= 27d567bc2d48d9f40e9ccc20ea370b15)
ctl comments on Morality is Awesome - 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 (437)
This may be a minor nit, but... is this forum collectively anti-orgasmium, now?
Because being orgasmium is by definition more pleasant than not being orgasmium. Refusing to become orgasmium is a hedonistic utilitarian mistake, full stop.[1] (Well, that's not actually true, since as a human you can make other people happier, and as orgasmium you presumably cannot. But it is at least on average a mistake to refuse to become orgasmium; I would argue that it is virtually always a mistake.)
[1] We're all hedonistic utilitarians, right?
... no?
http://lesswrong.com/lw/lb/notforthesakeofhappinessalone/
Interesting stuff. Very interesting.
Do you buy it?
That article is arguing that it's all right to value things that aren't mental states over a net gain in mental utility.[1] If, for instance, you're given the choice between feeling like you've made lots of scientific discoveries and actually making just a few scientific discoveries, it's reasonable to prefer the latter.[2]
Well, that example doesn't sound all that ridiculous.
But the logic that Eliezer is using is exactly the same logic that drives somebody who's dying of a horrible disease to refuse antibiotics, because she wants to keep her body natural. And this choice is — well, it isn't wrong, choices can't be "wrong" — but it reflects a very fundamental sort of human bias. It's misguided.
And I think that Eliezer's argument is misguided, too. He can't stand the idea that scientific discovery is only an instrument to increase happiness, so he makes it a terminal value just because he can. This is less horrible than the hippie who thinks that maintaining her "naturalness" is more important than avoiding a painful death, but it's not much less dumb.
[1] Or a net gain in "happiness," if we don't mind using that word as a catchall for "whatever it is that makes good mental states good."
[2] In this discussion we are, of course, ignoring external effects altogether. And we're assuming that the person who gets to experience lots of scientific discoveries really is happier than the person who doesn't, otherwise there's nothing to debate. Let me note that in the real world, it is obviously possible to make yourself less happy by taking joy-inducing drugs — for instance if doing so devalues the rest of your life. This fact makes Eliezer's stance seem a lot more reasonable than it actually is.
You can argue that having values other than hedonistic utility is mistaken in certain cases. But that doesn't imply that it's mistaken in all cases.
Very well, let's back up Eliezer's argument with some hard evidence. Fortunately, lukeprog has already written a brief review of the neuroscience on this topic. The verdict? Eliezer is right. People value things other than happiness and pleasure. The idea that pleasant feelings are the sole good is an illusion created by the fact that the signals for wanting something and getting pleasure from it are comingled on the same neurons.
So no, Eliezer is not misguided. On the contrary, the evidence is on his side. People really do value more things than just happiness. If you want more evidence consider this thought experiment Alonso Fyfe cooked up:
Damn but that's a good example. Is it too long to submit to the Rationality Quotes thread?
Choices can be wrong, and that one is. The hippy is simply mistaken about the kinds of differences that exist between "natural" and "non-natural" things, and about how much she would care about those differences if she knew more chemistry and physics. And presumably if she was less mistaken in expectations of what happens "after you die".
As for relating this to Eliezer's argument, a few examples of wrong non-subjective-happiness values is no demonstration that subjective happiness is the only human terminal value. Especially given the introspective and experimental evidence that people care about certain things that aren't subjective happiness.
I see absolutely no reason that people shouldn't be allowed to decide this. (Where I firmly draw the line is people making decisions for other people on this kind of basis.)
I'm not arguing that people shouldn't decide that. I'm not arguing any kind of "should."
I'm just saying, if you do decide that, you're kind of dumb. And by analogy Eliezer was being kind of dumb in his article.
Okay. What do you mean by "dumb"?
In this case: letting bias and/or intellectual laziness dominate your decision-making process.
So if I wanted to respond to the person dying of a horrible disease who is refusing antibiotics, I might say something like "you are confused about what you actually value and about the meaning of the word 'natural.' If you understood more about about science and medicine and successfully resolved the relevant confusions, you would no longer want to make this decision." (I might also say something like "however, I respect your right to determine what kind of substances enter your body.")
