On Fish Suffering

Vegan message board for support on vegan related issues and questions.
Topics include philosophy, activism, effective altruism, plant-based nutrition, and diet advice/discussion whether high carb, low carb (eco atkins/vegan keto) or anything in between.
Meat eater vs. Vegan debate welcome, but please keep it within debate topics.
User avatar
brimstoneSalad
neither stone nor salad
Posts: 10273
Joined: Wed May 28, 2014 9:20 am
Diet: Vegan

Re: On Fish Suffering

Post by brimstoneSalad »

carnap wrote: Fri Mar 16, 2018 1:09 am Why would operant conditioning prove interests? If that is true than means basic robots would have interests as well and may have consciousness.
Correct.
carnap wrote: Fri Mar 16, 2018 1:09 amComputationally all that is required for operant conditioning is that two variables can become associated based on "positive" or "negative" outcomes which can be evaluated in any way.
It is in discovering those negative or positive evaluations that we probe the interests of the robot (programmed though they may be).
carnap wrote: Fri Mar 16, 2018 1:09 am
brimstoneSalad wrote: Thu Mar 15, 2018 1:43 pm Which is why I prefer to talk about sentience and interests.
Sentience is related to consciousness, its no more well understood than consciousness. "Interests" refers to various concepts depending on the context.
Sentience refers more specifically to sense experience and cognition of those experiences in some meaningful way (like associative learning). It is at least a bit better nailed down and minimalist.

Consciousness could refer to something more broad, or even a concept of "self-knowledge" or meta-cognition.

carnap wrote: Fri Mar 16, 2018 1:09 amBut we aren't talking about intelligence, we are talking about sentience. Also I didn't suggest its 1:1 but there is no obvious way to compare the sentience of one animal to the next. Does higher intelligence mean the animal has greater sentience? Why?
The only good way to compare something that is fundamentally demonstrated by learning is through just that -- learning -- which is roughly comparable to intelligence.

We can also try to bean count number and variation of interests, but I think intelligence is a good enough indication to make ballpark estimations by.

carnap wrote: Fri Mar 16, 2018 1:09 amAnimals will eat more plants but they don't eat the same plants we do.
Animals fed on grain eat essentially the same plants we do.
carnap wrote: Fri Mar 16, 2018 1:09 amMost insects are killed when insecticides are applied to crops and some crops use more than others, vegetable and fruit crops make pretty heavy use of insecticides which aren't given to farm animals.
Not something people following a sustainable diet are going to be eating huge amounts of for calories.

Meat is properly replaced by plant-based protein, from grains and legumes.

We would assume a health-minded meat-eater is going to be eating fruits and veggies too; the difference is primarily in the protein choice.

I don't advocate vegans live on lettuce; that's not a very sustainable diet at all.
carnap wrote: Fri Mar 16, 2018 1:09 amFarm animals can also be pastured which would require little to no cultivated crops.
That may or may not be an improvement, but it is not an option at current consumption levels, and it also has more deleterious environmental consequences.
carnap wrote: Fri Mar 16, 2018 1:09 amSo there is no obvious answer here, you'd have to compare various methods of raising animals and define a calculus for comparing the sentience of one animal to another before you could make a meaningful distinction between the raising of crops vs animals.
If you were to regard insect lives as equally valuable to human lives, so you could functionally ignore the harms of global warming and other negative effects to human civilization, and compared the best case of meat like grass fed beef to the worst case in plants as some fruit or vegetable with high adult insect killing pesticide demands (or assume that an insect egg is as important as an adult human), then you might be able to obfuscate the difference.

