I can’t follow the flow of logic which seems to me to rely on strange assumptions: e.g. considering a group of people who are exactly the same and react the same way. This is not a puzzle about an island full of logicians with some words on their foreheads, presumably you’re talking about real life.
I understand the standard dialectic spiral, but I don’t understand what “philosophically correct” means. I am also not clear how your concern for which way the incentive gradient goes is different from the simple “there should be more of our kind”.
I can’t follow the flow of logic which seems to me to rely on strange assumptions: e.g. considering a group of people who are exactly the same and react the same way. This is not a puzzle about an island full of logicians with some words on their foreheads, presumably you’re talking about real life.
This seems like standard objection #353, (it often appears e.g. when people argue that Newcomb’s dilemma is irrelevant etc). The standard answer applies here: in the real world, you replace “equality” with “correlation”.
I understand the standard dialectic spiral
Thanks for the reference, it seems that I have reinvented this particular wheel
but I don’t understand what “philosophically correct” means
Something like “agents with enough decision-theoretic sophistication will converge on this outcome”
I am also not clear how your concern for which way the incentive gradient goes is different from the simple “there should be more of our kind”.
Yeah, this is a good question, and there’s a lot of subtle differences. First of all, “there should be more of our kind” is wishful thinking (a useless, impossible counterfactual), it draws an arbitrary distinction between groups of people, it assumes it’s still OK for some people to be in a bad situation by design, etc. I’m sure you can come up with more complaints. OTOH, looking at the incentive gradient for the whole population avoids these problems, and also is not morally abhorrent.
The problem with this is that the setup suddenly becomes a whole lot handwavy. Equality is mathematically “hard”, correlation is not, it’s a statistical “soft”. Which correlation (Pearson’s? why this one?), how much is enough (is 5% sufficient? 50%? does it depend on the amount of noise present?), etc.
For example you say:
Everyone can expect everyone else in the group to follow the same logic, and get the same conclusion. It’s not possible to game the system
but if equality is replaced with correlation, this doesn’t hold.
“agents with enough decision-theoretic sophistication will converge on this outcome”
Regardless of their values?
“there should be more of our kind” is wishful thinking
I disagree. Human history is full of “there should be more of our kind and we will make it happen”. The typical route is killing competitors and taking their resources. There are also cultural factors (e.g. the “protect the women” virtue), religious/ideological influences (see the population growth in Roman Catholic countries until recently), etc.
I can’t follow the flow of logic which seems to me to rely on strange assumptions: e.g. considering a group of people who are exactly the same and react the same way. This is not a puzzle about an island full of logicians with some words on their foreheads, presumably you’re talking about real life.
I understand the standard dialectic spiral, but I don’t understand what “philosophically correct” means. I am also not clear how your concern for which way the incentive gradient goes is different from the simple “there should be more of our kind”.
This seems like standard objection #353, (it often appears e.g. when people argue that Newcomb’s dilemma is irrelevant etc). The standard answer applies here: in the real world, you replace “equality” with “correlation”.
Thanks for the reference, it seems that I have reinvented this particular wheel
Something like “agents with enough decision-theoretic sophistication will converge on this outcome”
Yeah, this is a good question, and there’s a lot of subtle differences. First of all, “there should be more of our kind” is wishful thinking (a useless, impossible counterfactual), it draws an arbitrary distinction between groups of people, it assumes it’s still OK for some people to be in a bad situation by design, etc. I’m sure you can come up with more complaints. OTOH, looking at the incentive gradient for the whole population avoids these problems, and also is not morally abhorrent.
The problem with this is that the setup suddenly becomes a whole lot handwavy. Equality is mathematically “hard”, correlation is not, it’s a statistical “soft”. Which correlation (Pearson’s? why this one?), how much is enough (is 5% sufficient? 50%? does it depend on the amount of noise present?), etc.
For example you say:
but if equality is replaced with correlation, this doesn’t hold.
Regardless of their values?
I disagree. Human history is full of “there should be more of our kind and we will make it happen”. The typical route is killing competitors and taking their resources. There are also cultural factors (e.g. the “protect the women” virtue), religious/ideological influences (see the population growth in Roman Catholic countries until recently), etc.
And I still cannot follow your train of logic.