|
I'm not using this data to treat people differently: If I'm alone at 3AM and you come out of nowhere, I wouldn't care if you were blue, green, black or white. I wouldn't trust anyone equally. My point was, that one can understand how people think and stereotype given the data, as long as the data isn't corrupted. But coming back to the data: Sure, the majority of people aren't criminals, but the few who are, belong to a certain category. I'm not talking absolute, I'm talking relative. I'm not talking "divine truth", I'm talking correlation. And you can be a sunflower all you want, you can't deny it. You can deny it if you want, but that doesn't make it any less true, and that doesn't solve the problems in urban areas, poverty, lack of education, etc. So: One can either say it's not true and be liked by people as an open mind and non racist (that doesn't fix the root problems), or one acknowledges the problem and think about solutions (at the risk of being called names). I do feature extraction and classification of data, can't go against such stats. That's why I asked the commenter to put herself in their shoes: She's got a kid, between a 45 year old dude and a more usual nanny, she'll choose the nanny every single time (from the 60%). To be closer to her experience: She'll do an ad, a couple comes, she'll acknowledge the guy and ask the woman how much experience she's had with kids .. While it's the guy who came for the job. |