I suppose you want me to say that Eliezer is also confused about what he actually values, namely that he thinks he values science but he only values the ability of science to increase human happiness. (I don't think he's confused about the meaning of any of the relevant words.)
I disagree. One reason to value science, even from a purely hedonistic point of view, is that science corrects itself over time, and in particular gives you better ideas about how to be a hedonist over time. If you wanted to actually design a process that turned people into orgasmium, you'd have to science a lot, and at the end of all that sciencing there's no guarantee that the process you've come up with is hedonistically optimal. Maybe you could increase the capacity of the orgasmium to experience happiness further if you'd scienced more. Once you turn everyone into orgasmium, nobody's around to science anymore, so nobody's around to find better processes for turning people into orgasmium (or, science forbid, find better ethical arguments against hedonistic utilitarianism).
In short, the capacity for self-improvement is lost, and that would be terrible regardless of what direction you're trying to improve towards.
I surmise from your comments that you may not be aware that Eliezer's written quite a bit on this matter; http://wiki.lesswrong.com/wiki/Complexityofvalue is a good summary/index ( is one of my favorites). There's a lot of stuff in there that is relevant to your points.
However, you asked me what I think, so here it is...
The wording of your first post in this thread seems telling. You say that "Refusing to become orgasmium is a hedonistic utilitarian mistake, full stop."
Do you want to become orgasmium?
Perhaps you do. In that case, I direct the question to myself, and my answer is no: I don't want to become orgasmium.
That having been established, what could it mean to say that my judgment is a "mistake"? That seems to be a category error. One can't be mistaken in wanting something. One can be mistaken about wanting something ("I thought I wanted X, but upon reflection and consideration of my mental state, it turns out I actually don't want X"), or one can be mistaken about some property of the thing in question, which affects the preference ("I thought I wanted X, but then I found out more about X, and now I don't want X"); but if you're aware of all relevant facts about the way the world is, and you're not mistaken about what your own mental states are, and you still want something... labeling that a "mistake" seems simply meaningless.
On to your analogy:
If someone wants to "keep her body natural", then conditional on that even being a coherent desire[1], what's wrong with it? If it harms other people somehow, then that's a problem... otherwise, I see no issue. I don't think it makes this person "kind of dumb" unless you mean that she's actually got other values that are being harmed by this value, or is being irrational in some other ways; but values in and of themselves cannot be irrational.
This construal is incorrect. Say rather: Eliezer does not agree that scientific discovery is only an instrument to increase happiness. Eliezer isn't making scientific discovery a terminal value, it is a terminal value for him. Terminal values are given.
Why are we doing that...? If it's only about happiness, then external effects should be irrelevant. You shouldn't need to ignore them; they shouldn't affect your point.
[1]Coherence matters: the difference between your hypothetical hippie and Eliezer the potential-scientific-discoverer is that the hippie, upon reflection, would realize (or so we would like to hope) that "natural" is not a very meaningful category, that her body is almost certainly already "not natural" in at least some important sense, and that "keeping her body natural" is just not a state of affairs that can be described in any consistent and intuitively correct way, much less one that can be implemented. That, if anything, is what makes her preference "dumb". There's no analogous failures of reasoning behind Eliezer's preference to actually discover things instead of just pretend-discovering, or my preference to not become orgasmium.
I have never used the word "mistake" by itself. I did say that refusing to become orgasmium is a hedonistic utilitarian mistake, which is mathematically true, unless you disagree with me on the definition of "hedonistic utilitarian mistake" (= an action which demonstrably results in less hedonic utility than some other action) or of "orgasmium" (= a state of maximum personal hedonic utility).[1]
I point this out because I think you are quite right: it doesn't make sense to tell somebody that they are mistaken in "wanting" something.
Indeed, I never argued that the dying hippie was mistaken. In fact I made exactly the same point that you're making, when I said:
What I said was that she is misguided.
The argument I was trying to make was, look, this hippie is using some suspect reasoning to make her decisions, and Eliezer's reasoning looks a lot like her's, so we should doubt Eliezer's conclusions. There are two perfectly reasonable ways to refute this argument: you can (1) deny that the hippie's reasoning is suspect, or (2) deny that Eliezer's reasoning is similar to hers.
