Article image

AI outperforms humans in creativity tests

Artificial intelligence (AI), particularly in the realm of creativity and divergent thinking, has marked another milestone. A fascinating study recently revealed that ChatGPT-4, a cutting-edge AI language model, outshined human participants in tests designed to gauge divergent thinking — a core indicator of creativity.

This exploration into the capabilities of AI showcases its potential to generate unique and complex solutions to problems that lack a single correct answer.

Understanding AI creativity and divergent thinking

Divergent thinking allows for the generation of multiple solutions to a given problem, exemplified by questions like, “What is the best way to avoid talking about politics with my parents?”

In this particular study, ChatGPT-4 demonstrated an ability to provide answers that were not only more original but also more detailed than those given by human participants.

Published in Scientific Reports, the study was led by Kent F. Hubert and Kim N. Awa, Ph.D. students in psychological science at the University of Arkansas. They were joined by assistant professor Darya L. Zabelina, who directs the Mechanisms of Creative Cognition and Attention Lab at the same institution.

ChatGPT-4 takes on human creativity

Their research, titled “The current state of artificial intelligence generative language models is more creative than humans on divergent thinking tasks,” utilized three specific tests to measure creativity:

  1. The Alternative Use Task: Participants were asked to think of creative uses for everyday items such as a rope or a fork.
  2. The Consequences Task: This task required participants to envision potential outcomes of hypothetical scenarios, for instance, the implications of humans not needing sleep.
  3. The Divergent Associations Task: Here, participants generated ten nouns that were semantically unrelated, challenging them to think beyond closely related concepts.

The evaluation criteria focused on the quantity and length of responses, as well as the semantic diversity between words.

Understanding AI’s creative boundaries

The findings were clear. ChatGPT-4 consistently delivered more original and comprehensive responses across all tasks, even when adjusting for the fluency of those responses, highlighting its superior creative potential.

However, the researchers caution that these findings, while significant, represent only one dimension of creativity — creative potential.

They remind us that creativity also encompasses active participation in creative activities and achievements, which were not the focus of this study.

Creativity unbound: The human factor

Hubert and Awa pointed out a critical difference between AI and human creativity: AI lacks agency and requires human interaction to activate its creative capabilities, which otherwise remain dormant.

Furthermore, the study did not assess the appropriateness of the AI’s responses, leaving open the question of how human participants balanced creativity with the need for realism.

Awa raised additional questions regarding the measurement of creativity, pondering the effectiveness of these tests in capturing the full spectrum of creative thinking across different populations.

Evolving role of AI in creativity

This study is not a definitive measure of human versus AI creativity but a demonstration of the advancing capabilities of large language models.

The real question of whether AI will replace human creativity remains unanswered. Instead, the authors are optimistic about the future, envisioning AI as a source of inspiration and a tool to enhance human creativity, promising avenues to overcome creative barriers.

As we move forward, the intersection of AI and creativity opens up new possibilities for collaboration between humans and machines, offering a glimpse into a future where AI aids in the creative process, potentially leading to unprecedented levels of innovation.

More about ChatGPT-4, AI creativity, and humans

As discussed above, ChatGPT-4 represents the latest milestone in artificial intelligence, showcasing remarkable advancements over its predecessors.

Developed by OpenAI, this iteration of the Generative Pre-trained Transformer series excels in understanding and generating human-like text, offering a more nuanced and coherent interaction experience.

Core features of ChatGPT-4

ChatGPT-4 demonstrates a profound grasp of language nuances, enabling it to comprehend complex queries and provide responses that are not only accurate but also contextually relevant.

This improvement stems from its training on a diverse dataset that includes a wide range of languages, dialects, and jargon.

One of the most significant advancements in ChatGPT-4 is its ability to maintain context over longer conversations. This ability allows for more meaningful and sustained interactions, as the model can recall and reference previous exchanges within a conversation, offering responses that build on earlier dialogue.

ChatGPT-4 is remarkably versatile, capable of engaging in conversations across various domains such as technology, medicine, law, and more. This flexibility makes it an invaluable tool for professionals seeking information or advice, as well as for individuals looking for casual conversation on a wide range of topics.

Applications and impact

ChatGPT-4 serves as an excellent educational tool, providing explanations, summarizing complex concepts, and offering tutoring in subjects ranging from mathematics to literature. Its ability to adapt to the learner’s level and offer tailored feedback can significantly enhance the learning experience.

Businesses can leverage ChatGPT-4 to revolutionize their customer service, offering instant, accurate, and personalized responses to customer inquiries. This not only improves the customer experience but also reduces the workload on human customer service representatives.

Writers can use ChatGPT-4 as a creative partner, generating ideas, suggesting plot developments, or even composing entire pieces of text. Its capacity to understand and mimic various writing styles makes it a versatile tool for creative endeavors.

Ethical considerations and future directions

As with any powerful technology, the deployment of ChatGPT-4 raises ethical considerations, including privacy concerns, the potential for misuse, and the impact on employment. OpenAI continues to address these challenges, ensuring that ChatGPT-4 and future models are used responsibly and for the benefit of society.

In summary, ChatGPT-4 marks a significant step forward in the field of artificial intelligence, offering enhanced language understanding, improved contextual awareness, and versatility across domains.

Its applications in education, customer service, and creative writing demonstrate its potential to positively impact various aspects of life. As we look to the future, the ongoing development and ethical use of AI technologies like ChatGPT-4 will continue to be of paramount importance.

The full study was published in the journal Scientific Reports.


Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.

Check us out on EarthSnap, a free app brought to you by Eric Ralls and


News coming your way
The biggest news about our planet delivered to you each day