If Racial Equity is Our North Star, How Can We Navigate Racial Bias in AI/LLMs?
John Kenyon
APRIL 11, 2024
They may associate words like “fantastic” and “great” more with white-sounding names versus Black-sounding names. Feed the system explicit instructions and examples displaying equitable behavior. Researchers have consistently found LLMs exhibiting dangerous biases along racial lines. Humans must apply an equity lens.
Let's personalize your content