These are both perfectly fine things to do, since I never elaborated on either point. (You seem to be trying option 1.) My comment can only possibly convince people who feel instinctively that both of these points are true.
All that said, I think that I am meaningfully right — in the sense that, if we debated this forever, we would both end up much closer to my (current) view than to your (current) view. Maybe I'll write an article about this stuff and see if I can make my case more strongly.
[1] Please note that I am ignoring the external effects of becoming orgasmium. If we take those into account, my statement stops being mathematically true.
I don't think those are the only two ways to refute the argument. I can think of at least two more:
(3) Deny the third step of the argument's structure — the "so we should doubt Eliezer's conclusions" part. Analogical reasoning applied to surface features of arguments is not reliable. There's really no substitute for actually examining an argument.
(4) Disagree that construing the hippie's position as constituting any sort of "reasoning" that may or may not be "suspect" is a meaningful description of what's going on in your hypothetical (or at least, the interesting aspect of what's going on, the part we're concerned with). The point I was making is this: what's relevant in that scenario is that the hippie has "keeping her body natural" as a terminal value. If that's a coherent value, then the rest of the reasoning ("and therefore I shouldn't take this pill") is trivial and of no interest to us. Now it may not be a coherent value, as I said; but if it is — well, arguing with terminal values is not a matter of poking holes in someone's logic. Terminal values are given.
As for your other points:
It's true, you didn't say "mistake" on its own. What I am wondering is this: ok, refusing to become orgasmium fails to satisfy the mathematical requirements of hedonistic utilitarianism.
But why should anyone care about that?
I don't mean this as a general, out-of-hand dismissal; I am asking, specifically, why such a requirement would override a person's desires:
Person A: If you become orgasmium, you would feel more pleasure than you otherwise would.
Person B: But I don't want to become orgasmium.
Person A: But if you want to feel as much pleasure as possible, then you should become orgasmium!
Person B: But... I don't want to become orgasmium.
I see Person B's position as being the final word on the matter (especially if, as you say, we're ignoring external consequences). Person A may be entirely right — but so what? Why should that affect Person B's judgments? Why should the mathematical requirements behind Person A's framework have any relevance to Person B's decisions? In other words, why should we be hedonistic utilitarians, if we don't want to be?
(If we imagine the above argument continuing, it would develop that Person B doesn't want to feel as much pleasure as possible; or, at the least, wants other things too, and even the pleasure thing he wants only given certain conditions; in other words, we'd arrive at conclusions along the lines outlined in the "Complexity of value" wiki entry.)
(As an aside, I'm still not sure why you're ignoring external effects in your arguments.)
If I become orgasmium, then I would cease to exist, and the orgasmium, which is not me in any meaningful sense, will have more pleasure than I otherwise would have. But I don't care about the pleasure of this orgasmium, and certainly would not pay my existence for it.
The difficulty here, of course, is that Person B is using a cached heuristic that outputs "no" for "become orgasmium"; and we cannot be certain that this heuristic is correct in this case. Just as Person A is using the (almost certainly flawed) heuristic "feel as much pleasure as possible", which outputs "yes" for "become orgasmium".
Why do you think so?
What do you mean by "correct"?
Edit: I think it would be useful for any participants in discussions like this to read Eliezer's Three Worlds Collide. Not as fictional evidence, but as an examination of the issues, which I think it does quite well. A relevant quote, from chapter 4, "Interlude with the Confessor":
Humans are not perfect reasoners.
[Edited for clarity.]
I give a decent probability to the optimal order of things containing absolutely zero pleasure. I assign a lower, but still significant, probability to it containing an infinite amount of pain in any given subjective interval.
No. Most of us are preferentists or similar. Some of us are not consequentialists at all.
For as long as I've been here, which admittedly isn't all that long.
Here's your problem.
No thanks. Awesomeness is more complex than can be achieved with wireheading.
I can't bring myself to see the creation of an awesomeness pill as the one problem of such huge complexity that even a superintelligent agent can't solve it.
