= 27d567bc2d48d9f40e9ccc20ea370b15)
wedrifid comments on A Challenge for LessWrong - Less Wrong
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 (158)
The examples listed are not rational. They are examples of 'altruism' for the sake of a 'warm feeling' and signalling. Writing a letter, ringing a politician or giving blood are not actions that maximise your altruistic preferences!
You have responded to this 'Potential Objection' with the "better than nothing" argument but even with that in mind this is not about being rational. It is just a bunch of do-gooders exhorting each other to be more sacrificial. When we used to do this at church we would say it was about God... and premising on some of the accepted beliefs that may have been rational. But it definitely isn't here.
I make a call for a different response. I encourage people to resist the influence, suppress the irrational urge take actions that are neither optimal signals nor an optimal instrument for satisfying their altruistic values.
This isn't a religious community and 'rational' is not or should not be just the local jargon for 'anything asserted to be morally good'.
If my preferences were such that I valued eating babies then it would be rational for me to eat babies. Rational is not nice, good, altruistic or self sacrificial. It just is.
Well, you're right that rationality is just a system for achieving a goal; it is the same process regardless of whether that goal is making the world a better place or turning it into a desert wasteland.
But, the OP is asking us to use rationality in a practical way and report back. That means we have to pick a goal, or there's nothing to point our rationality at. Making the world a better place for the people living in it (or to use a more utilitarian phrasing, reducing the net amount of potential and actual suffering in the world) seems like a pretty good one. It matches my own personal goals, at any rate.
Therefore: if you don't think the specific steps outlined in the OP are optimal for achieving that goal, please describe your alternative! I'm not being sarcastic; to use the chant, if the OP's steps are effective, I want to believe they're effective, and if they're not, I want to believe they're not.
But, please don't confuse that practical matter with the issue of choosing a goal; that argument is outside the bounds of rationality (except for the specific case of trying to justify one value as a sub-goal of another one).
Absolutely. What did Eliezer call this "must have a goal" principle way back when in the sequences? He explained it well, whatever it was.
What my point is is that none of the the actions listed are an effective way of achieving anything. Neither of the two purposes of altruistic actions are served (that being signalling and actually changing the world to match altruistic preferences.)
(For this response I'm going to focus on the goal of improving the world, not on signalling.)
One of the options was to give blood, which contributes directly to the reduction of suffering. I admit that I haven't personally looked into the effectiveness of the blood donation system, but as a basic medical technology it's quite sound, right? Why do you feel that donating blood is ineffective?
Two of the options were about donating to charities; one to a specific charity that seeks to defend a college student falsely accused of murder, and another a more general request to donate to any "reputable charity". I can understand that you might reasonably default to the null hypothesis on evaluating the effectiveness of any particular charity, particularly a minor one with little reputation like the Amanda Knox Defense charity... but it's a much stronger statement that reputable charities in general are not "an effective way of achieving anything"! Could you describe in more detail what leads you to that conclusion?
Finally, the remaining two options were about letter-writing or otherwise contacting people with political power in the hopes of influencing their actions. In terms of cost vs. benefit, this strikes me as being very hard to attack. Communication is cheap and easy, and public approval is a major factor in most political systems. By telling politicians explicitly what will earn your approval or disapproval, you're taking advantage of this system. I like the description here of this idea.
Do you disagree and feel that communicating with politicians is an ineffective way of influencing their decisions? If so, do you have a more effective alternative to propose?
Something to Protect?
That's the one, thanks guys. Time for me to do some revision.
Something to Protect?
That's okay, dammit!
Look, scoffing at less-than-optimal philanthropy is ultimately just another form of counterproductive negativity. If you're really serious about efficacy, you should be adding to the list of causes, not subtracting from it. That is, instead of responding to a post like this by encouraging people to
(!)
how about answering with "hey, you know what would be really, really helpful?" and proceeding to list some awesome utility-maximizing charity.
Warm feelings are good. Someone who donates a few spare frequent-flyer miles to help Curt Knox and Edda Mellas visit their daughter imprisoned 6,000 miles away doesn't need to feel ashamed of themselves for not being "rational" -- except in the extremely unlikely event that that action actually prevented them from doing something better. Does anyone honestly, seriously believe that discouraging people from doing things like this is a way of making the world a better place?
Speaking of challenges for LW, I propose a new rule: anybody who comes across an ostensibly good cause, but scoffs at its suboptimality, or thinks "well, it's not that I'm not willing to sacrifice $10, but surely there are better uses of that money" should be immediately required, right then and there, to donate that $10 to the Singularity Institute -- no ifs, ands, or buts.
