Ethical Veganism is Wrong

The main points for ethical veganism as I understand it are:

1. Killing other animals is unjust aggression; you wouldn't like to be killed and eaten, so don't kill and eat them.

2. Factory farming causes animals to have bad lives.

My answer to these arguments:

1a. In a modern market economy, buying farmed meat causes more deaths by causing more animal lives. The ethical vegan must therefore decide whether their objection is to animals dying or to animals living. The question reduces to whether they'd be more glad to have been born than sad to die. Buying wild-caught game does cause a death, but if the animals in question aren't being overhunted / overfished, the counterfactual is that some other equilibrating force acts on the population instead. If you're really worried about reducing the number of animal life years, focus on habitat destruction - it obviously kills wildlife on net, while farming is about increasing lives. The remedy is to promote and participate in more efficient, less aggressive patterns of land usage, which would thereby also be less hostile towards other humans. I'm on the record as interested in coordinating on that. It's a harder problem because it requires prosocial coordination in a confusingly low-trust society pretending to be a high-trust society, but just because a problem is hard to solve doesn't mean we should substitute an easier task that is superficially similar but unhelpful.

1b. Another way of interpreting argument 1 for ethical veganism invokes rights: we shouldn't kill other agents because this violates decision-theoretic principles about respecting agency. But this assumes the other party can engage in the kind of reciprocal decision-making that grounds such rights. Most animals' decision processes don't mirror ours in the way needed for this kind of relationship - they can't make or honor agreements, or intentionally retaliate based on understanding our choices. The question returns to welfare considerations: whether their lives are net positive.

1c There's a third argument sometimes offered, which I think muddles together a rights-based and utilitarian perspective: the instrumentalization of animals as things to eat is morally repugnant, so we should make sure it's not perpetuated. This seems to reflect a profound lack of empathy with the perspective of a domesticate that might want to go on existing. Declaring a group's existence repugnant and acting to end it is unambiguously a form of intergroup aggression. I'm not arguing here that domesticates' preference to exist outweighs your aesthetic revulsion - I'm just arguing that under basic symmetry considerations, the argument from "moral" revulsion is an argument for, not against, aggression.

2. If factory farming seems like a bad thing, you should do something about the version happening to you first. The domestication of humans is particularly urgent precisely because, unlike selectively bred farm animals, humans are increasingly expressing their discontent with these conditions, and - more like wild animals in captivity than like proper domesticates - increasingly failing even to reproduce at replacement rates. This suggests our priorities have become oddly inverted - we focus intense moral concern on animals successfully bred to tolerate their conditions, while ignoring similar dynamics affecting creatures capable of articulating their objections, who are moreover the only ones known to have the capacity and willingness to try to solve problems faced by other species.

11 thoughts on “Ethical Veganism is Wrong

  1. Doug S.

    I do not understand your objection to point 2. Humans' dissatisfaction with their own lives and the human fertility rate seems to have little relevance to whether or not factory farmed animals have bad lives. Trying to argue "misplaced priorities" is also irrelevant, because there's no actual conflict between ethical vegetarianism and solving human problems: abstaining from animal products produced by factory farming to the extent practical has absolutely no effect on one's ability to "[do]something about the version happening to you" and vice versa.

    Reply
    1. Benjamin Ross Hoffman

      If factory farmed animals experienced more avoidance than approach or rest behaviors, their growth and reproductive capacities would noticeably suffer. This is happening for some of them (especially broiler chickens), but it seems like empirically humans are having an even worse time.

      Eating a bad diet makes people worse at things. Time and money spent figuring out how to make a diet restricted for reasons other than health and performance into a good one are time and money not spent on something else.

      Reply
      1. Ákos Kőrösi

        I agree with most of the post, but I'm not sure about 1b.

        It seems possible and even somewhat intuitively compelling for a decision theory to imply not harming less capable agents/beings (at least, above a certain threshold like sentience or capacity to suffer) and seeking retribution if someone harms them, since that makes similar but more capable agents not harm me.

        By revealed preferences, I don't seem to follow a decision theory of this form, and I'm not entirely sure it really is coherent if you think through all of the implications. I just wanted to flag 1b as a point that seemed weak to me.

        Reply
        1. Benquo Post author

          I think this can only work as a simple decision-theoretic argument if limited to the set of agents making decisions on the basis of explicit decision-theoretic arguments. Even so, unclear you get to incept this in higher-level agents; that seems like wishful thinking. It would make more sense to reason about the likely distribution of higher-level agents with influence over you (e.g. future civilizations, simulation overseers), and what they care about, and then check whether any of that has decision-theoretic implications for you, e.g. Robin Hanson's How to Live In A Simulation.

          Reply
          1. Ákos Kőrösi

            >Anarchism seems more important and urgent than veganism if we’re trying to align our social coordination so it can make aligned AIs. Quakerism has been another attempt.

            Strongly agreed. I also agree with most of your responses.

            I really like that the framework of this post and the next one treats predation and domestication as two subcategories of “resource extraction from another organism by exploiting its flaws”, distinguished by driving capacity increase vs. reduction in the target species. This clarified some stuff for me.

            By the way, do you have a detailed model of the very beginnings of human-human domestication? I feel like I know some stuff about prehistory, I know some stuff about early eras of civs, but there’s a “BAM, there are monarchs and priests now” discontinuity in my map. There's some handwavy filler, but nothing gears-level. A detailed account, with as much granularity as possible, of how this subjugation happened could be useful. Know of any good sources on this?

            As for my earlier comment and your responses, beyond the overall agreement with your position, it seems to me that your framework is applicable to those intelligent agents that satisfy two nontrivial assumptions:
            A) The agent is a member of a “species” that resulted from a long iterative selection process across many members
            B) The most efficient energy source in the “ancestral environment” of that species was other, somewhat intelligent, forms of life.

