But, like, the world is on fire and you can do something about it and you do obviously need to be healthy. And part of being healthy is not just saying things like “okay, I guess I can indulge things like not spending 100% of my resources on saving the world in order to remain healthy but it’s a necessary evil that I feel guilty about.”
AFAICT, the only viable, sane approach is to acknowledge all the truths at once, and then apply a crude patch that says “I’m just going to not think about this too hard, try generally to be healthy, put whatever bit of resources towards having the world not-be-on-fire that I can do safely.
My view is that there is indeed a principled resolution, which should be pursued by people aspiring to be more lawful and coherent. But the resolution requires nontrivial skills to implement. The key insight is that certain gut reactions should be viewed as policy-level choices, instead of correct, primitive moral evaluations of the situation.
Many people are deeply confused about their feelings, and what they’re “for”—should they be modified? Are feelings reflections of ground-truth morality? In particular, if I feel guilty about spending money on myself, does that mean it is reflectively-correct to feel that way?
The answer, of course, is a resounding No!.
Unfortunately, this can be hard to see, since our motivational systems are not well-typed—expected-utility and utility feel the same from the inside, and executed-heuristic versus reflectively-consistent-judgement are not primitive internal observables.
Notice that spending money on myself is not an intrinsic bad. Therefore, any guilt I feel must be the result of an instrumental value-function heuristic which fires when I take such actions, perhaps because “take selfish action while someone else needs your help” is (societally-, personally-)usually an action which leads to lower-value states, and—eventually—to outcomes with lower terminal utility.[1]
Since my guilt does not reflect an intrinsic bad, it is “up for grabs”—the “guilt” heuristic constitutes a cognitive strategy, which I can choose to execute or not, depending on its logical and causal effects.
After all, I’m (basically) optimizing for outcomes. From the FDT standpoint, there is no need to have an angsty internal struggle over these facts, as if I were living out the script of a hero who is dutifully remorseful about daring purchase a luxury for themselves. I simply choose the way of being which works out bestTurnTrout's values. The rest is noise.
(I know that from a certain inferential distance, the advice may seem trivial and laughably impractical, and if it does, I don’t immediately see how to bridge the gap. And I describe the closer-to-ideal standard as I understand it. I have moved some good distance towards it, but I am not yet able to fluently interact with myself in this way.)
My view is that there is indeed a principled resolution, which should be pursued by people aspiring to be more lawful and coherent. But the resolution requires nontrivial skills to implement. The key insight is that certain gut reactions should be viewed as policy-level choices, instead of correct, primitive moral evaluations of the situation.
Many people are deeply confused about their feelings, and what they’re “for”—should they be modified? Are feelings reflections of ground-truth morality? In particular, if I feel guilty about spending money on myself, does that mean it is reflectively-correct to feel that way?
The answer, of course, is a resounding No!.
Unfortunately, this can be hard to see, since our motivational systems are not well-typed—expected-utility and utility feel the same from the inside, and executed-heuristic versus reflectively-consistent-judgement are not primitive internal observables.
Notice that spending money on myself is not an intrinsic bad. Therefore, any guilt I feel must be the result of an instrumental value-function heuristic which fires when I take such actions, perhaps because “take selfish action while someone else needs your help” is (societally-, personally-)usually an action which leads to lower-value states, and—eventually—to outcomes with lower terminal utility.[1]
Since my guilt does not reflect an intrinsic bad, it is “up for grabs”—the “guilt” heuristic constitutes a cognitive strategy, which I can choose to execute or not, depending on its logical and causal effects.
After all, I’m (basically) optimizing for outcomes. From the FDT standpoint, there is no need to have an angsty internal struggle over these facts, as if I were living out the script of a hero who is dutifully remorseful about daring purchase a luxury for themselves. I simply choose the way of being which works out bestTurnTrout's values. The rest is noise.
(I know that from a certain inferential distance, the advice may seem trivial and laughably impractical, and if it does, I don’t immediately see how to bridge the gap. And I describe the closer-to-ideal standard as I understand it. I have moved some good distance towards it, but I am not yet able to fluently interact with myself in this way.)
I don’t think people’s desires are well-represented by utility functions, but I think the theory works fine in this situation.