{"items": [{"author": "Daniel", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553177748050559", "anchor": "fb-553177748050559", "service": "fb", "text": "What does it mean to attach confidence intervals to your moral positions? Do you just mean it's less than 1% likely you will at some later point decide that you aren't going to follow the \"Evaluate Actions\" principle any more? As a side note, my ethical system makes a sharp distinction between sins of commission and sins of omission. I also, somewhat regretfully, place negative value on some people (my enemies). I think both of these positions are nearly universally accepted by humans whether or not we admit such.", "timestamp": "1360812597"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553177944717206", "anchor": "fb-553177944717206", "service": "fb", "text": "\"People in the future are just as important as people now.\" Does this mean you don't apply any future discounting to utility? How do you avoid exploding/infinite results?", "timestamp": "1360812661"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553178454717155", "anchor": "fb-553178454717155", "service": "fb", "text": "Daniel, drawing a difference between sins of commission and omission is nowhere near universally accepted. I don't, Jeff doesn't, and the majority of people I have moral debates with (which is admittedly a biased sample) don't. Nor is placing negative value on some people universally accepted - consider e.g. Quakers.", "timestamp": "1360812790"}, {"author": "Daniel", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553179668050367", "anchor": "fb-553179668050367", "service": "fb", "text": "Ben, I retract \"universally accepted\" in favor of \"widely adhered to in practice\".", "timestamp": "1360813111"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553180288050305", "anchor": "fb-553180288050305", "service": "fb", "text": "@Daniel: \"What does it mean to attach confidence intervals to your moral positions?\"

Roughly, it's my estimate of the chance that they would stay unchanged even if I were to spend huge amounts of time thinking them over (including reading, discussing, etc).", "timestamp": "1360813280"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553181028050231", "anchor": "fb-553181028050231", "service": "fb", "text": "@Daniel: \"sharp distinction between sins of commission and sins of omission\"

I think this distinction goes away if you compare outcomes. Do you disagree that we should compare outcomes, or do you think this distinction is compatibile with looking only at outcomes.", "timestamp": "1360813451"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553181444716856", "anchor": "fb-553181444716856", "service": "fb", "text": "@Daniel: \"widely adhered to in practice\"

People do lots of immoral things in practice. We can still try and do better.", "timestamp": "1360813533"}, {"author": "Daniel", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553183171383350", "anchor": "fb-553183171383350", "service": "fb", "text": "Let me elaborate my position with a thought experiment. Say you are on vacation in Africa and you notice a small child carrying a gold ring. No one is looking, so you kill the child, pocket the ring, and sell it for $X at a pawn shop when you return home. Compare that to simply failing to donate $X to help cure malaria, where X is the cost to save a child's life. Everyone agrees the former sin (of commission) is crazily evil while the latter sin (of omission) is barely even an offense, even though they both result in the same net differences in outcome: one less child in the world, $X more in your pocket.", "timestamp": "1360813951"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553186951382972", "anchor": "fb-553186951382972", "service": "fb", "text": "Daniel, I have a couple clarifying questions about this experiment.
-Do you believe it's moral to treat people preferentially solely because they are nearby? (As opposed to treating them preferentially because e.g. it's easier to help nearby people, so you're doing more net good.)
-Do you have an answer for Jeff's question? I think this would help locate the fundamental disagreement.", "timestamp": "1360814768"}, {"author": "Christopher", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553227794712221", "anchor": "fb-553227794712221", "service": "fb", "text": "I think the thought experiment fails to account for collateral damage - quality of life of others in society. As a society, we tend to see living in peace and without fear as positive and desirable. So while the most moral thing is to help everyone, simple inaction (not donating money) is a lot less threatening than harmful action (killing someone). Put simply, I will live in greater fear of someone who might kill me for a gold ring, than someone who simply won't jump onto the tracks to save me from the train. So while net deaths may be the same, the results are very different.", "timestamp": "1360825072"}, {"author": "Aaron", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553249828043351", "anchor": "fb-553249828043351", "service": "fb", "text": "I'm not quite sure what's meant about the distinction between sins of omission and sins of commission going away when you evaluate consequences. If you could give me the value-relevant description of the whole history of the world, I would definitely agree that if two world histories are equally desirable, then it doesn't matter whether one is caused by omission or commission.

When I look at it without a whole-history view, it seems like the psychological consequences of a sin of commission and a sin of omission are very different in a way that actually influences your future actions in important ways.