I don't think that's going to seem reasonable to anybody, and I don't know why you'd want to make that argument...
carnap
Anti-Vegan Troll
Posts: 414
Joined: Wed Feb 01, 2017 12:54 pm

Re: On Fish Suffering

Post by carnap »

brimstoneSalad wrote: Fri Mar 16, 2018 4:22 am It is in discovering those negative or positive evaluations that we probe the interests of the robot (programmed though they may be).
I'm not sure what you're trying to say here, associative learning is just being able to associate two or more variables. For animals that typically means associating some stimuli to some response.

brimstoneSalad wrote: Fri Mar 16, 2018 4:22 am Sentience refers more specifically to sense experience and cognition of those experiences in some meaningful way (like associative learning). It is at least a bit better nailed down and minimalist.
Consciousness can refer to a few different notions so in that sense sentience is narrower, but its no more well understood.

brimstoneSalad wrote: Fri Mar 16, 2018 4:22 am The only good way to compare something that is fundamentally demonstrated by learning is through just that -- learning -- which is roughly comparable to intelligence.
Again, we are discussing sentience so I'm not sure why comparing learning (or intelligence) is relevant here. Is intelligence a proxy for sentience? Based on what exactly? Are intelligent computing systems sentient?
brimstoneSalad wrote: Fri Mar 16, 2018 4:22 am Not something people following a sustainable diet are going to be eating huge amounts of for calories.

Meat is properly replaced by plant-based protein, from grains and legumes.
Not sure what you mean by "plant-based protein", but if you have in mind processed foods constructed from protein isolates those have a pretty high impact and aren't all that sustainable. It takes a lot of energy to process and store them and a good deal is wasted in their production.

In any case, even in the US a good deal of animal fed is not grain that is edible to humans but instead various byproducts and plants we cannot consume. But you can raise animals on pasture and byproducts and in that case its not clear which would harm more insects.

If you were to regard insect lives as equally valuable to human lives, so you could functionally ignore the harms of global warming and other negative effects to human civilization, and compared the best case of meat like grass fed beef to the worst case in plants as some fruit or vegetable with high adult insect killing pesticide demands (or assume that an insect egg is as important as an adult human), then you might be able to obfuscate the difference.
brimstoneSalad wrote: Fri Mar 16, 2018 4:22 am I don't think that's going to seem reasonable to anybody, and I don't know why you'd want to make that argument...
I haven't made anything close to that argument......nowhere have I suggested insects lives are "equally valuable to human lives", etc. Honestly, not even sure where that is coming from. My point is that without a way of comparing relative sentience and degrees of suffering its not clear what subsistence methods would result in less harm to sentient animals. Including insects into the picture complicates matters considerably.
I'm here to exploit you schmucks into demonstrating the blatant anti-intellectualism in the vegan community and the reality of veganism. But I can do that with any user name.
User avatar
brimstoneSalad
neither stone nor salad
Posts: 10273
Joined: Wed May 28, 2014 9:20 am
Diet: Vegan

Re: On Fish Suffering

Post by brimstoneSalad »

carnap wrote: Fri Mar 23, 2018 4:51 am I'm not sure what you're trying to say here, associative learning is just being able to associate two or more variables. For animals that typically means associating some stimuli to some response.
Yes, but they do it for a reason: that they *want* the response. Without an interesting response, most animals won't learn effectively.

There is of course another kind of response to a specific instinct: curiosity. That can stimulate learning for its own sake, but seems to be limited to higher animals with the cognitive resources to spare. I don't think that has been demonstrated in insects.

There could also be some more or less reflexive processes, but those are poor explanations for behavioral interaction with learning.

My point is that certain kinds of associative learning like operant conditioning, which result in training certain behavior, demonstrate interests with respect to experiences and/or conditions of the outside world.

I'm on the fence about other kinds of associative learning (e.g. a cockroach may unconsciously associate some smell with food), but operant conditioning seems to be a smoking gun since it requires active participation in a way that wouldn't be found in nature so couldn't be hardwired.

Does operant conditioning happen in insects, or are they pure instinct? That's a good question, and there's limited evidence for it. But unfortunately existing evidence seems to suggest they do, in some limited capacity.
carnap wrote: Fri Mar 23, 2018 4:51 am Consciousness can refer to a few different notions so in that sense sentience is narrower, but its no more well understood.
That was my point regarding the definition.

As to understanding, it depends on what you mean: we don't need to know what's going on inside the black box to infer certain things from behavior beyond a reasonable doubt.

