Beliefs can pay rent without necessarily directly resulting in anticipated experiences. They could only figure into broader networks of thoughts and concepts that form a unified and internally consistent whole that pays (or tends to pay) rent. Perhaps one might say, then, that models or worldviews should pay rent, even if they have some spandrels that don't, themselves, pay rent.
Once I saw a tweet that endorsed “verificationism, but as a theory of whether I care if a sentence is true.” I haven’t clicked through to read Eliezer’s article, but it sounds like he might be sympathetic to that view.
i see this kind of view espoused a lot, but it seems completely untenable to me. imagine i’m choosing between fruit loops and cocoa puffs for breakfast. i’m leaning towards cocoa puffs, but then i consider the claim “through a nonphysical, unobservable mechanism, having cocoa puffs will cause hundreds of innocents to be tortured in parallel universe we will never have evidence of.”
this claim is unverifiable. do i care whether it is true? YES! massively! if it were true, i definitely would want to go with fruit loops! my choosing cocoa puffs is massively conditional on my strong belief that that claim is false! the reason to disregard the claim is because it’s wildly implausible, not because its implications are insignificant!
Perhaps one ought not to spend very much time concerned with beliefs that cannot pay rent in anticipated experience. You in particular might be happier if you spent less time thinking about things like your five examples. Although, number four can be rephrased in a way that does pay rent by saying that you can anticipate feeling bad about yourself if you cause harm to sentient beings.
Doesn't this mean "the world is on net happer if you don't torture other people", according to you? If so, this is going to cash out in anticipated experience.
Beliefs can pay rent without necessarily directly resulting in anticipated experiences. They could only figure into broader networks of thoughts and concepts that form a unified and internally consistent whole that pays (or tends to pay) rent. Perhaps one might say, then, that models or worldviews should pay rent, even if they have some spandrels that don't, themselves, pay rent.
Once I saw a tweet that endorsed “verificationism, but as a theory of whether I care if a sentence is true.” I haven’t clicked through to read Eliezer’s article, but it sounds like he might be sympathetic to that view.
i see this kind of view espoused a lot, but it seems completely untenable to me. imagine i’m choosing between fruit loops and cocoa puffs for breakfast. i’m leaning towards cocoa puffs, but then i consider the claim “through a nonphysical, unobservable mechanism, having cocoa puffs will cause hundreds of innocents to be tortured in parallel universe we will never have evidence of.”
this claim is unverifiable. do i care whether it is true? YES! massively! if it were true, i definitely would want to go with fruit loops! my choosing cocoa puffs is massively conditional on my strong belief that that claim is false! the reason to disregard the claim is because it’s wildly implausible, not because its implications are insignificant!
I feel like I could somehow realise I'd been wrong on 2 and 5.
Perhaps one ought not to spend very much time concerned with beliefs that cannot pay rent in anticipated experience. You in particular might be happier if you spent less time thinking about things like your five examples. Although, number four can be rephrased in a way that does pay rent by saying that you can anticipate feeling bad about yourself if you cause harm to sentient beings.
> It’s wrong to torture other people.
Doesn't this mean "the world is on net happer if you don't torture other people", according to you? If so, this is going to cash out in anticipated experience.
That’s not what it means but that is what determines if it’s true