I have no doubt that you could make a pill that would convince someone that they were living an awesome life, complete with hallucinations of rocket-powered tyrannosaurs, and black leather lab coats.
The trouble is that merely hallucinating those things, or merely feeling awesome is not enough.
The average optimizer probably has no code for experiencing utility, it only feels the utility of actions under consideration. The concept of valuing (or even having) internal experience is particular to humans, and is in fact only one of the many things that we care about. Is there a good argument for why internal experience ought to be the only thing we care about? Why should we forget all the other things that we like and focus solely on internal experience (and possibly altruism)?
Can't I simulate everything I care about? And if I can, why would I care about what is going on outside of the simulation, any more than I care now about a hypothetical asteroid on which the "true" purpose of the universe is written? Hell, if I can delete the fact from my memory that my utility function is being deceived, I'd gladly do so - yes, it will bring some momentous negative utility, but it would be a teensy bit greatly offset by the gains, especially stretched over a huge amount of time.
Now that I think about it...if, without an awesomeness pill, my decision would be to go and do battle in an eternal Valhalla where I polish my skills and have fun, and an awesomeness pill brings me that, except maybe better in some way I wouldn't normally have thought of...what is exactly the problem here? The image of a brain with the utility slider moved to the max is disturbing, but I myself can avoid caring about that particular asteroid. An image of an universe tiled with brains storing infinite integers is disturbing; one of an universe tiled with humans riding rocket-powered tyrannosaurs is great - and yet, they're one and the same; we just can't intuitively penetrate the black box that is the brain storing the integer. I'd gladly tile the universe with awesome.
If I could take an awesomeness pill and be whisked off somewhere where my body would be taken care of indefinitely, leaving everything else as it is, maybe I would decline; probably I won't. Luckily, once awesomeness pills become available, there probably won't be starving children, so that point seems moot.
[PS.] In any case, if my space fleet flies by some billboard saying that all this is an illusion, I'd probably smirk, I'd maybe blow it up with my rainbow lasers, and I'd definitely feel bad about all those other fellas whose space fleets are a bit less awesome and significantly more energy-consuming than mine (provided our AI is still limited by, at the very least, entropy; meaning limited in its ability to tile the world to infinity; if it can create the same amount of real giant robots as it can create awesome pills, it doesn't matter which option is taken), all just because they're bothered by silly billboards like this. If I'm allowed to have that knowledge and the resulting negative utility, that is.
[PPS.] I can't imagine how an awesomeness pill would max my sliders for self-improvement, accomplishment, etc without actually giving me the illusion of doing those things. As in, I can imagine feeling intense pleasure; I can't imagine feeling intense achievement separated from actually flying - or imagining that I'm flying - a spaceship - it wouldn't feel as fulfilling, and it makes no sense that an awesomeness pill would separate them if it's possible not to. It probably wouldn't have me go through the roundabout process of doing all the stuff, and it probably would max my sliders even if I can't imagine it, to an effect much different from the roundabout way, and by definition superior. As long as it doesn't modify my utility function (as long as I value flying space ships), I don't mind.
This is a key assumption. Sure, if I assume that the universe is such that no choice I make affects the chances that a child I care about will starve -- and, more generally, if I assume that no choice I make affects the chances that people will gain good stuff or bad stuff -- then sure, why not wirehead? It's not like there's anything useful I could be doing instead.
But some people would, in that scenario, object to the state of the world. Some people actually want to be able to affect the total amount of good and bad stuff that people get.
And, sure, the rest of us could get together and lie to them (e.g., by creating a simulation in which they believe that's the case), though it's not entirely clear why we ought to. We could also alter them (e.g., by removing their desire to actually do good) but it's not clear why we ought to do that, either.
Do you mean to distinguish this from believing that you have flown a spaceship?
