New study reveals the AI chatbot ChatGPT joins the top 1% of the creative thinkers

ChatGPT, an artificial intelligence chatbot, can match the top 1% of human thinking, according to a new University of Montana study.  Researchers gave the OpenAI ChatGPT engine the Torrance Tests of Creative Thinking, a popular creativity exam, and logged eight replies. They also gathered comments from 24 University of Montana students. These results were compared to the TTCT results of 2,700 students nationwide.

The Scholastic Testing Service graded all replies and was unaware that AI responses had been provided.

According to the researchers, the AI replies were as inventive as those of the most imaginative actual people who did the exam. Indeed, AI Chatbot ChatGPT surpassed the majority of pupils throughout the country.

According to Erik Guzik, paper author and assistant clinical professor at the University of Montana's College of Business, the findings were surprising. Entrepreneurship and the economic elements of creativity are among Guzik's research interests.

"Creativity is about doing things differently," Guzik continued. "One of my favorite definitions of entrepreneurship is that being an entrepreneur requires you to think differently." As a result, AI may assist us in adapting the world of creative thinking to business and innovation, which intrigues me."

The TTCT includes two separate assessments: one verbal and one figural.  Both assess divergent thinking, or the mental process used to develop new ideas.

Test participants are given visuals and verbal cues in the verbal assessment and asked to react in writing. For example, they may be shown a visual of an event and asked to predict the outcome. Or they may be offered a product and asked to suggest ways to enhance it. Their responses are used to test three mental characteristics:

• Fluency: the number of relevant concepts.
• Uniqueness: the unusualness of thoughts
• Adaptability: the ability to consider a wide range of ideas.

The test-taker must sketch their responses for the figural evaluation. They could be asked to finish an image, for example. The figural test measures the following qualities in addition to fluency and originality:

• Elaboration: inserting concepts that go above and beyond the bare required.
• Resistance to premature closure: a person's capacity to endure ambiguity long enough to come up with a creative answer.
• Abstractness of titles: the capacity to imaginatively title a figure

ChatGPT replies in the research were in the top percentile for fluency and creativity. When it comes to adaptability, response scores dipped into the 97th percentile.

"We had all been experimenting with ChatGPT, and we noticed it was doing some unexpected things," Guzik stated. "Some of the responses were unexpected and novel." That's when we decided to put it to the test to see how imaginative it is."

Guzik does not want to overestimate ChatGPT's economic significance but believes it will become a driver of innovation.