No, no, NO! I desire to correct a fundamental mistake that is counter to whatever good 'rationality' may happen to provide. Raising the sanity waterline is an important goal in itself and particularly applicable in rare communities that have some hope of directing their actions in a way that is actually effective. Not only that, but seeing the very concept of rationality abused to manipulate people into bad decision making is something that makes me feel bad inside. Yes, it is the opposite of a warm fuzzy.
You are fundamentally wrong and the use of labeling things that disagree with you as 'negative' is non-rational influence technique that works in most places but should be discouraged here. It is not counterproductive to not do things that are stupid. It is not intrinsically better to add things to a list of normatively demanded behaviors while never removing them. If the list is wrong (for a given value of wrong) then it should be fixed by adding to it or removing from it in whatever way necessary.
People being manipulated into actions by the inclusion of irrelevant things in the definition of 'rational' is what I am discouraging. Tell people that Knox is a good way to purchase warm fuzzies, that's fine. But don't dare try to call it a 'challenge for rationality', piggybacking on the human instinct to avoid the shame of not supporting the tribal value ('rational').
No ifs and buts? Not everyone here needs to consider the SIAI to be the best use of their money. That's not required by 'rationality' either. You're in the wrong place if you think that approach is at all appropriate. Don't try to force your obsession with Knox on everyone else. It's not my priority and for most people it just isn't the rational way to maximise their preferences either.
While I agree with pretty much all of your points here, you may have better luck persuading those who do not if you take a less confrontational approach (I still fail at this occasionally, despite much effort). It's easier for me to accept a line of reasoning if that line of reasoning does not include the conclusion that I, personally, am evil. This would not be true if I were a perfect rationalist, but unfortunately it is not yet possible for me to escape my existence as a sack of neurons. And so it is with everyone.
The most persuasive arguments are the ones we want to believe. If you believe you are right (and you should), you should make it as easy as possible for people to agree with you.
Optimal persuasion was not my priority, emphasizing the nature of disagreement was. If deceptive use 'rational' was not a violation of both terminal and instrumental values then accusations of 'counterproductive negativity' or generally poor thinking, etc could be taken to have some approximation credibility - "it doesn't matter so you should shut up" vs "Direct attack on core values! Destroy!". I did originally explicitly explain things in these terms in the comment, including reference to ethical theory but that was turning into a seriously long winded tangent.
As for optimal persuasion, it would not involve a complete reply, a targeted paragraph responding to a cherry picked quote would be more effective. In fact, better still would be to make no reply at all, but create a new (much needed) post on the subject of ethics and rationality in order to direct all attention away from this one. Arguing with something gives it the credibility of 'something worth arguing with'.
I perceive the tone of the parent comment as needlessly inflammatory (constituting a violation of niceness) and will therefore take some time out before replying to the substance (no concession on which is to be inferred from my temporary silence).
ETA: The above was written before the sentence calling me "evil" was removed. I continue to take exception to the part about an alleged "obsession with Knox" that I am attempting to "force" on anyone. I defy anyone to justify such a characterization; my charitable interpretation is that wedrifid has misunderstood something I said, and/or forgotten that my comment and the original post were written by two different people.
It should be noted that I observe the tone of the parent of my rebuttal to be aggressive, with vigorous use of shaming to present a position that undermines a core value of this community. A vigorous response should be expected.
At WrongBot's suggestion I have removed the sentence containing the word 'evil'. Since almost nobody except myself uses that word in a technical sense it was foolish of me to include it here. I went through planning to edit out anything else that I wrote in haste that I would remove on reflection but I was surprised to find that was the only edit I needed to make. What remains has my reflective endorsement.
No, I cannot let you get away with that. The position I was presenting was that small good deeds should not be discouraged. If you are going to assert that that undermines a core value of this community (which one?), you are going to have to present a serious (and almost certainly novel) argument before you get to call me "evil".
Absolutely no "shaming" was used in presenting this position. The charge is an ironic one, because I am in fact attempting to defend myself and any other warm-fuzzy-enthusiasts who may happen to consider themselves members of this community from being "shamed" by those who would regard with contempt any activity not (e.g.) calculated to minimize the expected number of deaths.
Epistemic rationality (which, by the way, is what I presented the Knox case as a lesson in in the first place) is, as you know, not an end in itself. At least, it isn't the ultimate end. There has to be something to protect. And, at least in my own case, part of what I protect is that part of myself that is capable of caring about specific, individual humans, apart from "humanity" as an aggregate.
For the sake of cutting to the chase, let me now present what I think this disagreement is really about, and you can correct me if necessary. I think what is going on here is that you perceive the kind of "caring" I described above as an obstacle to epistemic rationality, which should therefore be Destroyed. Is that right, or am I being unfair?
