I’ve seen AI-based animal communication technologies starting to be involved in some EA events / discussions (e.g. https://www.earthspecies.org/ ). I’m worried these initiatives may be actively negative, and I’m wondering if anyone has / will articulate a stronger defense of why they’re good?
The high-level argument I’ve heard is that communicating with animals will make humans be more empathetic towards them. But I don’t see why this would be the most likely outcome:
Humans are already fairly empathetic to animals, especially around things that we’d consider important welfare issues. We don’t need a hen to articulately describe why she’d prefer not to have her beak cut off or be kept in a cage, I think it would be fairly obvious to most people.
Animals might become less sympathetic if we knew what they were saying. It seems possible that most of their thoughts and words are about food, sex, and ingroup / outgroup dynamics.
A similar argument is that communication would allow us to see that animals are actually intelligent, but again I don’t see why this is necessarily the case. If their thoughts are things people would generally consider crude, it’s possible people would become more confident in their lack of intelligence (despite still deserving moral consideration).
More importantly, a large effect of being able to communicate with animals is that they’ll become more useful to humans. If animals had political power or legal rights, this might open the door to mutually beneficial trade. But in reality, they don’t have these things, so it seems more likely that this would allow humans to exploit these species more easily. They reason chickens, cows, and pigs are in such a bad state is because they’re very useful to humans, and I’m worried animal communication technologies will subject more species to similar fates.
“Articulate a stronger defense of why they’re good?”
I’m no expert on animal-welfare stuff, but just thinking out loud, here are some benefits that I could imagine coming from this technology (not trying to weigh them up versus potential harms or prioritize which seem largest or anything like that):
You imagine negative PR consequences once we realize that animals might mostly be thinking about basic stuff like food and sex, but I picture that being only a small second-order consequence—the primary effect, I suspect, is that people’s empathy for animals might be greatly increased by realizing they think about stuff and communicate at all. The idea that animals (especially, like, whales) have sophisticated thoughts and communicate, and the intuition that they probably have valuable internal subjective experience, might both seem “obvious” to animal-welfare activists, but I think for most normal people globally, they either sorta believe that animals have feelings (but don’t think about this very much) or else explicitly believe that animals lack consciousness / can’t think like humans because they don’t have language / don’t have full human souls (if the person is religious) / etc. Hearing animals talk would, I expect, wake people up a little bit more to the idea that intelligence & consciousness exist on a spectrum and animals have some valuable experience (even if less so than humans).
In particular, I’m definitely picturing that the journalists covering such experiments are likely to be some combination of 1. environmentalists who like animals, 2. animal rights activists who like animals, 3. just think animals are cute and figure that a feel-good story portraying animals as sweet and cute will obviously do better numbers than a boring story complaining about how dumb animals are. So, with friendly media coverage, I expect the biggest news stories will be about the cutest / sweetest / most striking / saddest things that animals say, not the boring fact that they spend most of their time complaining about bodily needs just like humans do.
Compare for instance “news coverage” (and other cultural perceptions of) human children. To the extent that toddlers can talk, they are mostly just demanding things, crying, failing to understand stuff, etc. Yet, we find this really cute and endearing (eg, i am a father of a toddler myself, and it’s often very fun). I bet animal communication would similarly be perceived positively, even if (like babies) they’re really dumb compared to adult humans.
Talk-to-animals tech also seems potentially philosophically important in some longtermist, “sentient-futures” style ways:
What’s good versus bad for an animal? Right now we literally just have to guess, based on eyeballing whether the creature seems happy. And if you are less of a total-hedonic-utilitarian, more of a preference utilitarian, the situation gets even worse. It would be nice if we could just ask animals what their problems are, what kind of things they want, etc! Even a very small amount of communication would really increase what we are able to learn about animals’ preferences, and thus how well we are able to treat them in a best-case scenario.
Maybe we could use this tech to do scientific studies and learn valuable things about consciousness, language, subjective experience, etc, in a way that clarifies humanity’s thinking about these slippery issues and helps us better avoid moral catastrophes (perhaps becoming more sympathetic to animals as a result, or getting a better understanding of when AI systems might or might not be capable of suffering).
Perhaps humanity has some sort of moral obligation to (someday, after we solve more pressing problems like not destroying the world or creating misaligned AI) eventually uplift creatures like whales, monkeys, octopi, etc, so they too can explore and comprehend the universe together with us. Talk-to-animals tech might be an early first step toward such future goals, might set early precedents, might help us learn about some of the philosophical / moral choices we would need to make if we embarked on a path of uplifting other species, idk.
Are there also just, concerns about misinterpretation? There’s not really a good way of checking baselines on hallucinations or unconfident predictions from the AI, since 99% of humans don’t know what these sounds mean.
