= 27d567bc2d48d9f40e9ccc20ea370b15)
DanArmak comments on Rationality Quotes May 2012 - 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 (696)
Don't remove the value. Remove just the experience of feeling bad due to not yet achieving the value.
If I have a value/goal of being rich, this doesn't have to mean I will feel miserable until I'm rich.
What you're implicitly doing here is divorcing goals from values (feelings are a value). Either that or you're thinking that there's something especially wrong related to negative incentives that doesn't apply to positive ones.
If you don't feel miserable when you're poor or, similarly, if you won't feel happier when you're rich, then why would you value being rich at all? If your emotions don't change in response to having or not having a certain something then that something doesn't count as a goal. You would be wanting something without caring about it, which is silly. You're saying we should remove the reasons we care about X while still pursuing X, which makes no sense.
There's something terribly wrong about the way negative incentives are implemented in humans. I think the experience of pain (and the fear or anticipation of it) is a terrible thing and I wish I could self-modify so I would feel pain as damage/danger signals, but without the affect of pain. (There are people wired like this, but I can't find the name for the condition right now.)
Similarly, I would like to get rid of the negative affect of (almost?) everything else in life. Fear, grief, etc. They're the way evolution implemented negative reinforcement learning in us, but they're not the only possible way, and they're no longer needed for survival; if we only had the tools to replace them with something else.
Being rich is (as an example) an instrumental goal, not a terminal one. I want it because I will use the money to buy things and experiences that will make me feel good, much more than having the money (and not using it) would.
"pain asymbolia"
Treating it as an instrumental goal doesn't solve the problem, it just moves it back a step. Even if you wouldn't feel miserable by being poor because you magically eliminated negative incentives you would still feel less of the positive incentives when you are poor than when you were rich, even though richness is just the means to feeling better. All of this:
still applies.
(Except insofar as it might be altered by relevant differences between positive and negative incentives.)
To clarify, what I'm contending is that this would only make sense as a motivational system if you placed positive value on achieving certain goals which you hadn't yet achieved, I think you agree with this part but am not sure. But I don't think we can justify treating positive incentives differently than negative ones.
I don't view the distinction between an absence of a positive incentive and the presence of a negative incentive the same way you do. I'm not even sure that I have any positive incentives which aren't derived from negative incentives.
Negative and positive feelings are differently wired in the brain. Fewer positive feelings is not the same as more negative ones. Getting rid of negative feelings is very worthwhile even without increasing positive ones.
But the same logic justifies both, even if they are drastically different in other sort of ways.
Forcing yourself to feel maximum happiness would make sense if forcing yourself to feel minimum unhappiness made sense. They both interact with utilitarianism and preference systems which are the only relevant parts of the logic. The degree or direction of the experience doesn't matter here.
Removing negative incentives justifies maxing out positive incentives = nihilism.
I mean, you can arbitrarily only apply it to certain incentives which is desirable because that precludes the nihilism. But that feels too ad hoc and it still would mean that you can't remove the reasons you care about something while continuing to think of it as a goal, which is part of what I was trying to get at.
So, given that I don't like nihilism or preference paralysis but I do support changing values sometimes, I guess that my overall advocacy is that values should only be modified to max out happiness / minimize unhappiness if happiness / no unhappiness is unachievable (or perhaps also if modifying those specific values helps you to achieve more value total through other routes). Maybe that's the path to an agreement between us.
If you have an insatiable positive preference, satiate it by modifying yourself to be content with what you have. If you can never be rid of a certain negative incentive, try to change your preferences so that you like it. Unfortunately, this does entail losing your initial goals. But it's not a very big loss to lose unachievable goals while still achieving the reasons the goals matter, so fulfilling your values by modifying them definitely makes sense.
Reducing bad experience was the original subject of discussion. As I said, it's worthwhile to reduce them even without increasing good experience. I never said I don't want to increase good experience - I do! As you say, both are justified.
I didn't mean to imply that I wanted one but not the other; I just said each one is a good thing even without the other. I'm sorry I created the wrong impression with my comments and didn't clarify this to begin with.
Of course when self-modifying to increase pleasure I'd want to avoid the usual traps - wireheading, certain distortions of my existing balance of values (things I derive pleasure from), etc. But in general I do want to increase pleasure.
I also think reducing negative affect is a much more urgent goal. If I had a choice between reducing pain and increasing pleasure in my life right now, I'd choose reducing pain; and the two cannot (easily) be traded. That's why I said before that "there's something wrong about negative [stuff]".
I agree with a lot of what you're saying, I made errors too, and IMHO apologizing doesn't make much sense, especially in the context of errors, but I'll apologize for my errors too because I desire to compensate for hypothetical status losses that might occur as a result of your apology, and also because I don't want to miss out any more than necessary on hypothetical status gains that might occur as a result of (unnecessary) apologies. But the desire to reciprocate is also within this apology, I'm not just calculating utilons here.
Sorry for my previous errors.
You said:
I said:
I don't know how you avoid this problem except by only supporting modifying incentives in cases of unachievable goals. I'd like to avoid it but I would like to see a mechanism for doing so explicitly stated. If you don't know how to avoid this problem yet, that's fine, neither do I.
Apologizing is indeed status signaling; I feel better in conversations where it is not necessary or expected.
When I said I was sorry, I meant it in the sense of "I regret". I didn't mean it as an apology and wasn't asking for you to reciprocate. (Also, the level of my idiomatic English tends to vary a lot through the day.)
Now I regret using the expression "sorry"!
I'm glad we agree about apologies :-)
As for the problem of modifying (positive) preferences: I don't have a general method, and haven't tried to work one out. This is because I don't have a way to self-modify like this, and if I acquire one in the future, it will probably have limitations, strengths and weaknesses, which would guide the search for such a general method.
That said, I think that in many particular cases, if I were presented with the option to make a specific change, and enough precautions were available (precommitment, gradual modifications, regret button), making the change might be safe enough - even without solving the general case.
I think this also applies to reducing negative affect (not that we have the ability to that, either) - and the need is more urgent there.
It's not just about status. It also communicates "This was an accident, not on purpose" and/or "If given the opportunity, I won't do that again" which are useful information.
I APOLOGIZE FOR MY APOLOGY. :(
Neat, then we agree on all of that. I also would prefer something ad hoc to the "solution" I thought of.