A USC Dornsife opinion paper published in Trends in Cognitive Sciences argues that AI chatbots are homogenizing human language, reasoning, and thought. As billions of people rely on the same LLMs, individual writing styles lose distinctiveness, group creativity declines, and reasoning becomes biased toward linear, Western-centric patterns. The researchers call on AI developers to intentionally incorporate global cognitive and linguistic diversity into LLM training data to preserve humanity's collective intelligence and problem-solving capacity.
Table of contents
Large language models dampen individualityLLMs can reduce the variety of reasoning stylesAbout the studySort: