AI increases creativity and reduces content diversity

AI News


Stories written with the help of AI are judged to be more creative, better written, and more enjoyable.

A new study published in the journal Science Advances finds that AI can enhance creativity by increasing the novelty of story ideas and a story's “usefulness” – its ability to engage with a target audience and its likelihood of being published.

AI was found to “professionalize” stories, making them more enjoyable, enhancing story development, improving writing quality, and making them less boring.

In a study that tasked 300 participants with writing short, eight-sentence “microstories” for a young adult audience, researchers found that participants judged to be less creative produced work that was up to 26.6% higher quality and 15.2% less boring with the AI.

However, AI was not judged to improve work produced by more creative authors.

The study also warned that while AI has the potential to enhance individual creativity, it could also result in a loss of collective novelty, as AI-assisted stories were found to be more similar to each other and less diverse and varied.

Researchers from the University of Exeter Business School, the Data Science and Artificial Intelligence Laboratory, and UCL Management School split 300 study participants into three groups: one group received no AI assistance, a second group was provided with one three-sentence idea using ChatGPT, and a third group of writers were able to choose from up to five AI-generated ideas for inspiration.

The researchers then recruited 600 people to judge the stories, rating them on novelty (whether the story does something new or unexpected) and “usefulness” (whether it's relevant to the target audience and whether the idea could be developed and published).

They found that writers with the most access to AI experienced the greatest creative benefits, with their stories' novelty scores being 8.1% higher and novelty scores being 9% higher compared to stories written without AI.

Writers who used up to five AI-generated ideas also scored higher on emotional traits, producing stories that were better written, more enjoyable, less boring and more entertaining.

The researchers assessed writers' innate creativity using a divergent association task (DAT) and found that the more creative writers who earned the highest DAT scores benefited the least from generative AI's ideas.

Conversely, less creative writers saw a big boost in creativity: using the five AI ideas increased novelty by 10.7% and usefulness by 11.5% compared to writers who didn't use any AI ideas. Their stories were rated as up to 26.6% easier to write, up to 22.6% more enjoyable, and up to 15.2% less boring.

These improvements would bring writers with low DAT scores on par with writers with high DAT scores, essentially equalizing the creativity of low and high creative writers.

The researchers also used OpenAI's embedded application programming interface (API) to calculate similarities between stories.

The researchers found a 10.7% increase in similarity between writers who used one generative AI idea in their story compared to the non-AI group.

Oliver Hauser, professor of economics and deputy director of the Data Science and Artificial Intelligence Institute at the University of Exeter Business School, said: “This is the first step in investigating a question that is fundamental to all human behavior: how will generative AI affect human creativity?”

“Our findings provide insight into how generative AI can enhance creativity and eliminate disadvantages or advantages based on a writer's natural creativity.”

Anil Doshi, Associate Professor at the London School of Management, University of London, added: “These results suggest that individual creativity may increase, but there is also a risk that collective novelty may decrease. If the publishing industry adopts more generative storytelling inspired by AI, our findings suggest that stories as a whole may become less unique and more similar to each other.”

Professor Hauser warns: “This downward spiral resembles a new social dilemma: if individual authors find that their work inspired by generative AI is judged to be more creative, they will be motivated to use more generative AI in the future, but in doing so, they may make stories overall even less novel.”

“In sum, our findings suggest that despite the enhancing effects that generative AI may have on individual creativity, there may be caveats to be observed if generative AI is widely adopted for creative tasks.”

A paper by Professor Oliver Hauser of the University of Exeter Business School and Professor Anil Doshi of the London School of Management, University of London, entitled “Generative AI enhances individual creativity but reduces the collective diversity of novel content”, has been published in Science Advances.

/Public Release. This material from the originating organization/author may be out of date and has been edited for clarity, style and length. Mirage.News does not take any organizational stance or position and all views, positions and conclusions expressed here are solely those of the authors. Read the full article here.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *