Math can’t solve everything: Questions we need to be asking before deciding an algorithm is the answer

Relying entirely on algorithms to make critical decisions and take immediate action is so remarkably dystopian that I am surprised so many companies are pushing towards doing it. The best you can expect with algorithms currently, as the article points out, is absolutely cold calculation with no fairness, nuance, compassion, ability to discern intent or taking circumstances into account. And with no human oversight or verification before action is taken, a tremendous amount of people fall between the cracks as a result, as evidenced by what's happening on YouTube with demonetization.

There is no doubt that algorithms will eventually become better at humans for many things (and already are, as is the case with diagnosis of multi-factorial diseases and illnesses)... however I am extremely wary of removing the human element.

IMO the best of both worlds is to have the algorithms identify potentials and then have a human review the options and make the actual final decision before action is taken.