I think we do understand it in a pretty general sense, and that we have synthetic intelligence that probably fits the bill that we've created, but at the same time AI/SI etc. aren't really human readable and probably won't ever be due to the complexity of these emergent systems. I don't think that means we don't understand them, but you could argue that in a very narrow sense where understanding demands complete comprehension of the whole... but then by that definition we don't understand how computers work either... or pretty much anything.
carnap wrote: Fri Mar 23, 2018 4:51 am Again, we are discussing sentience so I'm not sure why comparing learning (or intelligence) is relevant here. Is intelligence a proxy for sentience? Based on what exactly? Are intelligent computing systems sentient?
I'm not sure how intelligence could occur without sentience. Any intelligent system must have sense input and a goal/interests to function.
Likewise, I don't think sentience can occur without intelligence.

Anyway, I can say that what I'm concerned with are interests and their absolute magnitude. I don't think there's any way to say that an insect is any more than barely interested in doing things given how it goes about doing them (assume it's a black box) vs. a human.
A human is better able to express operant conditioning to earn a penny than an insect is to save its own life. You can say that's just begging the question on intelligence, but if we assume these mental black boxes there's no other way to really assess relative interests.
carnap wrote: Fri Mar 23, 2018 4:51 amNot sure what you mean by "plant-based protein", but if you have in mind processed foods constructed from protein isolates those have a pretty high impact and aren't all that sustainable. It takes a lot of energy to process and store them and a good deal is wasted in their production.
My concern is promoting better options right now, less so the subtle differences beyond that.
carnap wrote: Fri Mar 23, 2018 4:51 amIn any case, even in the US a good deal of animal fed is not grain that is edible to humans but instead various byproducts and plants we cannot consume.
The overwhelming majority of food fed to animals is grown for them. However:
carnap wrote: Fri Mar 23, 2018 4:51 amBut you can raise animals on pasture and byproducts and in that case its not clear which would harm more insects.
I agree with this; it is possible to do. But in order to do it, people would have to drastically decrease meat consumption to a few percent of what it currently is (it would be interesting to do the math and figure out what percent).

Cows would have to be avoided due to enteric fermentation, but a purely waste stream based pork operation, for example, would probably be fairly sustainable and low impact.

Then the question becomes one of the relative sentience of the pig vs. the number of insects, and one of opportunity cost: it's always going to be better to grow something with a lower metabolism and less complexity like mushrooms on that feed stock.
carnap wrote: Fri Mar 23, 2018 4:51 amI haven't made anything close to that argument......nowhere have I suggested insects lives are "equally valuable to human lives", etc. Honestly, not even sure where that is coming from.
In threads I try to cover everything that a poster might be implying, or any inferences somebody might draw, since responses are slow. If that's not your position that's cool (I thought it would be weird if you were suggesting that, but I wanted to make sure to address it because many people read these threads)... just wanted to make sure I covered all of the bases.
carnap wrote: Fri Mar 23, 2018 4:51 amMy point is that without a way of comparing relative sentience and degrees of suffering its not clear what subsistence methods would result in less harm to sentient animals.
I think we can pretty effectively ball-park by looking at things like total neuron count. That will overestimate the contribution of insects since there's a lot of overhead and only a fraction is devoted to the cognitive function we're interested in, but it's a good test of whether it's even a plausible problem to consider.
carnap wrote: Fri Mar 23, 2018 4:51 amIncluding insects into the picture complicates matters considerably.
Maybe a little. I agree it would be easier if insects weren't sentient at all... but them being barely so doesn't shake things up much.
The only cases where it opens some questions are things we don't typically do and don't make up any meaningful part of our food system (like 100% waste stream fed pork).
carnap
Anti-Vegan Troll
Posts: 414
Joined: Wed Feb 01, 2017 12:54 pm

Re: On Fish Suffering

Post by carnap »

brimstoneSalad wrote: Fri Mar 23, 2018 3:31 pm Yes, but they do it for a reason: that they *want* the response. Without an interesting response, most animals won't learn effectively.
Based on what evidence? The capacity for desires is based on advanced cognition, for an entity to have desires it has to have some sense of itself, it has to be conscious and it has to have a conceptual model of its behavior. The ability for associative learning doesn't hinge on any of these capacities. The response doesn't need to be "interesting" or understood at all.

