Re:Facebook fires employee for publicly scolding a colleague
by musicale on Friday June 12, 2020 @09:30PM
(#23505843)
Facial recognition may not work reliably if there is training data bias, which can be disadvantageous if you want it to work.
And as you note it can also be used for reducing privacy at scale in a way that humans can't normally do on their own, for example personally identifying thousands of people in a crowd at a sporting event or political rally.
Re:Facebook fires employee for publicly scolding a colleague
by SpicyLemonZest on Saturday June 13, 2020 @10:16AM
(#23509485)
Recoil isn't facial recognition software, it's a React state management library. That's why the employee's conduct was so toxic here; the project he demanded a statement from just has nothing to do with racism in the first place.
While racism is the hot button topic right now, widespread governmental and corporate (or indeed, private) use of facial recognition software is problematic for many reasons. Erosion of privacy, sexual and non-sexual harassment and stalking, racial profiling, gender profiling, profiling for membership of any other kind of group, cute Facebook apps tied to shadowy political organizations, even crazy dystopian inventions like the Chinese social credit scheme. This tech needs to be very strongly regulated, and soon.
Anyone working in this field, and creating this type of software for companies with a historically weak moral compass, _should_ think deeply about their involvement.