← Back to context

Comment by jakupovic

2 years ago

I'll answer for the previous post. The most disturbing part is stating main criteria is being a male and their models have 10% error rate.

I don't think you're parsing the article correctly.

There is no allegation that the main criteria for the algorithm is "being male."

The allegation is that the human double-checking of the algorithm confirms the target is male (as opposed to woman/child.)