In his book Talking with Strangers, Malcolm Gladwell discusses an AI experiment that looked at 554,689 bail hearings conducted by New York City judges. As one online publication noted, "Of the more than 400,000 people released, over 40% either failed to appear at their subsequent trials or were arrested for another crime." However, decisions recommended by the machine learning algorithm on whom to detain or release would have resulted in 25% fewer crimes. This is an example of an AI system that is less biased than a human. However, that is not always the case. Read more on Forbes.com