Entry

Homogenizing effect of large language models (LLMs) on creative diversity: An empirical comparison of human and ChatGPT writing

Kibum Moon, Adam E. Green, Kostadin Kushlev

Synopsis

Three studies analyzing 2,200 essays; quantifies "homogenizing effect" — each additional human essay contributes more new ideas than additional GPT-4 essays.

Keywords

·homogenization ·creative diversity ·GPT-4 ·essays ·novelty rate

Open paper ↗ Report issue ↗

Related entries

Does Writing with Language Models Reduce Content Diversity?

September 11, 2023 · ICLR 2024
Large Language Model Influence on Diagnostic Reasoning: A Randomized Clinical Trial

October 28, 2024 · JAMA Network Open 2024
Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality

September 15, 2023 · Organization Science 2026
Strategic Algorithmic Monoculture: Experimental Evidence from Coordination Games

April 10, 2026 · arXiv
Modeling Distinct Human Interaction in Web Agents

February 19, 2026 · arXiv
When Stereotypes GTG: The Impact of Predictive Text Suggestions on Gender Bias in Human-AI Co-Writing

September 30, 2024 · CHI 2026