Companies are beginning to use facial recognition technology in job interviews and for god’s sake can we stop “disrupting” things with technology that don’t need to be disrupted? As pointed out on Twitter this effectively is a new phrenology.
In a world where we keep having light and motion sensors that can’t detect the skin of black people, a world where, say, autistic people might not have the same “language, tone and facial expressions” as neurotypicals, it’s absolutely untrue that algorithms somehow are “free of human bias”, since they are written by humans with biases.
If the algorithm in question compares the “language, tone and facial expressions” to those of successful employees, it’s just going to loop through the same types of people hired before the algorithm was involved, reinforcing whatever human biases already existed in the hiring system.
This sort of thing will just be used to let human decision-makers off the hook for biases. The algorithm did it, so it must be fair. No, none of this.