>> "The systems were most accurate with cisgender men and women, who on average were accurately classified 98% of the time. Researchers found that trans men wer

>> "The systems were most accurate with cisgender men and women, who on average were accurately classified 98% of the time. Researchers found that trans men were wrongly categorized roughly 30% of the time. The tools fared far worse with non-binary or genderqueer people, inaccurately classifying them in all instances. " I guarantee this sort of tech will be misused by every kiosk in malls and every sample person in stores to push products on individuals. Amazon at least tried to pre-empt this concern, but this really isn't enough. Not that I know what is. I know the technology is coming, so pointing out what an ill-considered folly the whole thing is won't do much. >> "Scheuerman said that while he believed Amazon’s guidelines were well-intended, the study points out what’s wrong with the service. There’s no guarantee that Rekognition’s clients are using the facial analysis tool as Amazon intended. And even if such data is viewed in aggregate, it would very likely still be incorrect."