I would argue that the vast majority of the populace agrees that industrialised animal farming is harmful. Defence of meat consumption tends to rest on «these animals live good, relatively natural, relatively free lives, are killed humanely, used completely, and integrated into their local ecosystem, and the humans who work with them know them well and care for them». I know many people who will defend e.g. fishing and hunting deer to promote forest rejuvenation, organic farming, cows who graze on mountain meadows where we cannot farm and their meadows provide biodiversity, chickens who forage in food forests, sheep that stabilise water infrastructure, animals that eat things humans cannot eat, promote soil health, seeing this as a natural and fair trade. But I know practically noone who would defend e.g. battery hens, animals kept in abject misery and captivity their whole lives without sunlight, air or movement while fed foods that humans could directly eat. Characteristically, when people see footage of this, they are disgusted and upset and feel that this is not right. And yet we also know that with industrialised products being much cheaper and easier to access, most people will mostly consume them, even though their rationale for meat consumption uses a far more rare more ethical reference frame.
The issue is that PLF can be trivially implemented in a huge industrialised farm, and also has the most benefits for the farmer there; at a small organic farm, the scale makes it too costly, and the farmers tend not to need it, because they actually have caring relationships with the animals. But this then frames the industrialised PLF farm as the one meeting supposedly objective modern welfare standards with documented proof while the small organic one will not.
Combine that with the fact the PLF often measures metrics that do not reflect choices the animals would make (e.g. wanting to freely roam, even though that risks getting a stone in one's hoof or eating something spiky), but ones farmers are interested in (e.g. weight gain), and the fact that the baseline data for «these animals are fine» is complete bullshit, because it is itself taken in factory farms where they are all depressed, and PLF both furthers the worst kind of animal farming there is, and frames it as much more ethical than it is. And I do think one can absolutely make both a rational and an emotional appeal against that with the public.
Final note: I've been involved on the periphery of PLF projects. They always talk about welfare. You know what they were modelling? How to more efficiently move animals into the fucking slaughter house. Literally, just animals moving down a corridor, each one tagged by an AI, and the speaker happily telling me that they redesigned the area based on this for better flow, which is more efficient and also supposedly keeps them much happier walking into the slaughterhouse cause it is less confusing.
Your figure for insects is off by multiple orders of magnitude. There are far more of them, making this more pressing.
The list of sentience candidates misses a huge relevant group that is still systematically neglected - spiders. Portia spiders are especially intriguing candidates.
The plans to disenhance animals specifically do not seek to suppress reflexive responses to pain, but conscious experience of it, so the intended outcome would actually be animals that show signs of pain they do not experience. (Which has its own severe risks, especially in light of the fact that we are not in fact sure that this result could be achieved, and farmers may well numb as a result to animal pain expressions, a la Descartes the screeching is that of a malfunctioning clock.)
A development I fear in using AI in factory farming to monitor welfare, based on how I have seen this actually researched and implemented, is over-reliance on AI to signal welfare problems, while having even less human oversight for all the animals, which is troubling in light of the fact that the systems trained to recognise suffering animals have very dubious training data. E.g. it is assumed that your average factory farm animal that doesn't look egregiously sad to the current vet is probably doing fine, then volunteers are briefly trained to label data accordingly, have their own speciesist bias affect the results, and then this all gets a fancy AI label and suddenly seems objective when it really is not.
Insect farming is indeed turning into a huge thing, and they currently have no protective legislation, even while we are finding increasing indicators of a capacity for suffering. Insect farming is potentially very cheap and space, water, food turnover, climate etc. friendly, and so may well scale up significantly while there are zero welfare steps taken, and then be economically entrenched and hard to reverse, especially as the public tends not to be sympathetic to such alien creatures and many aspects of it are also quite "green".
Animal communication is also not trivially solvable by just throwing AI at it.
And something relevant people keep forgetting; while decoding wild animal communication is very challenging, teaching non-human animals human language systems (verbal for parrots, sign for primates, or generally working with symbol tables or buttons) has proven surprisingly successful (albeit also ethically disastrous); we've gotten far more communication going that way. (They understand us better than we understand them.) A lot of the consequences and issues that follow have already played out that way.