Furthermore, since these seem to be based on the behaviours humans observe co-occurring with the communication, they’d necessarily be lower-fidelity than that animal’s thought process (as you note). That seems a bit lame and the website certainly isn’t trying to dispel their own mythmaking around ‘talk to animals’, which isn’t really what’s happening here in any meaningful sense.
Hey, useful question! This is probably the strongest case I’ve seen, focused on the potential legal implications. But it still rests on the assumption that opening the door for legal recognition of certain charismatic species (like whales) will spill over to other species that people are primed to disregard (like chickens), which definitely isn’t a given. If you’ve not already seen it, the MOTH Project at NYU has also come up with a list of principles that seem like a useful guide for anyone doing work in this space.
On the empathy argument: In theory it could still be more persuasive, and harder for people to ignore, if animals were able to more directly communicate feelings of pain, stress, isolation, etc. And this could also reinforce people’s understanding of animals as individuals if different animals express different emotions in similar situations, which in theory could improve their empathy towards them. But yeah, this depends on a lot of things going right: e.g., the technology accurately relaying animals’ communications, animals coming across as complex/sympathetic, etc., and this being relayed by credible people to key audiences in a persuasive way. There’s likely to be a lot of misleading noise in this area as well—maybe you’ll get some big industry players commissioning studies that ‘translate’ animals’ communications in humane-washing ways. And on top of the points you raised, using this kind of tech on e.g. farmed chickens will give a skewed perspective on the intelligence of a typical chicken, given that farmed meat chickens are killed at a few weeks old so you’d be listening to the equivalent of toddlers who’ve been raised in abnormal conditions that probably stifle normal development.
Broader AI uses for understanding animals better could still be really useful (e.g. by enabling more sophisticated bio-sensors and tracking devices, and the software required to process and interpret all the resulting data), so it would be helpful to pivot people more to thinking about all the messier ways that AI could help us understand animals’ lives on their own terms rather than looking for a clean animal-to-human translation. Mal Graham has written about some of the potential applications in this post.
It seems like there’s recently been a noticeable uptick in the quality and quantity of animal-related posts by group like Rethink Priorities, Animal Ask, and many others. This puts the movement in a much better place than just a few years ago where it was very hard to know how to effectively help animals.
Just wanted to say this is awesome, and keep up the good work!
I’ve seen AI-based animal communication technologies starting to be involved in some EA events / discussions (e.g. https://www.earthspecies.org/ ). I’m worried these initiatives may be actively negative, and I’m wondering if anyone has / will articulate a stronger defense of why they’re good?
The high-level argument I’ve heard is that communicating with animals will make humans be more empathetic towards them. But I don’t see why this would be the most likely outcome:
Humans are already fairly empathetic to animals, especially around things that we’d consider important welfare issues. We don’t need a hen to articulately describe why she’d prefer not to have her beak cut off or be kept in a cage, I think it would be fairly obvious to most people.
Animals might become less sympathetic if we knew what they were saying. It seems possible that most of their thoughts and words are about food, sex, and ingroup / outgroup dynamics.
A similar argument is that communication would allow us to see that animals are actually intelligent, but again I don’t see why this is necessarily the case. If their thoughts are things people would generally consider crude, it’s possible people would become more confident in their lack of intelligence (despite still deserving moral consideration).
More importantly, a large effect of being able to communicate with animals is that they’ll become more useful to humans. If animals had political power or legal rights, this might open the door to mutually beneficial trade. But in reality, they don’t have these things, so it seems more likely that this would allow humans to exploit these species more easily. They reason chickens, cows, and pigs are in such a bad state is because they’re very useful to humans, and I’m worried animal communication technologies will subject more species to similar fates.
“Articulate a stronger defense of why they’re good?”
I’m no expert on animal-welfare stuff, but just thinking out loud, here are some benefits that I could imagine coming from this technology (not trying to weigh them up versus potential harms or prioritize which seem largest or anything like that):
You imagine negative PR consequences once we realize that animals might mostly be thinking about basic stuff like food and sex, but I picture that being only a small second-order consequence—the primary effect, I suspect, is that people’s empathy for animals might be greatly increased by realizing they think about stuff and communicate at all. The idea that animals (especially, like, whales) have sophisticated thoughts and communicate, and the intuition that they probably have valuable internal subjective experience, might both seem “obvious” to animal-welfare activists, but I think for most normal people globally, they either sorta believe that animals have feelings (but don’t think about this very much) or else explicitly believe that animals lack consciousness / can’t think like humans because they don’t have language / don’t have full human souls (if the person is religious) / etc. Hearing animals talk would, I expect, wake people up a little bit more to the idea that intelligence & consciousness exist on a spectrum and animals have some valuable experience (even if less so than humans).
