The Human Toll of Making ChatGPT Safe for Users

6,133 次觀看・4 個月前
Generative AI programs rely on human workers to train them to filter out harmful content, including violent and sexual material. WSJ reporter Karen Hao traveled to Kenya to speak with some quality analysts, who say the work took a toll on their mental well-being. Photo: Tony Karumba/Agence France-Presse/Getty Images