Can AI Develop Its Own Biases? – Ep. 18
Researchers discover large language models can spontaneously develop new social biases. A conversation with Addison Wu, a co-author of the study and research student at Tom Griffiths' Computational Cognitive Science Lab.