Mapping Fun Theory onto the challenges of ethical foie gras
Foie gras, the delicacy made from the liver of a very fat goose (or sometimes duck), is believed to be unethical and is therefore frequently banned. For a long time, it was believed that the only way to properly fatten a goose is to continually force-feed it through a tube over several weeks, which is probably a highly unpleasant experience, although it’s difficult to tell. Recently, Spanish farmer Eduardo Sousa revealed that under highly specific conditions, you can get geese to fatten themselves voluntarily.
Geese will instinctively gorge themselves when winter is coming on. Eat a goose right after it’s fattened itself up for the winter, and you get a delicious treat that died happy. The problem is that geese will only do this if they believe food may become scarce during the winter (or their instinct to gorge only kicks in when the environment is such that that would be a reasonable inference; it’s not clear whether it’s the goose or evolution doing the analysis). If they realize that food will remain available during the winter, they eat normally. And there are quite a few possible clues—farmers trying to replicate Sousa’s setup have discovered that cheating on any part leads to unfatted livers.
Even as chicks, geese cannot be handled by a human, or encounter other geese who have been.
There can be no visible fences.
Geese cannot be “fed,” rather a variety of food must be distributed randomly throughout a large space, with the placement constantly changing, so that the geese happen to come across it.
We’re late.
So… suppose hypothetically that it were analogously true of humans that our likelihood of voluntarily maximizing “Fun” is dependent on being in an environment in which our access to Fun appears primarily determined by chance and our own efforts, and in which we believe Fun may soon run out.
It seems to follow from that supposition that if an outside force (e.g., a superhuman FAI) wants to maximize the amount of Fun we have, while still respecting our agency, it has to create such an environment.
How does adding that hypothetical constraint affect the conclusions of the Fun Theory Sequence?
Or does it?
How does such an environment differ from the world we actually live in?
Or does it?
Not rhetorical questions.
The first requirement:
suggests that a FAI would not tell us that it exists. In other words, the singularity may already have happened.
Agreed with the first part of that.
The second part I’m less sure of. It may already have happened iff the world as it is is consistent with what we’d expect such an entity to create, which is precisely what I’m wondering about.
My instinct is to say “nonsense; it could certainly create a world where irrevocable suffering and death aren’t quite so common, without violating that constraint”… a point that the Fun Theory Sequence also makes at some length, IIRC. But on thinking about it I’m not at all sure that it could without spoiling the scenario altogether.
For example: even in the world as it is, we are remarkably willing as a species to behave as though we were being taken care of by an all-powerful supernatural force; it might be that any significant further reduction in our suffering that wasn’t visibly our own doing would bring us over a tipping point where that willingness became a literally irresistable temptation.
Dunno. It seems as though it ought to be possible to make more specific predictions based on that speculation, perhaps even falsifiable ones, but thus far I’ve failed to get traction.
unless suffering entities do not have an internal experience and are only simulated to satisfy our perverse utility functions. It seems possible that the maximum fun a human type entity can have is being born in a time of scarcity and experiencing the transition to non-scarcity, in which case we should expect lots of such entities to be living in a simulation similar to our current world.
Major objection: historically, humans have behaved like that regardless of the absolute amount of suffering they were witnessing. We only evaluate the goodness of our world on a relative scale, according to the worst and the best conditions we’ve seen.
I’m sympathetic to this line of reasoning, but not entirely convinced of it. People raised in more abusive environments do seem to come out different from people raised in less abusive environments, for at least some spreads, and it’s not clear to me that their result is improved by being kept unaware of people who are better off. Though it might be.
Did a medieval townsman behave as though he/she was protected from the bad things that could happen any less than a modern middle-class person (whether religious or not)? I don’t think so.
Difficult question. From what I’ve read I’d expect medieval townsmen to be more aware of particular dangers than modern middle-class people: there’s a huge volume of medieval charms and prayers against thieves that’ve come down to us, for example, and I’m fairly certain that reflects an actual preoccupation. But it’s not clear to me that this would be any stronger in absolute terms than, say, modern culture’s fear of pedophiles.
In any case the baseline seems to have been within the same order of magnitude, while the everyday threat of violence or theft would have been multiple orders of magnitude higher in medieval society. On the other hand, the medieval threat landscape may also have been less stable, which I’d expect to represent a source of stress in its own right.
Yeah, that’s exactly what struck me as unbelievable about Dave’s assertion.
Just to make sure I understand… the “assertion” we’re talking about is the possibility that “any significant further reduction in our suffering that wasn’t visibly our own doing would bring us over a tipping point where that willingness [to behave as though we were being taken care of by an all-powerful supernatural force] became a literally irresistable temptation”… yes?
