Epistemic effort: I wrote this on a plane flight. I’m often interested in Ribbonfarmian “consider a bad thing. What if it’s good? (Here’s my favorite example of this.)
As regards updating my beliefs, I’m drawn to motivated snobbery. “Motivated” means “this belief improves my experiences;” “snobbery” means “with this belief, I eliminate a class of problems other people have.”
An example of motivated snobbery is “tipping well.” Here’s my sales pitch: Tipping is an iterated prisoner’s dilemma, not an evaluation of their performance! I want servers and bartenders to be happy to see me. A friend said this well: If I’m getting rich we’re all getting rich.
That this makes bartenders happy and signals to my friends that I’m atypically generous[0] makes this “motivated.” That I’m implying, “The extra money I spend doesn’t hurt me” signals abundance; I can’t be pressured by what pressures others. This is snobbish.
Some nonexamples for me are the slogans-as-beliefs you see on hacker news: “Machine learning is just multiplying matrices.” “Meta-Language is the real ML” “Once we know how to do something with computers, it ceases to be AI, and becomes boring.” Though snobs say these things, I don’t because I’m not (right now) trying to hire an engineer who agrees with those things[1].
A reasonable objection: “So zlrth, you’ll change your beliefs depending on your circumstances?” In a trivial sense, yes. To be a motivated snob you must keep your identity small and avoid mind-killing. Those things suggest not being an ideologue[2]. I don’t think avoiding mind-killing is controversial (at lesswrong, anyway), but if you do, you should let me know. What I think is more up for debate is what heuristics you use to decide your rent-paying beliefs, and their end goals.
Another objection: Things HN’ers say are mostly phatic; not beliefs-that-are-supposed-to-pay-rent. Asked, “What kind of thing is, ‘Meta-Language is the real ML’?” more people would say that it’s an opinion than a belief. I’m not sure what opinions are; I’d rather have rationality-style beliefs. You should elevate your opinions into snobbish beliefs that eliminate for you classes of problems other people have, or dismiss them.
[0] There’s a cynical read: I, an alleged sociopath, get off on being perceived as formidibly rich or risk-tolerant. I don’t have a ready response. You should try out motivated snobbery; see how you feel : ). I don’t feel sociopathic when I do it.
[1] Perhaps there’s value in virtue signaling elitist beliefs about machine learning. I think there’s more, better virtue-signaling in declining to have a belief in the first place.
[2] I don’t think there’s anything wrong with ideologues, and I apologize if I sounds like I’m strawmanning them. My point is that having a small identity is a reasonable, uncontroversial thing to do, and a prerequisite for motivated snobbery.
Consider motivated snobbery
Epistemic effort: I wrote this on a plane flight. I’m often interested in Ribbonfarmian “consider a bad thing. What if it’s good? (Here’s my favorite example of this.)
As regards updating my beliefs, I’m drawn to motivated snobbery. “Motivated” means “this belief improves my experiences;” “snobbery” means “with this belief, I eliminate a class of problems other people have.”
An example of motivated snobbery is “tipping well.” Here’s my sales pitch: Tipping is an iterated prisoner’s dilemma, not an evaluation of their performance! I want servers and bartenders to be happy to see me. A friend said this well: If I’m getting rich we’re all getting rich.
That this makes bartenders happy and signals to my friends that I’m atypically generous[0] makes this “motivated.” That I’m implying, “The extra money I spend doesn’t hurt me” signals abundance; I can’t be pressured by what pressures others. This is snobbish.
Some nonexamples for me are the slogans-as-beliefs you see on hacker news: “Machine learning is just multiplying matrices.” “Meta-Language is the real ML” “Once we know how to do something with computers, it ceases to be AI, and becomes boring.” Though snobs say these things, I don’t because I’m not (right now) trying to hire an engineer who agrees with those things[1].
A reasonable objection: “So zlrth, you’ll change your beliefs depending on your circumstances?” In a trivial sense, yes. To be a motivated snob you must keep your identity small and avoid mind-killing. Those things suggest not being an ideologue[2]. I don’t think avoiding mind-killing is controversial (at lesswrong, anyway), but if you do, you should let me know. What I think is more up for debate is what heuristics you use to decide your rent-paying beliefs, and their end goals.
Another objection: Things HN’ers say are mostly phatic; not beliefs-that-are-supposed-to-pay-rent. Asked, “What kind of thing is, ‘Meta-Language is the real ML’?” more people would say that it’s an opinion than a belief. I’m not sure what opinions are; I’d rather have rationality-style beliefs. You should elevate your opinions into snobbish beliefs that eliminate for you classes of problems other people have, or dismiss them.
[0] There’s a cynical read: I, an alleged sociopath, get off on being perceived as formidibly rich or risk-tolerant. I don’t have a ready response. You should try out motivated snobbery; see how you feel : ). I don’t feel sociopathic when I do it.
[1] Perhaps there’s value in virtue signaling elitist beliefs about machine learning. I think there’s more, better virtue-signaling in declining to have a belief in the first place.
[2] I don’t think there’s anything wrong with ideologues, and I apologize if I sounds like I’m strawmanning them. My point is that having a small identity is a reasonable, uncontroversial thing to do, and a prerequisite for motivated snobbery.