4 ms·
You are assuming that discrimination in favor of males and discrimination in favor of females are like vectors that are parallel and opposite in direction. Und
by logfromblammo 12y ago
You are assuming that discrimination in favor of males and discrimination in favor of females are like vectors that are parallel and opposite in direction. Under that assumption, overall equality could be achieved by selecting and titrating an appropriate magnitude for your own bias, based on the observed bias in the system.
I believe the assumption is flawed. There is a perpendicular component to those vectors, such that any decision to discriminate, at any magnitude, can result in a decrease in fairness for everyone. Thus, the only way to ensure that you are not actually making things worse is to not discriminate based on irrelevant factors like founder gender at all.
If you want to show off the experiences of women founders in YC, make your random sample size large enough that several of them will be represented. Narrowing the pool by eliminating one gender or the other is cheating. So adding "here's 40 male founders" would actually just be making it worse.
The only way to make it better would be to include a representative sampling of founders, chosen without using gender as a consideration.
But the article was motivated by the same premise--that opposing biases can cancel, so that is why it was created with an obvious gender bias. I doubt the intent was malicious, but the results may not be as expected.