I’m curious why this is so. I feel like I get the intuitive pull of suspicious demographic composition is all-else-equal evidence of something wrong, but I have trouble formalizing that intuition into something large and concrete. I guess I just feel like the Bayes factor for such warning signs shouldn’t be very high. Obviously intersectionality/interaction effects are real, but if we vary obvious parameters for a social movement’s origin I don’t feel like our movement should counterfactually be noticeably less concerned. Consider the following phrases:
a largely Chinese, educated, Eastern and male group talking about how to fix everything should not raise flags
a largely white, uneducated, western and male group talking about how to fix everything should not raise flags
a largely white, educated, western and female group talking about how to fix everything should not raise flags
a largely white, educated, western and male group talking about how nothing needs to be fixed should not raise flags
a demographically diverse, uneducated, globalist, gender-neutral group talking about how it’s impossible to fix anything should not raise flags
Each of the above statements continue to seem intuitively suspicious to me, which is at least some evidence that we’re confused here if we ascribe overly high import to the demographics of origin.
I think all these groups need to be concerned, but about different things:
a largely white, educated, western, and male group talking about how to fix everything should (i) not repeat mistakes like racist eugenics, colonialism, etc., historically made by such groups, (ii) also think about other possible problems caused by its members being privileged/powerful
an uneducated group should not repeat mistakes made by past such groups (perhaps famines and atrocities caused by the Great Leap Forward), and anticipate other problems caused by its demographics by ensuring it has good epistemics, talent, and ways of being taken seriously
a largely black group should look at mistakes made by other predominantly black groups including black supremacists (perhaps becoming cults), …
a group talking about how it’s impossible to fix anything should look at mistakes made by past such groups (perhaps Calvinism), …
Also, any group that isn’t demographically diverse might want to become diverse if they think adding voices makes the direction of the movement better.
Some of these concerns can be easily dismissed (the NAACP doesn’t need to try especially hard to not become a black supremacist cult because the prior probability of that is very low). But when thinking about plausible failure modes even a ~3:1 Bayes factor from demographics can be important, until we know about the actual causes of these failures and whether they apply to us.
Or perhaps more simply, if a small, non-representative group disagrees with the majority of humans, we should wonder why, and given base rates and the outside view, worry about failure modes that have affected similar small groups in the past.
In terms of:
I’m curious why this is so. I feel like I get the intuitive pull of suspicious demographic composition is all-else-equal evidence of something wrong, but I have trouble formalizing that intuition into something large and concrete. I guess I just feel like the Bayes factor for such warning signs shouldn’t be very high. Obviously intersectionality/interaction effects are real, but if we vary obvious parameters for a social movement’s origin I don’t feel like our movement should counterfactually be noticeably less concerned. Consider the following phrases:
Each of the above statements continue to seem intuitively suspicious to me, which is at least some evidence that we’re confused here if we ascribe overly high import to the demographics of origin.
I think all these groups need to be concerned, but about different things:
a largely white, educated, western, and male group talking about how to fix everything should (i) not repeat mistakes like racist eugenics, colonialism, etc., historically made by such groups, (ii) also think about other possible problems caused by its members being privileged/powerful
an uneducated group should not repeat mistakes made by past such groups (perhaps famines and atrocities caused by the Great Leap Forward), and anticipate other problems caused by its demographics by ensuring it has good epistemics, talent, and ways of being taken seriously
a largely black group should look at mistakes made by other predominantly black groups including black supremacists (perhaps becoming cults), …
a group talking about how it’s impossible to fix anything should look at mistakes made by past such groups (perhaps Calvinism), …
Also, any group that isn’t demographically diverse might want to become diverse if they think adding voices makes the direction of the movement better.
Some of these concerns can be easily dismissed (the NAACP doesn’t need to try especially hard to not become a black supremacist cult because the prior probability of that is very low). But when thinking about plausible failure modes even a ~3:1 Bayes factor from demographics can be important, until we know about the actual causes of these failures and whether they apply to us.
Mostly endorsed.
Or perhaps more simply, if a small, non-representative group disagrees with the majority of humans, we should wonder why, and given base rates and the outside view, worry about failure modes that have affected similar small groups in the past.