UHD Business Professor's AI Research Cited in Nature — and It Has Good News for Creativity Page Title
Dr. Yun Wan of the Marilyn Davies College of Business finds that strategically designed AI personas can preserve — and even enhance — human creative diversity, challenging a widely accepted assumption about generative AI.
A UHD Marilyn Davies College of Business faculty member's research on generative artificial intelligence has earned a citation in Nature — one of the world's most prestigious scientific journals — placing the Marilyn Davies College of Business (MDCOB) at the center of one of today's most urgent conversations in technology research.
Dr. Yun Wan, Associate Professor of Management Information Systems in MDCOB's Department of Finance, Information Systems, Economics and Risk Management, co-authored the paper "Diverse AI Personas Can Mitigate the Homogenization Effect in Human-AI Collaborative Ideation" alongside Dr. Yoram M. Kalman. The paper was recently accepted for publication in Computers in Human Behavior: Artificial Humans and cited this March in a Nature news feature exploring whether AI is narrowing human creative expression.
The Problem: Is AI Making Everyone Think Alike?
A widely cited 2024 study established a troubling tradeoff: while generative AI tools can improve the quality of individual creative work, they tend to reduce the collective diversity of what groups produce. When everyone draws from the same AI for inspiration, the results converge. This is a phenomenon Nature describes as AI "same-ifying" human expression.
The implications reach well beyond creative writing. If AI homogenizes thought at scale, the effects touch business innovation, product development, academic inquiry, and the very originality that drives competitive advantage.
The MDCOB Study: A Design Fix, Not a Dead End
Dr. Wan and Dr. Kalman questioned whether this homogenization was an inherent flaw of generative AI or just a flaw in how AI is typically deployed.
Their answer came through an elegant redesign of the original experiment. Rather than exposing participants to a single, generic AI, the researchers built ten distinct generative AI personas each defined by unique combinations of cultural backgrounds, thinking styles, and creative preferences. These personas collectively generated a pool of 300 story plot ideas.
The findings were clear:
- When all story ideas came from a single AI persona, the results were strikingly similar to one another. As if every idea was coming from the same voice, with the same habits and assumptions baked in.
- When ideas came from ten different AI personas, each with distinct backgrounds, styles, and preferences, the variety increased dramatically. Different perspectives, different outputs.
- Most importantly, when human participants wrote stories using these varied AI-generated ideas, their collective work was just as diverse and original as stories written with no AI help at all — meaning the typical "AI flattening effect" was completely reversed.
The study also found that AI-assisted stories showed greater diversity in descriptive and emotional language compared to purely human-generated work suggesting that well-designed AI collaboration can enhance human expression.
Why Nature Took Notice
Nature's March 2026 news feature draws on emerging research in cognitive science, human-computer interaction, and information systems to examine whether AI is reshaping, and narrowing, human thought. Dr. Wan's paper is positioned as a significant counterpoint: evidence that AI-driven homogenization is not inevitable when systems are intentionally designed for diversity.
A citation in Nature's news section is a mark of exceptional research impact. Published by Springer Nature, Nature consistently ranks among the highest-impact scientific publications in the world. Recognition at this level signals that the research speaks not just to a specialist audience, but to the global scientific community at large.
What It Means for Business, Education, and AI Policy
For business teams: Organizations using AI for brainstorming, product ideation, or content creation should rethink prompt design. A single AI "voice" can narrow the ideation funnel. Dr. Wan's research points to persona diversity as a practical, implementable strategy for richer creative outputs — a meaningful competitive advantage.
For students and educators: How AI is used matters as much as whether it is used. This research provides an empirical foundation for classroom conversations about AI and originality. UHD's Center for Teaching and Learning Excellence offers resources to help faculty and students engage with generative AI thoughtfully.
For policymakers and AI developers: AI homogenization is not a technical inevitability — it is a design choice. Building for input diversity produces diverse output. That is a concrete, actionable principle for anyone shaping how AI tools are built and deployed.
About Dr. Yun Wan
Dr. Yun Wan is an Associate Professor of Management Information
Systems at the Marilyn Davies College of Business, UHD Marilyn Davies College of Business. His research spans electronic commerce, consumer decision-making, artificial intelligence, and human-computer interaction, with work published across leading.