See my ETA.
An inefficient small good deed is a negated greater good deed requiring the same effort. In this framing, the "small good deed" is actually a bad deed, and should be discouraged.
False. Time isn't fungible, and humans demonstrably don't make decisions that way.
Among other things, when humans are faced with too many alternatives, we usuallly choose "none of the above"... which means that the moment you complicate the question by even considering what those "greater good deeds might be", you dramatically reduce the probability that anything whatsoever will be accomplished.
False (at least I reject the incorrect generalization you use to contradict Vladmir). People who do small goods are less inclined to do subsequent goods. Given that the instincts evaluate 'good' more or less independently of any achievement fake 'good deeds' can prevent subsequent good deeds that make a difference. (This has been demonstrated.)
(Incidentally, Vladimir did not mention time at all.)
Oh really? What about the FITD effect?
He didn't have to. If time were unlimited, one could do any number of good deeds, and it would literally not matter how many of them you did, you could always do more... and thus there would be no competition between choices of how to use that time.
The assumption that not doing something now lets you do more later is false, however, because the time is already passing -- if you choose not to do something now, this doesn't give you any more time to do it later. Thus, a real thing done now beats an imaginary thing to be done later (which, given human psychology, probably won't actually be done).
That was apparently not the argument that wedrifid was making after all.
As for the argument itself: it says nothing more than that the good is bad because it isn't perfect. That is obviously wrong, because the good is better than nothing. It shouldn't be discouraged; rather, the better should be (separately) encouraged.
Again, see this post.
That is not right. I disagree specifically with the claims which I quoted in my reply and my disagreement is limited to precisely that which is contained in said reply.
I approve, for example, of seeking warm fuzzies and this is entirely in line with my stated position.
Then what, exactly, do we disagree about?
(Your earlier comment is of no help in clarifying this; in fact you explicitly described the pursuit of warm fuzzies -- as would be exemplified by contributing to the causes listed in the post -- as "bad decision making".)
This is not the case. I explicitly describe the equivocation of 'rational' with any meaning apart from 'rational' (and the application of said equivocation when decision making) as 'bad decision making'.
Okay, I think I see what happened. Your original point was really this:
-- with which I agree. However, the following statements distracted from that point and confused me:
These made it sound like you were saying "No! Don't contribute to those causes! Doing so would be irrational, since they're not philanthropically optimal!" (I unfortunately have a high prior on that type of argument being made here.) My natural response, which I automatically fired off when I saw that your comment had 17 upvotes, is that there's nothing irrational about liking to do small good deeds (warm fuzzies) separately from saving the planet.
However, as I understand you now, you don't necessarily see anything wrong with those causes; it's just that you disapprove of the label "rationality" being used to describe their goodness -- rather than, say, just plain "goodness".
Is this right?
What's a technical definition of "evil", then? I would say something about incompatible higher goals, but I'd find your take interesting.
That's a decent take. But how do we account for people that are not most effectively modeled as agents with goals? Deontologists for example, can be evil even if their (alleged) preferences entirely match mine.
Sure, but they beat the heck out of endless navel-gazing on an ethereal blog. Compared to reading your 3,000th LW comment, giving blood might be a strictly dominant strategy -- it beats "read another comment" in almost all of the possible worlds in which we might find ourselves.
How many years should we spend optimizing our decision trees before we begin to devote some fraction of our time and energy to action? Why?
If all someone does is post here, that does sound sad.
I doubt that's actually the case. People just aren't posting about everything else they do.
Maximize, no, but promote - yes. I concur with DSimon - if these are borderline useless, please suggest something better! (That was half the point of this post!)
Also, note that if I feel warm and fuzzy as a result of an action that promotes my goals, that is not a bad thing - on the contrary, you could make a pretty good argument that systematically ethical people are those who like doing ethical things.
I also (perhaps unfairly) assumed my audience would follow along easily enough in my slight equivocation between "ethical" and "rational."
Not unfair, just more wrong. This is human bias. We identify with the in group identity and associate all morality and even epistemic beliefs with it. It doesn't matter whether it is godly, spiritual, professional, scientific, spiritual, enlightened, democratic or economic. We'll take the concept and associate it with whatever we happen to think is good of or be approved of by our peers. People will call things 'godly' even when they violate explicit instructions in their 'Word of God'. Because 'godly' really means 'what the tribe morality says right now'. People make the same error in thought when they use 'rational' to mean 'be nice' or even 'believe what I say'. This is ironic enough to be amusing if not for the prevalence of the error.