In particular, I’m definitely picturing that the journalists covering such experiments are likely to be some combination of 1. environmentalists who like animals, 2. animal rights activists who like animals, 3. just think animals are cute and figure that a feel-good story portraying animals as sweet and cute will obviously do better numbers than a boring story complaining about how dumb animals are. So, with friendly media coverage, I expect the biggest news stories will be about the cutest / sweetest / most striking / saddest things that animals say, not the boring fact that they spend most of their time complaining about bodily needs just like humans do.
Compare for instance “news coverage” (and other cultural perceptions of) human children. To the extent that toddlers can talk, they are mostly just demanding things, crying, failing to understand stuff, etc. Yet, we find this really cute and endearing (eg, i am a father of a toddler myself, and it’s often very fun). I bet animal communication would similarly be perceived positively, even if (like babies) they’re really dumb compared to adult humans.
Talk-to-animals tech also seems potentially philosophically important in some longtermist, “sentient-futures” style ways:
What’s good versus bad for an animal? Right now we literally just have to guess, based on eyeballing whether the creature seems happy. And if you are less of a total-hedonic-utilitarian, more of a preference utilitarian, the situation gets even worse. It would be nice if we could just ask animals what their problems are, what kind of things they want, etc! Even a very small amount of communication would really increase what we are able to learn about animals’ preferences, and thus how well we are able to treat them in a best-case scenario.
Maybe we could use this tech to do scientific studies and learn valuable things about consciousness, language, subjective experience, etc, in a way that clarifies humanity’s thinking about these slippery issues and helps us better avoid moral catastrophes (perhaps becoming more sympathetic to animals as a result, or getting a better understanding of when AI systems might or might not be capable of suffering).
Perhaps humanity has some sort of moral obligation to (someday, after we solve more pressing problems like not destroying the world or creating misaligned AI) eventually uplift creatures like whales, monkeys, octopi, etc, so they too can explore and comprehend the universe together with us. Talk-to-animals tech might be an early first step toward such future goals, might set early precedents, might help us learn about some of the philosophical / moral choices we would need to make if we embarked on a path of uplifting other species, idk.
Are there also just, concerns about misinterpretation? There’s not really a good way of checking baselines on hallucinations or unconfident predictions from the AI, since 99% of humans don’t know what these sounds mean.
Furthermore, since these seem to be based on the behaviours humans observe co-occurring with the communication, they’d necessarily be lower-fidelity than that animal’s thought process (as you note). That seems a bit lame and the website certainly isn’t trying to dispel their own mythmaking around ‘talk to animals’, which isn’t really what’s happening here in any meaningful sense.
Hey, useful question! This is probably the strongest case I’ve seen, focused on the potential legal implications. But it still rests on the assumption that opening the door for legal recognition of certain charismatic species (like whales) will spill over to other species that people are primed to disregard (like chickens), which definitely isn’t a given. If you’ve not already seen it, the MOTH Project at NYU has also come up with a list of principles that seem like a useful guide for anyone doing work in this space.
On the empathy argument: In theory it could still be more persuasive, and harder for people to ignore, if animals were able to more directly communicate feelings of pain, stress, isolation, etc. And this could also reinforce people’s understanding of animals as individuals if different animals express different emotions in similar situations, which in theory could improve their empathy towards them. But yeah, this depends on a lot of things going right: e.g., the technology accurately relaying animals’ communications, animals coming across as complex/sympathetic, etc., and this being relayed by credible people to key audiences in a persuasive way. There’s likely to be a lot of misleading noise in this area as well—maybe you’ll get some big industry players commissioning studies that ‘translate’ animals’ communications in humane-washing ways. And on top of the points you raised, using this kind of tech on e.g. farmed chickens will give a skewed perspective on the intelligence of a typical chicken, given that farmed meat chickens are killed at a few weeks old so you’d be listening to the equivalent of toddlers who’ve been raised in abnormal conditions that probably stifle normal development.
Broader AI uses for understanding animals better could still be really useful (e.g. by enabling more sophisticated bio-sensors and tracking devices, and the software required to process and interpret all the resulting data), so it would be helpful to pivot people more to thinking about all the messier ways that AI could help us understand animals’ lives on their own terms rather than looking for a clean animal-to-human translation. Mal Graham has written about some of the potential applications in this post.
It seems like there’s recently been a noticeable uptick in the quality and quantity of animal-related posts by group like Rethink Priorities, Animal Ask, and many others. This puts the movement in a much better place than just a few years ago where it was very hard to know how to effectively help animals.
Just wanted to say this is awesome, and keep up the good work!