The conversation delves into the complexities of measuring biases and their potential harms in AI. Hal expresses the need for a comprehensive catalog of common biases, akin to a Wikipedia page, to streamline understanding and research. He reflects on his own journey through sociolinguistics, highlighting the importance of accessible resources in this evolving field.
