
A new paper found that large language models from OpenAI, Meta, and Google, including multiple versions of ChatGPT, can be covertly racist against African Americans when analyzing a critical part of their identity: how they speak.
A new paper found that large language models from OpenAI, Meta, and Google, including multiple versions of ChatGPT, can be covertly racist against African Americans when analyzing a critical part of their identity: how they speak.
Richard Vogel/AP Photo
Antonio Masiello via Getty Images; Christian Petersen via Getty Images
Christopher Willard/Getty Images
twitch/just_relax_kid
Chief video correspondent Graham Flanagan got an inside look at what it's like on the first day for new freshmen, known as knobs, at the Citadel in Charleston, South Carolina.
Nicolas Economou/NurPhoto via Getty Images
REUTERS/Ivan Alvarado/File Photo