Don't we have to do it (lying to people) because we value other people being happy? I'd rather trick them (or rather, let the AI do so without my knowledge) than have them spend a lot of time angsting about not being able to help anyone because everyone was already helped. (If there are people who can use your help, I'm not about to wirehead you though)
Yes. Thinking about simulating achievement got me confused about it. I can imagine intense pleasure or pain. I can't imagine intense achievement; if I just got the surge of warmth I normally get, it would feel wrong, removed from flying a spaceship. Yet, that doesnt mean that I don't have an achievement slider to max; it just means I can't imagine what maxing it indefinitely would feel like. Maxing the slider leading to hallucinations about performing activities related to achievement seems too roundabout - really, that's the only thing I can say; it feels like it won't work that way. Can the pill satisfy terminal values without making me think I satisfied them? I think this question shows that the sentence before it is just me being confused. Yet I can't imagine how an awesomeness pill would feel, hence I can't dispel this annoying confusion.
[EDIT] Maybe a pill that simply maxes the sliders would make me feel achievement, but without flying a spaceship, hence making it incomplete, hence forcing the AI to include a spaceship hallucinator. I think I am/was making it needlessly complicated. In any case, the general idea is that if we are all opposed to just feeling intense pleasure without all the other stuff we value, then a pill that gives us only intense pleasure is flawed and would not even be given as an option.
Regarding the first bit... well, we have a few basic choices: - Change the world so that reality makes them happy
- Change them so that reality makes them happy
- Lie to them about reality, so that they're happy
- Accept that they aren't happy
If I'm understanding your scenario properly, we don't want to do the first because it leaves more people worse off, and we don't want to do the last because it leaves us worse off. (Why our valuing other people being happy should be more important than their valuing actually helping people, I don't know, but I'll accept that it is.)
But why, on your view, ought we lie to them, rather than change them?
I attach negative utility to getting my utility function changed - I wouldn't change myself to maximize paperclips. I also attach negative utility to getting my memory modified - I don't like the normal decay that is happening even now, but far worse is getting a large swath of my memory wiped. I also dislike being fed negative information, but that is by far the least negative of the three, provided no negative consequences arise from the false belief. Hence, I'd prefer being fed negative information to having my memory modified to being made to stop caring about other people altogether. There is an especially big gap between the last one and the former two.
Thanks for summarizing my argument. I guess I need to work on expressing myself so I don't force other people to work through my roundaboutness :)
Fair enough. If you have any insight into why your preferences rank in this way, I'd be interested, but I accept that they are what they are.
However, I'm now confused about your claim.
Are you saying that we ought to treat other people in accordance with your preferences of how to be treated (e.g., lied to in the present rather than having their values changed or their memories altered)? Or are you just talking about how you'd like us to treat you? Or are you assuming that other people have the same preferences you do?
I don't understand this. If your utility function is being deceived, then you don't value the true state of affairs, right? Unless you value "my future self feeling utility" as a terminal value, and this outweighs the value of everything else ...
No, this is more about deleting a tiny discomfort - say, the fact that I know that all of it is an illusion; I attach a big value to my memory and especially disagree with sweeping changes to it, but I'll rely on the pill and thereby the AI to make the decision what shouldn't be deleted because doing so would interfere with the fulfillment of my terminal values and what can be deleted because it brings negative utility that isn't necessary.
Intellectually, I wouldn't care whether I'm the only drugged brain in a world where everyone is flying real spaceships. I probably can't fully deal with the intuition telling me I'm drugged though. It's not highly important - just a passing discomfort when I think about the particular topic (passing and tiny, unless there are starving children). Whether its worth keeping around so I can feel in control and totally not drugged and imprisoned...I guess that's reliant on the circumstances.
So you're saying that your utility function is fine with the world-as-it-is, but you don't like the sensation of knowing you're in a vat. Fair enough.
My first thought was that an awesomeness pill would be a pill that makes ordinary experience awesome. Things fall down. Reliably. That's awesome!
And in fact, that's a major element of popular science writing, though I don't know how well it works.
Psychedelic drugs already exist...
One time my roommate ate shrooms, and then he spent about 2 hours repeatedly knocking over an orange juice jug, and then picking it up again. It was bizarre. He said "this is the best thing ever" and was pretty sincere. It looked pretty silly from the outside though.
I'm anti-orgasmium, but not necessarily anti-experience-machine. I'm approximately a median-preference utilitarian. (This is more descriptive than normative)