Stanford University: Assessing Political Bias in Language Models

Stanford University: Assessing Political Bias in Language Models. “…new research from Stanford University has quantified exactly how well (or, actually, how poorly) [Large Language Models] align with opinions of U.S. demographic groups, showing that language models have a decided bias on hot-button topics that may be out of step with general popular sentiment.”