Omission and Commission seem to have slightly different consequences in a way that's important but not overwhelming. In this thought experiment I think that, in terms of total outcomes, the two scenarios aren't actually equal because of different factors like your propensity to be kind to others in the future and the general state of unrest and discomfort in the area caused by a marginally higher murder rate, etc.

In theory, you could roll this distinction back into \"evaluate consequences\", but in practice it seems like people don't really do it that way, or adequately account for psychology-ish stuff in general.", "timestamp": "1360830614"}, {"author": "Keller", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553293728038961", "anchor": "fb-553293728038961", "service": "fb", "text": "You see a hundred dollar bill on a deserted street, pick it up, and notice a charitable collection box a few feet away. It is well sealed, so you have no reason to believe that the bill came from the box. You continue on your way.
Is that distinct from coming down that deserted street, seeing the box, and taking one hundred dollars out of it?
This doesn't address the psychological affects on you, but I believe that it covers almost everything else.", "timestamp": "1360841490"}, {"author": "Jeremy", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553299441371723", "anchor": "fb-553299441371723", "service": "fb", "text": "That's a lot of 90+'s! Are you just a very sure person, or is that a selection of the most important (and presumably most considered) principles? Or is there a sequel about the parts you don't trust so much?", "timestamp": "1360842541"}, {"author": "William", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553390731362594", "anchor": "fb-553390731362594", "service": "fb", "text": "I'm really pleased to see you do this - very few people do! But the numbers do seem high, especially in light of extensive disagreement among lay people and moral philosophers. (See the PhilPapers surveys). You might have a meta-ethical view on which disagreement isn't epistemically relevant - but how certain are you in that view?; with enough people to talk to, including those who defend alternative views, how likely would you be to change your beliefs? The chance of changing your meta-ethical view and thereby changing your first-order moral view needs to be factored into your credences as well.", "timestamp": "1360857818"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553395694695431", "anchor": "fb-553395694695431", "service": "fb", "text": "@Jeremy: \"is that a selection of the most important (and presumably most considered) principles?\"

Yes.

\"a sequel about the parts you don't trust so much?\"

There are definitely moral views I have that I'm much less confident in. For example the idea that you shouldn't consume substances to change the way you think. But there are so many of these, and they're so much in flux that I'm not sure I could make much of a post out of them.", "timestamp": "1360858674"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553396654695335", "anchor": "fb-553396654695335", "service": "fb", "text": "@Will: \"extensive disagreement among lay people and moral philosophers\"

Of the ones I have at 99% or more, my understanding is that only one (consequentialism) is controversial. Are the others also controversial? Do you have any anti-consequentialist writing you'd like to recommend?", "timestamp": "1360858877"}, {"author": "William", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553563514678649", "anchor": "fb-553563514678649", "service": "fb", "text": "I was thinking of \"living humans are what count\" and \"utilitarianism\" as well - 90%+ is still pretty high! As for recommended non-consequentialist reading... sorry, it's all really bad. Sam Scheffler \"The rejection of consequentialism\" and \"Utilitarianism, for and against\" by Smart and Williams are the standard readings. The Stanford Encyclopedia of Philosophy is normally pretty good - e.g. http://plato.stanford.edu/entries/ethics-deontological/ - but I haven't read that article. I'm also wondering: would you aggregate your uncertainty in the same way you'd aggregated empirical uncertainty? E.g. you might still give 10% chance to a pig's having positive utility, and so benefit a pig over a human when the pig would gain 10x as much?", "timestamp": "1360887834"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553598891341778", "anchor": "fb-553598891341778", "service": "fb", "text": "@Keller: \"You see a hundred dollar bill on a deserted street...\"