And what you find unbelievable about this is that when you compare modern willingness to behave that way to medieval willingness to behave that way, you find that they’re roughly the same, but that when you compare modern threat level to medieval threat level, you find the medieval threat level is significantly higher. Yes?
If I got both of those right, I am confused. Can you unpack the relationship between those two assertions more precisely?
(Another objection: it sounds more and more like the Christian take on the problem of evil, which is to say: fuck the deity that thinks it’s best for us, we should take care of ourselves instead. Even if we should grow lazier and more complacent, real suffering has a higher moral cost than whatever qualities our life might lose.
Didn’t most people here agree that the society of Brave New World, whatever its flaws and vices, is better than the current one?)
No, they don’t. Put “Brave New World” into the LessWrong Google box and see for yourself.
But I think you’re just asking a rhetorical question as a way of saying that you do think that, without actually saying so.
Um, I’m surprised. I definitely saw something to the effect of that.
I’m pretty sure I’m not altogether following your thought process here, but yes, insofar as “the Christian take on the problem of evil” is to posit that human nature is such that current evil is a necessary condition of ultimately improving the human condition, then this hypothetical sounds somewhat like it. (Of course, the Christian version also has to account for why an all-powerful creator God caused human nature to have that property, which this hypothetical does not.)
As for relative moral costs… (shrug) as I’ve said repeatedly in this thread, I don’t know. More precisely… I agree that in the world where we choose between (lazy, complacent, and non-suffering) and (non-lazy, non-complacent, and suffering) it might well be preferable to eliminate suffering at the cost of laziness and complacency. But in the world where we choose between (lazy, complacent, and non-suffering for eternity) and (non-lazy, non-complacent, and suffering for a while, followed by much higher levels of anti-suffering for eternity), it’s not clear the same conclusion arises.
More links on this wild-fattening and the research?
The radio series This American Life devoted part of their most recent episode to this. I was sure that there would be a linked to it in the OP, but I don’t see it.
I went to the transcript and couldn’t help but be struck by this quote:
It’s quite clear that the goose is doing the analysis, in that whatever analysis is happening is happening on goose hardware. All that is unclear is just how much of it is cognitive, and just how much any cognitive processes there interact with the nearest analogues of consciousness within goose heads.
How about this solution:
Stop expending resources to bend people’s preferences in the direction of liking foie gras.
And the solution generalizes to lots of stuff beyond foie gras: critic-approved art, for example.
My recollection is that this guy was winning competitions until they kicked him out on the grounds that without force feeding, it’s not really foie gras.
That’s really interesting if true. Can you find where you read that?
I cannot verify right now, but I believe it was in this TED talk.
Yes, he mentions it there, in passing.
Thanks!
“He doesn’t look like a guy who’s paying off French judges for his foie gras. So that died down, and very soon afterward, new controversy. He shouldn’t win because it’s not foie gras. It’s not foie gras because it’s not gavage. There’s no force feeding. So by definition, he’s lying and should be disqualified.”
He doesn’t actually say that Sousa was disqualified, but that some people thought he should be.
Alternately we can work on completely cruelty free synthetic replacements for foie gras, instead of spending time pandering to geese.
Actually, wouldn’t that be … *puts on sunglasses* … gandering?
So far as I know it’s not an acquired taste (e.g., generally unpleasant), so people would probably want it even if nobody were “bending their preferences”.
They have to do some bending to get people to notice its existence in the first place, let alone deem it worthy of trying a first time. The fact that people insist on calling it by its French name, rather than “fat liver”, is a testament to the marketing that has to be done to support interest in it.
I don’t pretend this is a cure-all, or that we can always end livestock torture by not promoting its tasty products—that would be endorsing a “just world fallacy”. But sometimes we really do make our hard choices a lot harder than they need to be.
The study of variable quantities is called “al-jabr” not because mathematicians want to make it sound exotic, but because of historical accident. Unless you have particularly good reason to think otherwise, I would guess foie gras is the same way.
People are automatically repulsed by “fat liver”. They’re not repulsed by “the restoration”. Foie gras needs to hide its original-language meaning to avoid turning away some people; algebra doesn’t. Not a particularly relevant comparison, I think.
People who speak English are (possibly) automatically repulsed by “fat liver”. French speakers are not similarly repulsed by “fois gras”. The difference has little if anything to do with the practice of force feeding.
Because that culture spent correspondingly more effort bending people’s preferences in favor of eating that food.
Americans are repulsed by just “liver”, for the most part. It’s unfortunate, organ meat is really good for you, and for the most part much cheaper than muscle.
Would the latter remain true in America if Americans were to lose their beliefs that organ meat is repulsive?
Probably the price of organ meat would go up, while the price of “normal” meat would go down. That’s basically a winner for everyone.
Except those of us who already like organ meat...
