Why AI Must Be Biased, and How We Can Respond
Like physics and biology, computation is a natural process with natural laws. We are making radical progress in artificial intelligence because we have learnt to exploit machine learning to capture existing computational outputs developed and transmitted by humans with human culture. This powerful strategy unfortunately undermines the assumption that machined intelligence, deriving from mathematics, would be pure and neutral, providing a fairness beyond what is present in human society. In learning the set of biases that constitute a word's meaning, AI also learns patterns some of which are based on our unfair history. Addressing such prejudice requires domain-specific interventions.
Joanna J. Bryson is a transdisciplinary researcher on the structure and dynamics of human- and animal-like intelligence. Her research covers topics ranging from artificial intelligence, through autonomy and robot ethics, and on to human cooperation. She holds degrees in Psychology from Chicago (AB) and Edinburgh (MPhil), and Artificial Intelligence from Edinburgh (MSc) and MIT (ScD). She has additional professional research experience from Oxford, Harvard, and LEGO, and technical experience in Chicago's financial industry, and international organization management consultancy. Bryson is presently a Reader (associate professor) at the University of Bath, and an affiliate of Princeton's Center for Information Technology Policy.