Recent results from a study conducted by the University of Montana, in collaboration with its partners, suggest that artificial intelligence could compete with the top 1% of human thinkers on a standardized test of creativity.
The Torrance Tests of Creative Thinking, an established and respected measure of human creativity, were used in this research, conducted by Dr. Erik Guzik, clinical assistant at the College of Business at university.
see more
Sweet news: Lacta launches Sonho de Valsa e Ouro chocolate bar…
Brazilian wine wins label award at the 'Oscars' of…
The team presented the Scholastic Testing Service with eight responses generated by ChatGPT, an application powered by the AI engine GPT-4, along with responses from a control group of 24 students studying entrepreneurship and personal finance at the University of Montana. These were compared with responses from 2,700 college students who took the TTCT in 2016, without knowledge of AI participation by the testing service.
The performance of ChatGPT, an artificial intelligence application, has reached the forefront of creativity, according to the results of a recent study.
ChatGPT proved to be in the highest percentage when it came to fluency – the ability to generate a large number of ideas – and originality – the ability to bring new ideas to light. In the flexibility criterion, referring to the ability to produce different categories and types of ideas, the AI dropped a little, reaching a percentage of 97.
“It's the first time we've proven that ChatGPT and GPT-4 are in the top 1% for originality,” said Guzik, noting that this was a first-of-its-kind discovery.
Guzik conducted the study during the spring semester, with help from Christian Gilde of UM Western and Christian Byrge of Vilnius University. They presented their findings at the Southern Oregon University Creativity Conference in May.
“We were very careful at the conference not to misinterpret the data,” said Guzik. “We have just presented the results. But we share strong evidence that AI appears to be developing creative capacity equal to, or even superior to, human ability.”
Guzik, who has long had a keen interest in creativity since his seventh grade at Palmer, Massachusetts, while participating in a program for gifted students, asked ChatGPT about its performance in the TTCT. At the conference, the answer given by the AI was shared:
“ChatGPT told us that maybe we don't have the capacity to fully understand human creativity, which I think is correct,” Guzik commented. “Furthermore, AI has suggested that we need more sophisticated assessment tools capable of differentiating between human-generated and AI-generated ideas.”
“I think we know that the future will include AI in some form,” Guzik said. “We have to be careful about how it is used and consider the necessary rules and regulations. But companies are already using it for many creative tasks. In terms of entrepreneurship and regional innovation, this is a game changer.”