= 27d567bc2d48d9f40e9ccc20ea370b15)
Viliam_Bur comments on Open Thread, November 23-30, 2013 - Less Wrong Discussion
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (295)
Imagine that you find $1000 on the street. How much would you feel tempted to take it?
Imagine that you meet a person who has $1000 in their pocket. Assuming that you feel absolutely no empathy, how much would you feel tempted to kill the person and take their money? Let's assume that you believe there is almost zero chance someone would connect you with the crime -- either because you are in a highly anonymous situation, or because you are simply too bad at estimating risk.
Not very tempted, actually. In this hypothetical, since I'm not feeling empathy the murder wouldn't make me feel bad and I get money. But who says I have to decide based on how stuff makes me feel?
I might feel absolutely nothing for this stranger and still think "Having the money would be nice, but I guess that would lower net utility. I'll forego the money because utilitarianism says so." That's pretty much exactly what I think when donating to the AMF, and I don't see why a psychopath couldn't have that same thought.
I guess the question I'm getting at is, can you care about someone else and their utility function without feeling empathy for them? I think you can, and saying you can't just boils down to saying that ethics are determined by emotions.
I think that ethics, as it actually happens in human brains, are determined by emotions. What causes you to be an utilitarian?
There's more to it than that. How about upbringing and conditioning? Sure, it made you feel emotions in the past, but it probably has a huge impact on your current behaviour although it might not make you feel emotions now.
Caring about someone else's utility function is practically the definition of empathy.
Nitpick: I've seen a distinction between affective empathy (automatically feeling what other people feel) and cognitive empathy (understanding what other people feel), where the former is what psychopaths are assumed to lack.
In practice, caring without affective empathy isn't intuitive and does take effort, but that's how I view the whole effective altruism/"separating warm fuzzies from utilons" notion. You don't get any warm empathic fuzzy feelings from helping people you can't see, but some of us do it anyway.
That's the way I care and try to care about people.
This is a valid point and it actually makes my statement stronger. Simply understanding what people like/dislike may not be considered 'true empathy', but caring about what they like/dislike certainly is.
If I make chicken soup for my friend when he's sick, and then I feel good because I can see I've made him happy, that's empathy. If I give $100 to a charity that helps someone I will never see, that's not empathy. The reward there isn't "I see someone happy and I feel their joy as my own." It's knowing abstractly that I've done the right thing. I've done both, and the emotional aspects have virtually nothing in common.
All forms of empathy must necessarily be indirect. When you see your friend happy, you don't directly percieve his happiness. Instead, you pick up on cues like facial expression and movements. You extract features that correspond to your mental model of human happiness. Let me make this clear and explain why it's relevant to the discussion.
Let's say your friend is asleep. You make him friend chicken soup, leave it on the table, and go to work. He later sends you a single text, "Thanks, the chicken soup made me really happy." This puts a smile on your face. I'm pretty sure you would consider that the first form of empathy, even though you never saw your friend happy. Indeed, the only indication of his happiness is several characters on a phone display.
Now let's take this further. Let's say every time you make your friend chicken soup it makes him happy, so that you can predict with confidence that making him chicken soup will always make him happy. Next time you make him chicken soup, do you even need to see him or get a text from him? No, you already know it's making him happy. Is this type of empathy the first kind or the second kind?
I'd call it the first kind, because it actually causes warm-fuzzy-happy feelings in me. My emotion reflects the emotion I reasonably believe my friend is feeling. Whereas the satisfaction in knowing I have done the right thing for someone far away whom I don't know and will never meet is qualitatively more like my satisfaction in knowing that my shoes are tied symmetrically, or that the document I have just written is free of misspellings. I've done The Right Thing, and that's good in an abstract aesthetic way, but none of my feelings reflect those I would believe, on reflection, that the recipient of the good deed would now be feeling. It doesn't put a smile on my face the way helping my friend does.
Well, what you say you feel is subjective (as is what I say I feel) but when I personally donate to charity it's because helping people - even if I don't directly see the results of my help - makes me happy. If not the 'warm fuzzy feeling', at least a feeling comparable to that of helping my friend. That is my subective feeling.
Nah, you can care about someones utility function instrumentally. In fact I think that's the way most people care about it most of the time, and have no reliable evidence to suggest otherwise.
I meant 'caring' as in direct influence of their utility on your utility (or, at least, the perception of their utility on your utility), conditionally independent of what their utility results in. If you take 'care' to simply mean 'caring about the outcomes' then yes you're right. Saying that all people are that way seems quite a strong statement, on par with declaring all humans to be psychopaths.
So you meant instrumentally in the first place. I misuderstood you, so retracted both the comment and the downvote.
Definitely not. Psychopaths are far more anomalous than selfish people. Also, I said most people most of the time, not all people all the time.
I suppose the word 'psychopath' is itself problematic and ill-defined, so fair enough.
They could. But if you select a random psychopath from the whole population, what is the probability of choosing an utilitarian?
To be afraid of non-empathic people, you don't have to believe that all of them, without an exception, would harm you for their trivial gain. Just that many of them would.
You would also have to know in what proportion they exist to know that, and you don't have that information precisely because of such presumptions. You wouldn't even know what's normal if displaying certain qualities is useful enough, and detecting whether people really have them isn't reliable enough.
It's possible to steelman that hypothetical to the threshold that yeah, killing someone for their money would be tempting. It wouldn't have much resemblance to real life after that however.
There are several other reasons not to kill someone for their money than empathy, so I'm not sure how your hypothetical illustrates anything relevant.