brimstoneSalad wrote: Fri Mar 23, 2018 3:31 pm My point is that certain kinds of associative learning like operant conditioning, which result in training certain behavior, demonstrate interests with respect to experiences and/or conditions of the outside world.
Yes that is your point and my point is that its easy to show that this is wrong with computer models of learning, for example, you can easily get a robot to learn via conditioning without that robot having any any active capacity for "interests". The algorithm just needs a criteria for associating variables which could be anything. For example:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4110879/

So if operant conditioning "demonstrates interests" than you're committed to the claim that robots also have interests in similar ways as animals.

brimstoneSalad wrote: Fri Mar 23, 2018 3:31 pm I'm not sure how intelligence could occur without sentience. Any intelligent system must have sense input and a goal/interests to function.
Likewise, I don't think sentience can occur without intelligence.
Why does sensory input and "interests" imply sentience? Your computer also has sensory input and also has "interests" in the very general sense of the word. Though it doesn't have "interests" in the more meaningful sense but nor do most animals it would seem. When people discuss "interests" they seem to equivocate a lot.

brimstoneSalad wrote: Fri Mar 23, 2018 3:31 pm The overwhelming majority of food fed to animals is grown for them. However:
In general this isn't true but it also depends on the animal. Cattle in the US, for example, are predominately pastured and even when they are finished in feedlots their diet is composed of large amounts of byproducts. For example soy meal (a byproduct of soy oil production).

At least in the US, I don't see any reason why a shift in production methods to more pastured based methods would result in a big downward shift in production. The US has plenty of land to pasture on and there is a large amount of byproducts produced eat year.
brimstoneSalad wrote: Fri Mar 23, 2018 3:31 pm I think we can pretty effectively ball-park by looking at things like total neuron count. That will overestimate the contribution of insects since there's a lot of overhead and only a fraction is devoted to the cognitive function we're interested in, but it's a good test of whether it's even a plausible problem to consider.
That wouldn't be very effective as neuron count hinges on the size of the animal and the overall efficiency of the brain which vary greatly from animal to animal. But it also doesn't explain how you'd do a comparison. So animal A has 50% the neurons of animal B.....what does that mean in terms of relative sentience and relative suffering?
I'm here to exploit you schmucks into demonstrating the blatant anti-intellectualism in the vegan community and the reality of veganism. But I can do that with any user name.
User avatar
brimstoneSalad
neither stone nor salad
Posts: 10273
Joined: Wed May 28, 2014 9:20 am
Diet: Vegan

Re: On Fish Suffering

Post by brimstoneSalad »

carnap wrote: Sun Mar 25, 2018 1:38 pm Based on what evidence?
Again, we don't need to know exactly what's going on in a black box to know things for a moral certainty. It's mechanistically obvious to that degree.
carnap wrote: Sun Mar 25, 2018 1:38 pmThe capacity for desires is based on advanced cognition,
A desire is something pretty primitive; it's necessary to move a neural network to learn. However, some things that appear to be simple associative learning could be hardwired (computers can "remember" user settings without having a neural network to do it, but this function is limited to what they've been specifically programmed to be able to remember). Like I said before, a cockroach associating a certain smell with food may not be a smoking gun in itself.
carnap wrote: Sun Mar 25, 2018 1:38 pmfor an entity to have desires it has to have some sense of itself, it has to be conscious and it has to have a conceptual model of its behavior.
What a wonderful coincidence! All things needed for operant conditioning, which is what I said is the smoking gun.
carnap wrote: Sun Mar 25, 2018 1:38 pmThe ability for associative learning doesn't hinge on any of these capacities.
Not the most basic appearances of it, but operant conditioning does.
You seem to be ignoring what I'm saying about this, or reading what you want into it.

Like I said, I'm on the fence on expression of learning that falls short of operant conditioning.
carnap wrote: Sun Mar 25, 2018 1:38 pmYes that is your point and my point is that its easy to show that this is wrong with computer models of learning, for example, you can easily get a robot to learn via conditioning without that robot having any any active capacity for "interests".
:lol: Again, begging the question! You have a bad habit of that.

