4 ms·

What is the chances that after going down this rabbit hole, you end up with racist algorithms that match people who are within the same ethnic group? It seems q

What is the chances that after going down this rabbit hole, you end up with racist algorithms that match people who are within the same ethnic group? It seems quite possible that you would find more "compatible" matches with people who are more genetically similar.

matte_black 8y ago
Good question. First of all, people control the genes they are looking for. Algorithms just find people according to given parameters, they are not racist. Second, there’s nothing racist about preferring people in your own ethnic group. Most people find their own race most attractive for a partner, and there’s reason to believe that will still be the case in a post-racism world. Yes there will be mixed couples, but people who want a mate from their own race aren’t racist.
- King-Aaron 8y ago
  > people control the genes they are looking for. Algorithms just find people according to given parameters, they are not racist. Uuuh, while your algorithm isn't racist, if a person is using an algorithm to be racist, then it could be concluded that what is happening is... racism.
  - bdamm 8y ago
    Computers generally cannot solve for the user’s bad intentions. Of course companies do try, and that’s either a value add or a regulatory requirement. Stamping out racism by making it impossible to carry out racist ideas seems like a losing strategy.
    - Zobat 8y ago
      Algorithms are not racist but they might very well be biased, intentionally or not. As a programmer I make assumptions every time I write code and I rarely have 100% insight into the full implications of that code. Sometimes I generate machine learning (ML) solutions and then this holds more true than ever. All the best (least biased) matchmaking solutions will be ML, as well as all the worst (most biased). Also, matching for the best DNA/RNA/gene expression match, wouldn't that just separate us into different species of humans?
  - crypto-jeronimo 8y ago
    > Uuuh, while your algorithm isn't racist, if a person is using an algorithm to be racist, then it could be concluded that what is happening is... racism. This is an extremely weak "argument". Practically, anything can be used as a murder weapon by a murderer. Then "what is happening is ..." murder.
rjf72 8y ago
I'm really curious about your views on something, if you would share: 1) Let's say you could provide definitive data showing, without doubt, that intraracial relationships resulted in different longterm outcomes than interracial relationships, with an extremely high degree of statistical certainty. Would you say that consideration or utilization of this fact would be racist if done in a completely impartial fashion? 2) What if we instead of simply considering intra/inter relationships we instead look the outcomes differences between specific ethnic/racial matchups over the longterm. Would consideration or utilization of these data be considered racist if done in a completely impartial fashion? 3) Let's move away from algorithms and add a human element. There are a variety of, mostly ethnically homogeneous, religions that vehemently forbade marriage outside the religion and where joining the religion, as an outsider, is made extremely difficult. Are these religions inherently racist? Are their practitioners inherently racist? --- And to be clear, as this is a sensitive issue, I'm not secretly trying to find some contradiction or whatever to come back at you with. I just like to try to understand worldviews outside my own and have no intention of challenging any of your answers in any way.
- shard972 8y ago
  > 1) Let's say you could provide definitive data showing, without doubt, that intraracial relationships resulted in different longterm outcomes than interracial relationships, with an extremely high degree of statistical certainty. Would you say that consideration or utilization of this fact would be racist if done in a completely impartial fashion? Probably, it meets most definitions. > 2) What if we instead of simply considering intra/inter relationships we instead look the outcomes differences between specific ethnic/racial matchups over the longterm. Would consideration or utilization of these data be considered racist if done in a completely impartial fashion? Well yes it would be. ˘ 3) Let's move away from algorithms and add a human element. There are a variety of, mostly ethnically homogeneous, religions that vehemently forbade marriage outside the religion and where joining the religion, as an outsider, is made extremely difficult. Are these religions inherently racist? Are their practitioners inherently racist? Yep. Pretty much anything that seeks to understand differences between races will inherently be racist by it's definition.

matte_black 8y ago

Good question. First of all, people control the genes they are looking for. Algorithms just find people according to given parameters, they are not racist. Second, there’s nothing racist about preferring people in your own ethnic group. Most people find their own race most attractive for a partner, and there’s reason to believe that will still be the case in a post-racism world. Yes there will be mixed couples, but people who want a mate from their own race aren’t racist.

King-Aaron 8y ago
> people control the genes they are looking for. Algorithms just find people according to given parameters, they are not racist. Uuuh, while your algorithm isn't racist, if a person is using an algorithm to be racist, then it could be concluded that what is happening is... racism.
- bdamm 8y ago
  Computers generally cannot solve for the user’s bad intentions. Of course companies do try, and that’s either a value add or a regulatory requirement. Stamping out racism by making it impossible to carry out racist ideas seems like a losing strategy.
  - Zobat 8y ago
    Algorithms are not racist but they might very well be biased, intentionally or not. As a programmer I make assumptions every time I write code and I rarely have 100% insight into the full implications of that code. Sometimes I generate machine learning (ML) solutions and then this holds more true than ever. All the best (least biased) matchmaking solutions will be ML, as well as all the worst (most biased). Also, matching for the best DNA/RNA/gene expression match, wouldn't that just separate us into different species of humans?
- crypto-jeronimo 8y ago
  > Uuuh, while your algorithm isn't racist, if a person is using an algorithm to be racist, then it could be concluded that what is happening is... racism. This is an extremely weak "argument". Practically, anything can be used as a murder weapon by a murderer. Then "what is happening is ..." murder.

rjf72 8y ago

I'm really curious about your views on something, if you would share: 1) Let's say you could provide definitive data showing, without doubt, that intraracial relationships resulted in different longterm outcomes than interracial relationships, with an extremely high degree of statistical certainty. Would you say that consideration or utilization of this fact would be racist if done in a completely impartial fashion? 2) What if we instead of simply considering intra/inter relationships we instead look the outcomes differences between specific ethnic/racial matchups over the longterm. Would consideration or utilization of these data be considered racist if done in a completely impartial fashion? 3) Let's move away from algorithms and add a human element. There are a variety of, mostly ethnically homogeneous, religions that vehemently forbade marriage outside the religion and where joining the religion, as an outsider, is made extremely difficult. Are these religions inherently racist? Are their practitioners inherently racist? --- And to be clear, as this is a sensitive issue, I'm not secretly trying to find some contradiction or whatever to come back at you with. I just like to try to understand worldviews outside my own and have no intention of challenging any of your answers in any way.

shard972 8y ago
> 1) Let's say you could provide definitive data showing, without doubt, that intraracial relationships resulted in different longterm outcomes than interracial relationships, with an extremely high degree of statistical certainty. Would you say that consideration or utilization of this fact would be racist if done in a completely impartial fashion? Probably, it meets most definitions. > 2) What if we instead of simply considering intra/inter relationships we instead look the outcomes differences between specific ethnic/racial matchups over the longterm. Would consideration or utilization of these data be considered racist if done in a completely impartial fashion? Well yes it would be. ˘ 3) Let's move away from algorithms and add a human element. There are a variety of, mostly ethnically homogeneous, religions that vehemently forbade marriage outside the religion and where joining the religion, as an outsider, is made extremely difficult. Are these religions inherently racist? Are their practitioners inherently racist? Yep. Pretty much anything that seeks to understand differences between races will inherently be racist by it's definition.