Goto

Collaborating Authors

 eliminate unexpected source


Eliminate Unexpected Sources of Bias in Artificial Intelligence

#artificialintelligence

It is important for companies to implement artificial intelligence, personalization, and other new technologies into the strategy, because it will bring the opportunity to find, interpret, dissect and react on difficult market signals. The more we use predictive AI systems and automated decision-making, the more crucial it is to recognize and eliminate bias from data. We idealize smart systems because they include mathematics as the foundation, thinking that AI platforms are more objective than the developers. But researchers found that artificial intelligence absorbs bias from different resources. For example, Harvard professor Latanya Sweeney detected that the word "arrest" connected with names of black babies.