Generative AI and the Homogenization Trap
FLASH TAKE: How to START Thinking Like a Culture Futurist.
Quick observations on how to S.T.A.R.T. thinking like a Culture Futurist, where I connect research articles to bigger questions shaping creativity in work and culture.
TODAY’S FOCUS STUDY: Generative AI enhances individual creativity but reduces the collective diversity of novel content
Doshi, A. R., & Hauser, O. P. (2024). Generative AI enhances individual creativity but reduces the collective diversity of novel content. Science Advances. https://doi.org/adn5290
Let’s S.T.A.R.T. Looking at the Doshi & Hauser Study
SIGNAL — What is the signal that matters right now?
Study suggests that generative AI boosts the creative output of individual writers, particularly those who struggled before. When results are viewed collectively, however, the range of ideas contracts.
The signal is that AI can improve individual fluency while compressing the diversity of cultural expression.
TILT — How does this signal tilt the way we normally think?
AI has been promoted as a way to multiply creativity. The study suggests the opposite effect at the group level. As more people rely on the same generative system, the pool of outcomes becomes more uniform.
For leaders who expect AI to unlock a steady flow of novelty, the tilt is in recognizing that the same adoption curve that raises performance may also drain differentiation.
ANCHOR — What evidence, story, or framework grounds the tilt?
In controlled experiments, participants were asked to write stories with and without AI assistance. The results were consistent:
Writers with lower creativity scores produced work comparable to higher-scoring peers when assisted by AI.
AI-supported stories were judged more engaging, better written, and more creative.
Across the group, outputs converged toward similarity, reducing variety.
The anchor is that creativity measured only by individual polish gives an incomplete picture. Systems thrive on range. Without it, adaptation slows.
REVEAL — What hidden tension or deeper paradox does the anchor expose?
AI functions both as an equalizer and a limiter. It gives individuals new expressive capacity while narrowing the spread of ideas available to the collective.
The deeper reveal is that organizations depend on variation to adapt to disruption. Homogenization is not only a cultural concern; it is a structural risk to resilience and long-term competitiveness.
TURN — Where does this take us?
Leaders can focus on building practices that widen, rather than compress, the creative field. This includes deliberate variation in prompts, diverse team inputs, and rotating perspectives in how AI is applied.
A practical first step is to compare team outputs from shared prompts versus distinct prompts. The gap between those spreads shows how much space exists for competitive imagination inside your system.
Why This Matters for the Wonder Economy
The Wonder Economy treats creativity as infrastructure. If AI compresses novelty, organizations need to develop Creative Brain Capital: measurable capacity of teams and organizations to transform uncertainty into enterprise value.
AI is a powerful tool, but its direction depends on system design. With intention, it can be part of infrastructures that protect variation while extending access.
The arts have long served as laboratories of divergence, and neuroscience confirms the value of diversity of thought for adaptive intelligence. Together, they provide the frameworks for embedding AI into systems that keep imagination expansive.
The study by Doshi and Hauser (2024) is both signal and design brief. The central challenge is how to build infrastructures that preserve imagination at scale.
RESEARCH OBSERVATION
I value this research because it gives a clear experimental signal: AI can raise individual creative performance while shrinking the range of collective outcomes. At the same time, there are limits in how the study was designed that matter if we want to connect the findings to the larger systems where creativity lives.
First, the study used short story writing as its test. That is useful for a controlled experiment, but it represents only a narrow slice of what creativity looks like in practice. In business, science, and culture, creative work often happens under pressure, with long feedback cycles, and in teams rather than individuals writing alone.
Second, the way creativity was measured leaned on subjective judgments of how engaging or well-written a story seemed. That captures fluency and polish, but it does not get at originality, usefulness over time, or the ability of an idea to change a system. Those dimensions are where creativity creates lasting value.
Third, the research focused on immediate output. We do not see how patterns might shift if people work with AI over weeks, months, or years. (Note: there is high quality new research beginning to surface on this question). Homogenization may intensify or soften when longer creative cycles and iterative processes are in play.
Finally, the study’s results reflect the particular AI model used at the time. Generative systems evolve quickly, and different models may influence convergence in different ways. This makes the study an important signal but not a fixed forecast of the future.
For me, the takeaway is that this work points in the right direction, but it should be treated as a design brief rather than a conclusion. It shows us where the questions are, and that is valuable. What it cannot do is settle what creativity with AI will mean across the larger cultural and economic systems we are building.
CODA MIRROR: OTHER AREAS THAT STRUGGLE IN SIMILAR WAYS
Peer Review Process and the Narrowing of Ideas
There is an ongoing debate over the peer review process that echoes the same tensions found in creativity research. Peer review often favors safe bets. Researchers know that unconventional ideas risk rejection from reviewers who prefer familiar approaches. Established voices dominate citations, which concentrates attention on the same topics year after year. Quality control is a crucial goal of science. But it also reminds me of how much we need infrastructures that can also process divergence rather than compress it.
The field often defines creativity as “novel and useful,” but what counts as novel or useful is never fixed. Peer review process makes this visible. A reviewer may see an approach as too unconventional to be valid, while another might see the same work as exactly the kind of novelty the field needs.
The parallels matter. Just as AI can raise individual performance while reducing collective range, peer review can validate strong work while keeping new directions out. Both highlight how judgments about novelty and usefulness move depending on who is making them and what system they serve.
This mobility is not a flaw to eliminate but a condition to recognize. If we want innovation systems and workplaces that foster creativity—and that is a big assumption that I’m not sure is entirely accurate—whether in research, business, or culture, we need to design them with enough flexibility that the definitions of novel and useful are not locked into narrow bands.
The idea here is to find better ways to create balance between the old and the new so we benefit from them both.
This is the work.
For those wanting to dive deeper, my “Creative Economy is Dead” series goes deep into how this topic is radically changing the Creative Industries
©2013-2025 Theo Edmonds | All Rights Reserved.
This article contains original intellectual property. No part of it may be reproduced, distributed, or adapted without attribution. Quotation or reference is permitted for non-commercial use with proper credit. The views expressed here are mine alone and do not necessarily represent those of any affiliated organization.
Co-Edit Statement
As a neurodiverse writer, I use AI to support editorial clarity and structure. The conceptual frameworks, metaphors, and systems logic reflect my thinking, grounded in lived experience and interdisciplinary research. AI contributed to refining language, not generating ideas. Original analysis, arguments, and insights are my own.
My Other Places, Other Frames
Business & Innovation: Culture Futurist®Substack
Professional Site: Creativity America
Personal Hub: Culture Futurist