I started a long response and then turned it into a top level post: http://www.jefftk.com/news/2013-02-14", "timestamp": "1360894364"}, {"author": "Marcus", "source_link": "https://plus.google.com/115811589251174483775", "anchor": "gp-1360894366763", "service": "gp", "text": "I disagree with the solely analytical approach to morality and feel it is necessary to have some fix principles, particularly avoiding harm to others, except in exceptional cases. \u00a0Simply that it be beneficial to do so is insufficient. \u00a0Without such principles, even our rational minds can be clouded with unseen bias, leading us down an evil path. \u00a0Furthermore it is of vital importance that any decision that harm in necessary be limited to a single individual and as a direct consequence of their own actions and not their background.\n
\n
\nRationally looking at potential courses of action and choosing the one the seems most likely to cause the greatest good for the greatest number of people is an excellent way to cause great injustices. In fact, it accounts for most of the evil of the past century. \u00a0Once you disregard human dignity, particularly the worth of each individual person, it's easy to kill millions in an ideological crusade for a glorious future.\n
\n
\nSimilarly, harming a single individual, say through incarceration, can be justified as necessary to protect the community from their ill doings. \u00a0But when you increase the scale of such reasoning the consequences are often quite dire. \u00a0War, genocide and discrimination of kinds, come from applying the notion of acceptable harm without regard to scale.\n
\n
\nUltimately I feel it's hubris to think we can correctly estimate the consequences of our actions and quite dangerous to be certain that we are right in wronging others. \u00a0If the harm from inaction is immediate and obvious, there can be compulsion to intervene in some cases, but generally this should be done with great forbearance. We cannot \"look into the seeds of time and say which grain will grow and which will not\", but can merely hope to do \"the right as God gives us to see the right\". \u00a0Therefore the compulsion to avoid wrongdoing, even in the services of a greater cause, is strong.", "timestamp": 1360894366}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553599738008360", "anchor": "fb-553599738008360", "service": "fb", "text": "@Daniel: \"one less child in the world, $X more in your pocket\"

The social effects of people dying of malaria are pretty different from random foreigners killing your children. But I would say that you shouldn't kill the kid and that you should give $X to prevent malaria.", "timestamp": "1360894521"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=553600884674912", "anchor": "fb-553600884674912", "service": "fb", "text": "@Will: \"you might still give 10% chance to a pig's having positive utility, and so benefit a pig over a human when the pig would gain 10x as much?\"

If I changed my mind on animals not counting I still doubt I would count them as much as humans. But you could add that in as just another scaling factor.

\"would you aggregate your uncertainty in the same way you'd aggregated empirical uncertainty?\"