I would argue that the vast majority of the populace agrees that industrialised animal farming is harmful. Defence of meat consumption tends to rest on «these animals live good, relatively natural, relatively free lives, are killed humanely, used completely, and integrated into their local ecosystem, and the humans who work with them know them well and care for them». I know many people who will defend e.g. fishing and hunting deer to promote forest rejuvenation, organic farming, cows who graze on mountain meadows where we cannot farm and their meadows provide biodiversity, chickens who forage in food forests, sheep that stabilise water infrastructure, animals that eat things humans cannot eat, promote soil health, seeing this as a natural and fair trade. But I know practically noone who would defend e.g. battery hens, animals kept in abject misery and captivity their whole lives without sunlight, air or movement while fed foods that humans could directly eat. Characteristically, when people see footage of this, they are disgusted and upset and feel that this is not right. And yet we also know that with industrialised products being much cheaper and easier to access, most people will mostly consume them, even though their rationale for meat consumption uses a far more rare more ethical reference frame.
The issue is that PLF can be trivially implemented in a huge industrialised farm, and also has the most benefits for the farmer there; at a small organic farm, the scale makes it too costly, and the farmers tend not to need it, because they actually have caring relationships with the animals. But this then frames the industrialised PLF farm as the one meeting supposedly objective modern welfare standards with documented proof while the small organic one will not.
Combine that with the fact the PLF often measures metrics that do not reflect choices the animals would make (e.g. wanting to freely roam, even though that risks getting a stone in one's hoof or eating something spiky), but ones farmers are interested in (e.g. weight gain), and the fact that the baseline data for «these animals are fine» is complete bullshit, because it is itself taken in factory farms where they are all depressed, and PLF both furthers the worst kind of animal farming there is, and frames it as much more ethical than it is. And I do think one can absolutely make both a rational and an emotional appeal against that with the public.
Final note: I've been involved on the periphery of PLF projects. They always talk about welfare. You know what they were modelling? How to more efficiently move animals into the fucking slaughter house. Literally, just animals moving down a corridor, each one tagged by an AI, and the speaker happily telling me that they redesigned the area based on this for better flow, which is more efficient and also supposedly keeps them much happier walking into the slaughterhouse cause it is less confusing.
Multiple corrections/additions.
Your figure for insects is off by multiple orders of magnitude. There are far more of them, making this more pressing.
The list of sentience candidates misses a huge relevant group that is still systematically neglected - spiders. Portia spiders are especially intriguing candidates.
The plans to disenhance animals specifically do not seek to suppress reflexive responses to pain, but conscious experience of it, so the intended outcome would actually be animals that show signs of pain they do not experience. (Which has its own severe risks, especially in light of the fact that we are not in fact sure that this result could be achieved, and farmers may well numb as a result to animal pain expressions, a la Descartes the screeching is that of a malfunctioning clock.)
A development I fear in using AI in factory farming to monitor welfare, based on how I have seen this actually researched and implemented, is over-reliance on AI to signal welfare problems, while having even less human oversight for all the animals, which is troubling in light of the fact that the systems trained to recognise suffering animals have very dubious training data. E.g. it is assumed that your average factory farm animal that doesn't look egregiously sad to the current vet is probably doing fine, then volunteers are briefly trained to label data accordingly, have their own speciesist bias affect the results, and then this all gets a fancy AI label and suddenly seems objective when it really is not.
Insect farming is indeed turning into a huge thing, and they currently have no protective legislation, even while we are finding increasing indicators of a capacity for suffering. Insect farming is potentially very cheap and space, water, food turnover, climate etc. friendly, and so may well scale up significantly while there are zero welfare steps taken, and then be economically entrenched and hard to reverse, especially as the public tends not to be sympathetic to such alien creatures and many aspects of it are also quite "green".
Animal communication is also not trivially solvable by just throwing AI at it.
And something relevant people keep forgetting; while decoding wild animal communication is very challenging, teaching non-human animals human language systems (verbal for parrots, sign for primates, or generally working with symbol tables or buttons) has proven surprisingly successful (albeit also ethically disastrous); we've gotten far more communication going that way. (They understand us better than we understand them.) A lot of the consequences and issues that follow have already played out that way.
- But overall, stimulating read, thank you!