I think this was my way of saying that it makes sense as an instrumental rationality technique to afford people at least some positive level of respect (as opposed to negative respect levels, or overall disrespect) regardless of their current world saving position. I could say all that in the article, but it sounds mealy-mouthed that way.
So my advice is that if you’re really a “respect-Bayesian” and you have to account for evidence (so you’re duty bound to adjust downward), try not to update others’ total respect value below zero over this. Or move your zero-floor down so that almost everyone has a positive value both a priori and in practice.
At this point, when you start discussing “positive/negative respect”, I’d need to ask what that means in even more detail. What defines the “zero point”, why would you have a total order (“levels”), why is this an interesting concept. Again, I see the affect, the surface promise of meaning, but not any straightforward way of discerning what’s actually meant.
(I agree that with any reasonable guesses at the concepts, “respect” going into the “negative” because of not saving the world in the case of not being aware of the arguments is incorrect, but I don’t appreciate the abundance of apparently arbitrary detail in your explanation.)
At this point, when you start discussing “positive/negative respect”, I’d need to ask what that means in even more detail. What defines the “zero point”, why would you have a total order (“levels”), why is this an interesting concept. Again, I see the affect, the surface promise of meaning, but not any straightforward way of discerning what’s actually meant.
One possible definition of a zero point would be signaling (or being perceived to signal) neither a raising nor a lowering of the status of the person in question. So the imperative could be reformulated as “don’t make moves to lower other people’s status in interactions with them”.
One possible definition of a zero point would be signaling (or being perceived to signal) neither a raising nor a lowering of the status of the person in question.
(It isn’t your imperative but...) High status people will often take that as disrespectful.
I understand treating higher status people like you would treat equal status people as signaling a lowering of their status so I think that’s already taken into account.
I understand treating higher status people like you would treat equal status people as signaling a lowering of their status so I think that’s already taken into account.
Not necessarily. Status is transactional and dynamic. High status people (of a certain kind) demand a constant stream of ‘status raising’ behaviors in the same way governments demand taxes.
I would add that the advice would seem better replaced with “for the purpose of social signalling don’t be a respect-Bayesian”. Now it seems to be “bias your bayesian updating such that your posterior respect gives desirable signals”.
(Although in the absence of the unpacking I can only infer.)
I think this was my way of saying that it makes sense as an instrumental rationality technique to afford people at least some positive level of respect (as opposed to negative respect levels, or overall disrespect) regardless of their current world saving position. I could say all that in the article, but it sounds mealy-mouthed that way.
So my advice is that if you’re really a “respect-Bayesian” and you have to account for evidence (so you’re duty bound to adjust downward), try not to update others’ total respect value below zero over this. Or move your zero-floor down so that almost everyone has a positive value both a priori and in practice.
At this point, when you start discussing “positive/negative respect”, I’d need to ask what that means in even more detail. What defines the “zero point”, why would you have a total order (“levels”), why is this an interesting concept. Again, I see the affect, the surface promise of meaning, but not any straightforward way of discerning what’s actually meant.
(I agree that with any reasonable guesses at the concepts, “respect” going into the “negative” because of not saving the world in the case of not being aware of the arguments is incorrect, but I don’t appreciate the abundance of apparently arbitrary detail in your explanation.)
One possible definition of a zero point would be signaling (or being perceived to signal) neither a raising nor a lowering of the status of the person in question. So the imperative could be reformulated as “don’t make moves to lower other people’s status in interactions with them”.
(It isn’t your imperative but...) High status people will often take that as disrespectful.
I understand treating higher status people like you would treat equal status people as signaling a lowering of their status so I think that’s already taken into account.
Not necessarily. Status is transactional and dynamic. High status people (of a certain kind) demand a constant stream of ‘status raising’ behaviors in the same way governments demand taxes.
I would add that the advice would seem better replaced with “for the purpose of social signalling don’t be a respect-Bayesian”. Now it seems to be “bias your bayesian updating such that your posterior respect gives desirable signals”.
(Although in the absence of the unpacking I can only infer.)