Being uncertain about this makes everything really tricky.", "timestamp": "1360894729"}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1360895620160", "service": "gp", "text": "@Marcus\n\u00a0\"Rationally looking at potential courses of action and choosing the one the seems most likely to cause the greatest good for the greatest number of people is an excellent way to cause great injustices.\"\n
\n
\nIf trying to make the world better makes it worse then you're going about it very wrong.\n
\n
\n\"Once you disregard human dignity, particularly the worth of each individual person, it's easy to kill millions in an ideological crusade for a glorious future.\"\n
\n
\nHuh? \u00a0How do you get that from \"choose actions that will make people better off in aggregate\"? \u00a0Killing one person is very bad; killing a million is about a million times worse, which makes it extremely horrible. \u00a0What would make you think that would be the best way to make people better off?\n
\n
\n\"I feel it's hubris to think we can correctly estimate the consequences of our actions and quite dangerous to be certain that we are right in wronging others.\"\n
\n
\nAgreed. \u00a0Predicting the consequences of actions is extremely difficult. \u00a0But this doesn't mean we should make no actions, it means we should be careful to learn as much as we can and take the best actions.", "timestamp": 1360895620}, {"author": "Marcus", "source_link": "https://plus.google.com/115811589251174483775", "anchor": "gp-1360895884823", "service": "gp", "text": "Yes, but you need some basis for thinking that the death of one person is bad. Name a butcher of the 20th century who \ndidn't\n think he was making the world a better place.", "timestamp": 1360895884}, {"author": "Marcus", "source_link": "https://plus.google.com/115811589251174483775", "anchor": "gp-1360896418743", "service": "gp", "text": "I agree that\u00a0difficulty\u00a0of moral analysis shouldn't lead to complete inaction, just that there need to be additional considerations beyond expediency, both because of\u00a0uncertainty, but also because of people's biases can lead them astray.", "timestamp": 1360896418}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1360898159109", "service": "gp", "text": "@Marcus\n\u00a0\"you need some basis for thinking that the death of one person is bad\"\n
\n
\nBut I do think the death of one person is bad! \u00a0This leads me to work hard to to prevent such deaths.\n
\n
\n\"Name a butcher of the 20th century who didn't think he was making the world a better place.\"\n
\n
\nThe butchers of the 20th century were wrong, and not repeating their mistakes is very important. \u00a0Any plan that involves wars or otherwise killing people is probably a very bad one. \u00a0Any plan that involves giant social changes, especially ones that haven't been extensively tested at smaller scales, is probably also harmful if implemented.", "timestamp": 1360898159}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1360898545106", "service": "gp", "text": "@Marcus\n\u00a0\"there need to be additional considerations beyond expediency, both because of uncertainty, but also because of people's biases can lead them astray\"\n
\n
\nAgain, I agree. \u00a0We need to learn to understand our biases and account for them so that what we think will make the world better actually does.", "timestamp": 1360898545}, {"author": "Marcus", "source_link": "https://plus.google.com/115811589251174483775", "anchor": "gp-1360940796184", "service": "gp", "text": "It seems that in practical concerns we basically agree, but for me it's more important to have certain values explicitly stated.", "timestamp": 1360940796}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1360945822100", "service": "gp", "text": "@Marcus\n\u00a0\"for me it's more important to have certain values explicitly stated\"\n
\n
\nDo you disagree with any of the values I listed in my post?", "timestamp": 1360945822}, {"author": "Marcus", "source_link": "https://plus.google.com/115811589251174483775", "anchor": "gp-1360946719709", "service": "gp", "text": "Yes, scale, for the reasons I mentioned.\u00a0 I think harming one person to help a number of people is more acceptable than harming a larger number of people to help an equivalently large ratio of people.\u00a0 Basically, I don't trust people to get \"beneficial harm\" decisions right on a large scale and think that allowing such thinking leads to scapegoating a group of people. (eg. \"our nation will be better off without <minority group>\")\n
\n
\nAlso, doubts about utilitarianism, not because I completely reject the premise of it, but because I think that it's difficult to get the value function correct.\u00a0 Most utilitarian writings I've read seem to reconstruct values derived from principle, but I think that's because they are culturally embedded in the author's minds already. I feel that even people who cause more harm than good to their community or to the world have inherent rights that cannot be completely abrogated. While it's possible to define a utilitarian function in such a way to account for those rights, I think it is difficult and risky to do so without explicitly stating reasons why those should exist.\n
\n
\nLastly, I have some doubts about being able to accurately use estimation, though in discussion it sounds like we're actually in roughly the same place with that.", "timestamp": 1360946719}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1360949672788", "service": "gp", "text": "@Marcus\n\u00a0\"I don't trust people to get 'beneficial harm' decisions right on a large scale\"\n
\n
\nThere are cases where we do pretty well with this, mostly in the field of public health. \u00a0I do agree we need to be careful with it, though, but I guess I put that as an issue in the application of the principle, not in the acceptance of it.", "timestamp": 1360949672}, {"author": "Marcus", "source_link": "https://plus.google.com/115811589251174483775", "anchor": "gp-1360951226433", "service": "gp", "text": "It may just be an issue of getting the value function right and if you put a large enough penalty on death, imprisonment, denial of civil rights, etc, there never is a case that, even with flawed assumptions we will inevitably make, we will want to conclude that it makes sense to oppress a group.\n
\n
\nEven in health there are some controversies though, one that immediately came to mind is the rule that men cannot donate blood if they've have sex with another man since 1977.", "timestamp": 1360951226}, {"author": "Keller", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=554304957937838", "anchor": "fb-554304957937838", "service": "fb", "text": "@Jeff : I understand your response. However, by addressing the question of the effectiveness of a given charity, which is a valid point, you have not answered the question about do x vs !do !x. What if the charity you saw on the street was the Against Malaria Foundation, Givewell's current top charity?", "timestamp": "1361018495"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=554358434599157", "anchor": "fb-554358434599157", "service": "fb", "text": "@Keller: \"What if the charity you saw on the street was the Against Malaria Foundation\"

Then I think it would be right to put the $100 I found in the box in (a) and not take money out of the box in (b).", "timestamp": "1361027539"}, {"author": "Andrew", "source_link": "https://www.facebook.com/jefftk/posts/553161234718877?comment_id=554955791206088", "anchor": "fb-554955791206088", "service": "fb", "text": "I'd like to know why you give only 10% for animals mattering. There's a lot of room between neuron farms and humans. Mammals experience all basic emotions that humans do. Humans can savour and understand things much better but humans can also inhibit our emotions better (I think). I'd be interested to know whether you started with a higher than 90% estimate and have been updating downwards, started lower and have been updating upwards, or neither.", "timestamp": "1361130093"}]}