Hi, I’m a person who writes algorithms & I have a lot of thoughts on algorithms that result in biases against women & people of color, which seems to be coming up more and more as algorithms start to be deployed in various contexts that could have deep effects on people’s lives. https://twitter.com/dhh/status/1192540900393705474
There is evidence all over the place of algorithmic bias, not just in Apple credit cards, but also in algorithms used for hiring employees, sentencing and bails, facial recognition, and many other contexts that affect a person's line of credit or even length of prison sentence.
Companies who deploy algorithms often claim that algorithms are more unbiased than if a person were making the decision because people have biases but a piece of code does not.
Often they will claim that because the algorithm does not include gender or ethnicity its results cannot be biased by those metrics.
(I apparently messed up the twitter threading, but the thread continues here) https://twitter.com/uajamie/status/1194700584751915010?s=21
You can follow @UAJamie.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: