A World "Averaged" by AI: What We Lose Behind Convenience - The Brain Slacks Off, Language Becomes Bland - How LLMs Are Changing Cultural Diversity

Generative AI "averages" human thought and expression—recent research introduced by WIRED indicates that the brain activity of LLM users decreases, and the connectivity related to creativity and working memory tends to weaken. Students' essays become similar to each other, and the diversity of opinions also diminishes. A study from Cornell University found that writing from both the U.S. and India leaned towards Western norms (such as pizza and Christmas), and vocabulary became abstracted. At Santa Clara University, responses to creative tasks also converged. On social media, while some argue that "homogenization is not inherently bad" and that it can be avoided through human "editing," there are many concerns that "outsourcing thought" erodes individual voices and cultural uniqueness. In the latter part of the article, practical measures for education, marketing, and product development are proposed—such as "AI-first drafts with human counter-editing," the injection of specific proper nouns, blending multilingual corpora, breaking away from template prompts, and offline thinking time. The key is to design a "de-averaging" approach while leveraging the power of AI.