Ask a thousand married rationalists of a given school to estimate the probability that their spouses have cheated on them. Confidentially ask their spouses if they have. Measure group calibration.
ETA: This applies to any potentially painful, but verifiable question. Ask them to draw a probability distribution over their date of death, or the longevity of their marriages. Estimate the probability of various kinds of cancer appearing over the next (5,10,15) years, etc. etc.
I’ve thought of a problem with this: if rationality is about /Winning/, then it may be rational to not consider the hypothesis that your wife cheats on you. You may better serve your preferences if you remain in blissful ignorance. Also, human relationships have a very Newcomb-like feel to them, because other humans are very good at ascertaining your true beliefs. If you are entertaining the hypothesis seriously, your wife will probably detect it.
So in this case winning and having a map that accurately reflects the territory may be anti-aligned.
You may better serve your preferences if you remain in blissful ignorance.
There is a difference between wanting not to be a cuckold and wanting not to believe that you are a cuckold. I want the former.
Also, human relationships have a very Newcomb-like feel to them, because other humans are very good at ascertaining your true beliefs. If you are entertaining the hypothesis seriously, your wife will probably detect it.
Presumably, if you are entertaining the hypothesis—at least beyond a societal average, or some such—there is a root problem already in play.
But yes, this does have some self-fulfilling aspects which make it rather hard to model well.
On introspection, this does agree with my preferences, yes.
That does complicate things—I’m not sure how to resolve this one.
I think we are using the world “rationalist” to cover too many meanings. One highly socially useful meaning for the word would be “person who can be reliably expected to speak the truth”. Whatever you choose to call those, it’d certainly be useful to have some around for any society you’d like to build. We would want to have some tests to identify them.
Ask a thousand married rationalists of a given school to estimate the probability that their spouses have cheated on them. Confidentially ask their spouses if they have. Measure group calibration.
ETA: This applies to any potentially painful, but verifiable question. Ask them to draw a probability distribution over their date of death, or the longevity of their marriages. Estimate the probability of various kinds of cancer appearing over the next (5,10,15) years, etc. etc.
I’ve thought of a problem with this: if rationality is about /Winning/, then it may be rational to not consider the hypothesis that your wife cheats on you. You may better serve your preferences if you remain in blissful ignorance. Also, human relationships have a very Newcomb-like feel to them, because other humans are very good at ascertaining your true beliefs. If you are entertaining the hypothesis seriously, your wife will probably detect it.
So in this case winning and having a map that accurately reflects the territory may be anti-aligned.
There is a difference between wanting not to be a cuckold and wanting not to believe that you are a cuckold. I want the former.
Presumably, if you are entertaining the hypothesis—at least beyond a societal average, or some such—there is a root problem already in play.
But yes, this does have some self-fulfilling aspects which make it rather hard to model well.
For me the biggest problem is that many people’s preferences will be:
(a) wanting to not be cheated on
AND
(b) wanting to trust the other person so much that the possibility doesn’t even arise.
i.e. your preferences in this area are a function of your own mind-state.
On introspection, this does agree with my preferences, yes.
That does complicate things—I’m not sure how to resolve this one.
I think we are using the world “rationalist” to cover too many meanings. One highly socially useful meaning for the word would be “person who can be reliably expected to speak the truth”. Whatever you choose to call those, it’d certainly be useful to have some around for any society you’d like to build. We would want to have some tests to identify them.
You’d have to define ‘cheated on’. A fair number of the most rational folks I know live in non-traditional marriage arrangements.
This is entirely true. We’re going for emotional effect, so on that test, I’d keep it to the self-identified monogamists
Perhaps because they realise the real probability of cheating.