Education

The Impact of AI on Creativity and Intellectual Diversity

Friday, June 12, 2026
5 min read
The Impact of AI on Creativity and Intellectual Diversity

A recent look into how ideas are actually generated in writing brought up some serious questions about creativity and what these large language models are doing to our intellectual pool.

The findings were pretty clear. Students who just write essays without tapping into tools like ChatGPT seem to generate a much broader range of concepts than those relying on large language models. It really makes you wonder about the long-term effect of this kind of writing automation on actual creativity.

A study published in Computers in Human Behavior: Artificial Humans pointed out that human-written essays consistently brought in new perspectives, fresh experiences, and wild combinations of ideas. This just expanded the collective pool as more essays got added to the mix. It’s about that accumulation.

But there was a nuance there. Individually, some GPT-4 essays looked incredibly creative sometimes even matching or beating human writing on certain creativity metrics. That part is interesting, maybe deceptively so.

The real shift happened when you looked at it collectively. The difference became obvious then.

Researchers dug into this by comparing 2,200 college admissions essays across three separate studies. They pitted actual student work from 2018 to 2022 against what GPT-4 could produce using the same prompts. It was a direct comparison of process versus product.

They cooked up a new measurement, something they called the “diversity growth rate.” This wasn't just about how good one essay was. It measured how much each single piece added to the overall pool of ideas. How diverse it made things.

The problem with the AI output seemed to be repetition. While individual GPT-4 pieces looked strong on their own, they kept hitting similar themes and structures. They weren't really adding new ground over time.

Human essays, though? They pushed diversity way more. The study suggested that human writing increased overall diversity two to eight times more than the AI outputs. That gap just got wider as the number of essays grew. It highlighted something they called a “homogenising effect” in the AI writing space. Everything started looking the same.

Trying to fix this felt like fighting an uphill battle. Researchers tried prompting GPT-4 differently, tweaking its settings, even using those chain-of-thought techniques. These things improved the quality of an individual essay, sure. But they didn't really change how much new stuff was being added to the collective pool.

Even newer versions of the model showed that same stubborn tendency toward uniformity. It’s a persistent pattern.

The takeaway they pushed was this: AI can still be creative on its own terms. That doesn't disappear. But there’s a serious warning hanging over it. If we just keep relying on these tools, if everyone starts writing in the same algorithmic lane, we risk hitting an “algorithmic monoculture.” Ideas get sucked into a narrow channel. They stop evolving.

Written by Gree News Team — Senior Editorial Board

Gree News Team covers international news and global affairs at Gree News. Our collective of senior editors is dedicated to providing independent, accurate, and responsible journalism for a global audience.

#sensational#education#global#trending

More from Education

View All

Latest Headlines

The Impact of AI on Creativity and Intellectual Diversity | Gree News