Scientists from the University of Arkansas in the United States conducted a latest study that involved 151 human participants competing with the artificial intelligence (AI) tool ChatGPT-4 in three tests to evaluate their divergent thinking. The results show that AI performs better than humans. The relevant paper was published in the latest issue of the journal Science Reports. Divergent thinking is a measure of creative thinking, characterized by the ability to provide unique answers to problems without expected solutions. In the latest research, GPT-4 provides more novel and refined answers than human participants. The three tests used in the latest research are "alternative use task", "outcome task", and "divergent association task". The "Alternative Use Task" requires participants to propose creative uses for everyday items such as ropes or forks; The outcome task requires participants to imagine the possible outcomes of hypothetical situations, such as what would happen if humans no longer needed sleep; The divergent association task requires participants to provide 10 nouns that are semantically as distant as possible. The semantic distance between "dog" and "cat" is not significant, while the semantic distance between words such as "cat" and "ontology" is significant. The research team evaluated the number, length, and semantic differences between words answered by humans and AI tools. It was found that in each divergent thinking task, the answers provided by GPT-4 were more original and precise than those provided by humans. In other words, GPT-4 exhibits higher creativity across a range of divergent thinking tasks. Kent Hubert, one of the research leaders and a PhD student in psychology at the University of Arkansas, said that there are also some areas that need to be noted in the latest research, such as the indicators used in the study measuring creative potential, but participation in creative activities or achievement is another aspect of measuring creative ability. In addition, AI relies on the help of human users, and its creativity remains stagnant unless prompted. However, the research team emphasizes that large-scale language models are rapidly evolving and surpassing humans in unprecedented ways. Whether they can replace human creativity remains to be seen. However, using AI as an inspiration tool to assist human creative activities is just around the corner. (Lai Xin She)