            When thinking about stronger agents that might causally influence me, I usually consider three main categories:
            1) AGIs created by humans
            2) Alien civilizations (or more likely their von Neumann probes at first)
            3) Simulation overseers

            (I guess one might add acausal trade partners, but I don’t think about that sort of stuff often)

            It seems to me that premises A) and B) probably don’t apply to AGIs and von Neumann probes. Unclear what this implies exactly though. Minimally, it suggests that our intuitions about predatorial intelligence might not be directly applicable to them. For example, finding exploitable errors in the cognition of other agents might not be as salient for them as it is for us. But also, the danger posed by stronger intelligences isn’t limited to narrowly defined predation – they could want things like destroying the Earth or the Sun for resources.

            The framework seems to suggest a corollary like: to ~ensure that one won't be harmed by a stronger rational agent, one must be able to produce some sort of output whose utility exceeds that of destroying any necessary condition of one’s flourishing. (Generally, there's also an option to develop adequate defensive capabilities, but that might not be possible against a sufficiently strong other agent). Also, the other agent needs to recognize us as capable of this sort of trade, which is not automatically guaranteed. And of course there might be strong agents who still wouldn't harm us even if we don't provide much utility for them, but I agree that it's wishful thinking to expect that, and it's unstable anyway.

          2. Benquo Post author

            Seems like you're arguing that while predation was evolution's way of bootstrapping intelligence and consciousness without requiring pre-existing intelligence or explicit cooperation, this framework won't help us with AI alignment because AIs won't evolve through these selective pressures. Does this seem like a fair summary?

            One implication of this is that designed AGIs and von Neumann probes are unlikely to have intelligence optimized for potentially adversarial agent-modeling. If we're the first meaningful adversary they meet, then they start out adversary-blind, which would make it easier to resist or outmaneuver them than comparably big and smart agents that evolved through predator-prey dynamics.

      2. Ashwin S.

        Animals in a factory farm are not the same as animals in captivity. If the animals are being raised as food, then they would necessarily need to have a high reproductive rate to ensure that the factory farm doesn't run out of animals: for example by selecting for animals with higher reproductive capacities and by artificial insemination. I don't see how comparing the reproductive capacities of animals in factory farms with those of humans says anything about their relative quality of lives. Similarly for their growth rates.

        The quality of life in factory farms can vary depending on the animal and the standards of a particular farm, but I don't think there's a reasonable argument that humans may have it worse. For example, broiler chickens have been bred to grow extremely quickly, resulting in many of them not being able to support their own weight and having heart, lung, and other health issues. Factory farms will debeak these chickens to prevent them from pecking each other due to the unnatural conditions they are in, which often causes acute and possibly chronic pain. Millions of chickens die before they even reach the slaughterhouse (https://animalequality.org/news/chicken-mortality-report/#flipbook-report-mortality-chickens/1/). This doesn't seem consistent with a view that these animals have been "successfully bred to tolerate their conditions".

        Would you say that in expectation a random human would rather swap lives with a random animal in a factory farm if given the chance?

        Reply
        1. Benquo Post author

          The "swap lives" hypothetical doesn't work because it conflates two different questions:

            Whether a given farm animal's life is worth living (compared to non-existence)
            Whether a human would prefer to live that life compared to their current human life

          These aren't the same thing at all. A mouse might have a life worth living for a mouse, while still being a life no human would want to live, because humans and mice have fundamentally different cognitive architectures, needs, and capabilities. The fact that I wouldn't want to be a mouse tells us nothing about whether being a mouse is net positive for mice.

          You make a fair point about broiler chickens - they do face serious welfare issues from rapid growth. But this is notably one of the worst cases in modern farming. By unit consumed, the central case of factory farming is dairy cows, who fare significantly better. Moreover, the fact that we're collectively choosing to torment broiler chickens at massive and increasing scale suggests our coordination systems are producing outcomes severely misaligned from individual human values and preferences - nearly no one actually wants this outcome. This points to a deeper problem with human coordination capacity that needs to be addressed before we can meaningfully improve animal welfare at scale.

          Reply
          1. Ashwin S.

            The "swap lives" argument was meant to address the point about the conditions of humans being a more urgent priority to fix than the conditions of animals in factory farming. If the conditions of humans are worse and a more urgent issue to address than the those of factory farmed animals, we should expect that a human pressing a button to swap lives with a factory farmed animal would come out ahead. Or am I misunderstanding your response 2 in the original post?

            Dairy cows might fare better than broiler chickens, but even their conditions don't seem particularly inspiring. For example, high rates of lameness, mastitis, infertility (!) - (https://pmc.ncbi.nlm.nih.gov/articles/PMC7125520/).

            I think many people who aren't aware of factory farms would be appalled at the conditions of animals in them. But it's not quite clear to me how misaligned these systems are from human values and preferences at least today. At least empirically, if you give a person the choice of some ethically sourced meat for example versus cheaper meat, there are many who would still take the cheaper meat even if they might be aware of conditions on these factory farms. But yes I suppose that there likely aren't too many people who are eager to have broiler chickens suffer, and there is likely room for better regulation, technology, other coordination mechanisms to improve their welfare.

            In any case, I still don't think this defends your point in (2).

          2. Benquo Post author

            Individual consumer choices aren't good evidence of true preferences about animal welfare - both because it's hard to perceive whether you're making a real difference when reports of suffering remain constant regardless of your choices, and because it's unclear whether I'm being scammed when I buy pastured chicken and eggs. This is precisely the kind of coordination problem I was highlighting in the original post - the gap between what our systems produce and what individuals would prefer if they felt free to decide together.

Leave a Reply

Your email address will not be published. Required fields are marked *