact of where we stop scrutinising. Yes, we have strong evidence that bednets counterfactually avert statistical deaths. But we have much weaker evidence that the counterfactual life thereby preserved is net-positive over its remaining course, and weaker still that it's more net-positive than the resources spent would have produced elsewhere (even limiting the resources within just humans, or the global health cause). That second layer — the value of the outcome, not the efficacy of the intervention — usually gets carried by unstated assumptions rather than by data. (FWIW, part of the assumptions are philosophical. For instance, there are serious philosophers who think that each extra life year is a net-negative, regardless of people's preferences. Also, people who have their lives saved might go on to harm other humans, but some EAs and ethicists think we ought not consider this when it comes to saving kids.)

I want to be careful not to overstate this, because there are disanalogies: The human prior is genuinely stronger on the immediate impact level. It also seems that on the secondary or further levels, interventions targeting humans are often even less certain than AW ones. So I'm not claiming the two are equally uncertain. 

But the (in)consistency point still bites. If AW has a major evidence problem vis-à-vis whether overall welfare was indeed improved, life-saving human interventions have it too — it just happens that we rarely turn the skepticism in that direction (welfare).

P.S. I'm aware of problems raised by population ethics and the meat-eating problem (it's a more productive framing than the version you heard), so not a novel observation in general. I'm raising my points narrowly because the comment section (or maybe just Lewis, and David Reinstein, plus the post implicitly?) leans on global health being the secure benchmark, in comparison to AW interventions seeking to improve welfare.

P.P.S. I used Claude 4.8 to help me check whether my points were already made by someone else here, and to help me draft the reply, of which I modified.

Thanks for the nudge! I have something really important coming up in July, possibly the most important thing I might do so far in my career. I will consider after that. Feel free to nudge me again in August! 

Yes, thanks for the reminder. I have long (incorrectly) thought pond loach is just one species, until Ryan pointed out that there are at least 4 (but seems like only two are commercially popular). 

From what I learned, even though mud carp should be the biggest used fry for mandarin fish feed, many other species such as other carps and tilapia are also used in significant amounts. But in terms of cause priortization/conceptualization, grouping them together makes perfect sense!

Yes feeder fish for mandarin fish is a big category. But my understanding is that many species' fry are used, including pond loaches (rarely though, I believe). I am not sure the majority of them are one species (i.e. mud carp). 

Also, since we need to count fry to come up with a few trillion figure for feeder fish for mandarin fish, we also need to count pond loach numbers by the fry stage. Estimates of pond loach survival rates from fry to sellable fish vary widely, from 2% to 10%. Given that the number of pond loach slaughtered each year is roughly 10B, that's ~100B-500B pond loaches slaughtered each year. 

Yeah, maybe mud carp or some other species is no.1, but I am guessing it is also possible pond loach is still no.1 or close.

Thanks for asking, here are some ideas I personally wish funders would consider at least investigating. The epistemic status of some of these ideas is not great, and I never attempted any robust analysis on the expected values of these potential interventions/causes, but I hope they are worth investigating. 

• Stop/slow down the development of remotely controllable insects (not just for AW reasons; such a tech can be misused for surveillance, military uses, and bioterrorism)
• Stop the spread of caged broiler systems, particularly in African countries, and other LMICs.  
• Help the most (probably) farmed fish on earth (in terms of number of individuals, counting from actual quantities sold), pond loaches, who suffer from mortality rates like 20-80%, stressful and long transports, and long, excruciating slaughters.
  • FWI is investigating potential interventions, both on the on-farm condition, and slaughter
• Using "Virtual Control Groups" to reduce the number of animals used in the control groups in preclinical studies and basic research
  • think about it, the control group should be similar across different experiments using the same strain of animal. But animals in the control arm also suffer, because they have to be captivated in small spaces, eat boring food, fed "vehicle chemicals" meant to carry the drug in the test arm, and also have their blood drawn. So the idea is: why don't we reuse historical control group data (it's more complicated than that, but that's the spirit)
• Building an LLM-based agent that can allow regulators and drug discovery researchers to have an easy-to-use AI agent with a text-based chat interface (credit to Alexandra Hammond, who developed the idea with me). This idea specifically tackles the problem that many regulators, and some researchers, are reluctant to use ML/DL models that can predict toxicities of chemicals to replace animal tests because there are too many different models, with each of them working for only a narrow range of chemicals. This idea also partly solves the "blackbox concern" excuse used by some regulators to not adopt AI alternatives to animal tests. This idea might be achievable from two different architectures, or the combination of them
  • Allow the LLM-based agent to access and utiliize various ML/DL models that can be found online, open-sourced, and either teach researchers and regulators how to apply them to certain kinds of chemicals, or even just give the prediction directly in the chat.
  • Train a multi-modal LLM, with one of its modes being toxicity data.
• It seems to me that many alt-protein companies, particularly cultivated meat ones, are interested in using AI to take their research to a new level. But since each of these companies data are proprietary, none of them can get their models to become a large model. What if we fund an open-source large database for alt proteins which everyone can use to train models to help their work. Or maybe we even just fund both the database and the training of a "large-altpro-model".
• AI to help wild animal welfare. A particular intervention that might serve many values, such as signalling effect, signposting, experimentation, and piloting, is to use AI-controlled drones that can identify animals who are in serious injuries or diseases so bad that it means they will not survive for long (and will die from starvation, dehydration, being eaten alive, or multiple organ failures). The drone would have the capability to euthanize the animal on the spot, preferably with methods that would not leave a trace (e.g. if a chemical was used, it has to be non-toxic to at least predators and scavengers nearby). The design is to try to make sure there is as little counterfactual impact on the situation, other than the alleviation of suffering by killing the animal sooner
  • Note: My original idea was to even limit the use case to animals with broken spines or multiple broken limbs so that counterfactually, not even the location of the death would differ.
• Supporting the research of using AI-controlled drones, preferably of insect sizes, for investigating cruelty in factory farms
  • Of course, don't use remote-controllable insects.
  • The idea could be expanded to investigating any animal cruelty cases
  • Controversial and risky, socially, politically, and legally

Thank you for the post!

Animals are not one issue

I have long wished to write or see an EA forum post about this. Thanks for doing it!

I don't understand the negative karma for a comment that has 5 agreement votes to 0 disagreement. I wonder if some of you can explain the reason for agreeing but downvoting this?

Thank you for the great post! I don't have comments right now on your subject matter, but want to provide a bit of information about my region.

and then trying to steer the conversation as quickly as possible back towards safer, intuitive examples, like malaria nets, lead poisoning, factory farms, biosecurity, etc.

In China, my sense is that factory farming is not a "safe" option when it comes to explaining EA, cause prioritization, etc to non-EAs or even new EAs. On the other hand, for some reason, AI seems to be much more "safe" than factory farming.