Ep. 18: How AI Learns Racism, Sexism

Aylin Caliskan, a Princeton researcher and co-author of a groundbreaking paper on AI bias, discusses the alarming ways AI can inherit societal prejudices. She unpacks how AI language models reflect racism and sexism, highlighting findings from the Implicit Association Test. Caliskan explains the consequences of biased AI in recruitment processes, leading to unfair treatment of candidates. She advocates for increased human oversight to mitigate these biases, emphasizing both the challenges and the potential for AI to help reveal and address unconscious prejudices.