Well, you people are disgusting anyway :)
Though the related expression* might lead one to believe otherwise, Jewish Americans consume chopped liver that is labelled “chopped liver” pretty regularly. Though I’ve got to say that chopped beef liver is much better than chopped chicken liver, so I have my doubts about goose liver pate (which I’ve never tried). Also, a little minced onion and salt makes it a lot more appetizing than it might be alone.
On a different note, make sure you don’t eat liver from a carnivore.
*”What am I, chopped liver?”
Also, some hard boiled eggs.
Mmm… chopped liver...
I think most of the discussion of “bending preferences” and “acquired taste” underestimate the variance of tastes across the population. I’ve seen someone on LW saying that they didn’t enjoy wine, and therefore suspecting that whoever claims to enjoy wine must be doing it for signalling. The idea that maybe some people actually enjoy wine and some don’t doesn’t seem to have occurred to them (where by some I mean ‘a sizeable fraction of the population’). Likewise, I’d be shocked to find that the fraction of people who actually like foie grass the first time they try it is 90%.
Has there been a discussion here of the ethics of killing and eating an animal versus forcing it to live uncomfortably?
No, but I eat meat and don’t eat factory farmed eggs. I think I’m basically the only person that does that, though.
I once read an article arguing that separating animals from their young was worse than killing them.
I am not sure that you can count a distressed fowl running around like mad worrying about the coming winter scarceness a happy goose. Or at least any more happy than a force-fed one.
Well, there seems to be a real correlation with happy meat and taste quality, so that this farmer was decisively winning international foie gras taste competitions is probably real evidence that his geese were happier. Personally, I’d much rather be running around like mad worrying about the coming winter than being force fed. And in geese, the worry-based gorging is much more natural than the alternative of force feeding.
Visually pleasing food tastes better than visually unpleasing food; fragrant food tastes better than rank food; if ethical meant tastes better than unethical meat, it has nothing to do with the meat, and everything to do with the interconnected nature of the taster’s sensory network.
I have heard that too, but I fear it’s just halo effect i.e. if you are kind to animals they will automatically taste better.
Shouldn’t be too hard to test if the effect size is large enough. Next time you find yourself hosting a dinner for n meat-eating people, have an accomplice buy n/2 factory-farmed portions of meat and another n/2 ethically farmed portions and label them 1 through n at random, secretly keeping track of which is which. Then cook them all yourself, serve, and give out a survey about the quality of the food. Compare notes afterwards. If your guests protest, tell them they’re eating science. Better yet, conspire with a restaurant owner if you happen to know one.
Actually, someone’s probably already done this—although most of the people with an incentive to do so in an unbiased way would also have an incentive to keep the results secret.
Yes that would be interesting. What would a representative number of dinner guests be?
But it might be that people actually find ethical meat generally tastier when presented as ethical meat i.e. placebo. (I do understand that the study you suggest would be blinded).
My main message though was to draw attention to the halo effect in diet and medicine, I have encountered people for example that take for granted that antibiotic-resistant bacteria are generally stronger than it’s non-antibiotic-resistant counterpart.
This would clearly come down in favor of ethical meat tasting better, but that’s because ethical meat is given better food. You’d have to give the cage confined higher quality feed for the comparison to be proper.
I think that n would need to be a lot larger than the number of people which fit in your house or a restaurant for the results not to be swamped by random noise.
ETA: A way to get more useful data from the same number of people would be asking each person to taste both kinds of meat (without telling them which is which, of course), and asking them which one tastes better.
Can anyone figure out if it’s possible to buy this foie gras in the USA?
Did anyone look at the behavior (feeding and avoiding force-feeding places) differences between geese and ducks?
… I find myself once again perplexed by LW’s obsession with hedonistic utilitarianism as opposed to all other forms of utilitarianism which exist, such as what I call “Functional Utilitarianism”—which, in very simple terms, is the notion that utility is best measured by the impact of a given action on the abilities of others and one’s own self to fulfill or augment the principle of self-definition.
Have you made a post about this? If not, do so.
those terms aren’t simple at all
Who cares what birds think? They can’t understand the prisoner’s dilemma.
Depending on how sentient they’re judged to be, utilitarians.
Yes, I’m aware that there’s an alternate interpretation of the above sentence. :D
“Understanding the PD” per se is an arbitrary requirement. Recognition of an animal’s ethical worth should depend on a less parochial standard, like whether they act on gratefulness / reciprocity / Golden Rule / subjunctive consequences. I would consider it close enough that e.g. the animal does good things for you as a result of you having done good things for it.
I don’t know about birds, but domestic cats and dogs definitely meet that standard.
I’m not so much concerned about an animal’s “ethical worth”, I just don’t want anybody force-feeding me in order to eat my liver.
CooperateBots practice cargo cult decision theory. (I probably miss your point, in which case please clarify.)