Data biases can skew outcomes of AI-based systems
Artificial intelligence systems can be as beneficial or as detrimental as the data fed into them, according Fulton Schools Professor Subbarao Kambhampati, a recent former president of the international Association for the Advancement of Artificial Intelligence. In a talk presented by the Manthan Forum for Public Discourse in India, Kambhampati said the challenges posed by increasing reliance on AI technology stem from biased data being put into AI systems that skews the outcomes those systems produce.