Hacker News Reader

Comment by elvinyung | Hacker News Reader

I mean, don't you think it's kind of a strawman to assume anyone thinks machines are explicitly sexist? It seems much more likely that whatever method they use to do feature selection on this obviously high-dimensional data just happens to end up picking something that was a proxy for gender.

You can be implicitly biased without being aware of it. This is true for both humans and algorithms.

View on HN

> It seems much more likely that whatever method they use to do feature selection on this obviously high-dimensional data just happens to end up picking something that was a proxy for gender.

Yes, this is exactly what everyone thinks happened.

Nobody thinks people at Goldman-Sachs wrote

    if (applicant.gender == "F")
        limit /= 20

somewhere in their algorithm.

But regardless of how the thing happened, if millions of people are treated significantly differently for no reason other than their plumbing, that's a major problem. People have been talking about this "accidental proxy for gender" for years now; there's absolutely no excuse for no doing a basic sanity check to make sure that this kind of thing isn't happening.

edit: typo

"for no other reason than their plumbing" is exactly what you just excluded in the prior sentence.

It is absolutely not a strawman. It is incredibly easy to find people granted authority by the state that will claim that, eg, mathematics per se is an instance of exclusionist masculine thinking. It's even easier to find instances where people granted authority by the state will claim that any discrepancy in outcomes is ipso facto intentional.

Reply on news.ycombinator.com