You're assuming these robots are not sentient. What if sentience isn't "magical"? What if that's all sentience and subjective experience is?
carnap wrote: Sun Mar 25, 2018 1:38 pmThe algorithm just needs a criteria for associating variables which could be anything. For example:
Those variables are its interests and its sense experience. :roll:
If you think we're anything fundamentally different from that, then you'll need to put up some evidence.

Here, maybe this will help you out: https://www.gotquestions.org/human-soul.html
But if you're appealing to religious experience of your soul, it's not an argument that's going to pass here.
carnap wrote: Sun Mar 25, 2018 1:38 pm So if operant conditioning "demonstrates interests" than you're committed to the claim that robots also have interests in similar ways as animals.
Seriously? I just said that.
brimstoneSalad wrote:we have synthetic intelligence that probably fits the bill
I'm well aware of this. Currently, SI is around insect level and used in a very limited scope, so it's not something I'm worried about today, but I am concerned about the potential for abuse of Dorothy (as are all sensible secular individuals who don't write off robots for lacking souls) and other highly sentient robots.

carnap wrote: Sun Mar 25, 2018 1:38 pmWhy does sensory input and "interests" imply sentience?
Why does 1+1=2?
carnap wrote: Sun Mar 25, 2018 1:38 pmYour computer also has sensory input and also has "interests" in the very general sense of the word.
My computer is not running any artificial intelligence as far as I know. I don't know how you ascribe interests to it when it can't express any. You think it's interested in simply running?

It remembers very specific settings, but I don't think it's capable of basic operant conditioning.

carnap wrote: Sun Mar 25, 2018 1:38 pmWhen people discuss "interests" they seem to equivocate a lot.
They seem to employ magical thinking and wax theological.

carnap wrote: Sun Mar 25, 2018 1:38 pmIn general this isn't true but it also depends on the animal. Cattle in the US, for example, are predominately pastured
Pastured on managed forage which is grown for their consumption, including being fertilized much of the time (just a quick Google search shows how common this is http://www.beefmagazine.com/pasture-range/0331-fertilizing-pastures-spring ). And if it's not fertilized, they're just degrading the land (as occurs on some grasslands rented out for grazing).

People seem to think cows are these magical creatures that just capture an infinite and perfectly sustainable resources (grass) for our use.
That isn't the case. Forage is grown for cows much like any other crop, the only major difference is that the cows harvest it themselves (burning a lot of calories while they do it), and with cows (unlike with monogastric animals) there's substantial enteric fermentation which is a leading contributor to global warming. I assume you're just a climate change denialist because you consistently ignore that fact or seem indifferent to it.

There's no reason to support the beef industry or propagate fictions to defend it.
carnap wrote: Sun Mar 25, 2018 1:38 pmand even when they are finished in feedlots their diet is composed of large amounts of byproducts. For example soy meal (a byproduct of soy oil production).
:lol: Soy meal is a coproduct, not just a byproduct; the value on the market is similar to the oil. You should know that. Other oilseeds would be grown instead (with better oil yield) if not for that fact.
It's also something that should be fed to humans, as defatted soy flour in products, not be wasted on cows and contribute more to environmental damage in the process.

Cows also eat a lot of silage grown and fermented specifically for them.

There are very few things cows eat that humans fundamentally can not, and most of those are grown specifically for them in fields that could otherwise be growing food for human beings.
carnap wrote: Sun Mar 25, 2018 1:38 pm That wouldn't be very effective as neuron count hinges on the size of the animal and the overall efficiency of the brain which vary greatly from animal to animal. But it also doesn't explain how you'd do a comparison. So animal A has 50% the neurons of animal B.....what does that mean in terms of relative sentience and relative suffering?
I said it is probably an overestimation. The point is if it's even close, then that means there may be something to look into more.
I didn't say it's the final word on the issue (unless it's not close, then there's no point in examining it further at this point).
Post Reply