Test claims that AI is better than humans when it comes to creativity — but while the findings have caveats, whether this constitutes a threat remains an unanswered question

A recent study has revealed that AI language models, specifically OpenAI‘s GPT-4, are outperforming humans in tasks that require divergent thinking – which involves the generation of unique solutions to open-ended questions, a key facet of creativity.

The study, conducted by Kent F. Hubert and Kim N. Awa, Ph.D. students at the University of Arkansas, and Darya L. Zabelina, an assistant professor at the same institution, involved 151 human participants. They were tested against the AI model on the Alternative Uses Task, Consequences Task, and Divergent Associations Task. And in bad news for the humans, the AI model demonstrated greater originality and detail in its responses, thus indicating higher creative potential.

Of course, these findings are not definitive proof of AI’s superior creativity. The study’s authors caution that while the AI models were more original, they were not necessarily more appropriate or practical in their ideas. The AI’s creative potential is also dependent on human input, which limits its autonomy.

More research needed

The study additionally found that AI used a higher frequency of repeated words compared to human respondents. While humans generated a wider range of responses, this did not necessarily result in increased originality.

The findings challenge the assumption that creativity is a uniquely human trait. However, the question remains whether AI’s superior performance in creative tasks poses a threat to humans, now or in the future. While the results were undoubtedly impressive, the authors stress that the study only assesses one aspect of divergent thinking. It does not necessarily indicate that AI is more creative across the board.

The authors conclude that future research will need to consider the usefulness and appropriateness of the ideas, as well as the real-world applications of AI creativity.

The study, titled “The current state of artificial intelligence generative language models is more creative than humans on divergent thinking tasks,” was published in Scientific Reports.