Can generative artificial intelligence systems like ChatGPT genuinely create original ideas? A brand new study led by Professor Karim Jerbi from the Department of Psychology on the Université de Montréal, with participation from renowned AI researcher Yoshua Bengio, takes on that query at an unprecedented scale. The research is the most important direct comparison ever conducted between human creativity and the creativity of huge language models.
The study, published in Scientific Reports (Nature Portfolio), points to a major shift. Generative AI systems have now reached a level where they’ll outperform the common human on certain creativity measures. At the identical time, probably the most creative people still show a transparent and consistent advantage over even the strongest AI models.
AI Reaches Average Human Creativity Levels
Researchers evaluated several leading large language models, including ChatGPT, Claude, Gemini, and others, and compared their performance with results from greater than 100,000 human participants. The findings highlight a transparent turning point. Some AI systems, including GPT-4, exceeded average human scores on tasks designed to measure divergent linguistic creativity.
“Our study shows that some AI systems based on large language models can now outperform average human creativity on well-defined tasks,” explains Professor Karim Jerbi. “This result could also be surprising — even unsettling — but our study also highlights an equally essential remark: even one of the best AI systems still fall in need of the degrees reached by probably the most creative humans.”
Further evaluation by the study’s co-first authors, postdoctoral researcher Antoine Bellemare-Pépin (Université de Montréal) and PhD candidate François Lespinasse (Université Concordia), revealed a striking pattern. While some AI models now outperform the common person, peak creativity stays firmly human.
In reality, when researchers examined probably the most creative half of participants, their average scores surpassed those of each AI model tested. The gap grew even larger among the many top 10 percent of probably the most creative individuals.
“We developed a rigorous framework that enables us to check human and AI creativity using the identical tools, based on data from greater than 100,000 participants, in collaboration with Jay Olson from the University of Toronto,” says Professor Karim Jerbi, who can also be an associate professor at Mila.
How Scientists Measure Creativity in Humans and AI
To guage creativity fairly across humans and machines, the research team used multiple methods. The first tool was the Divergent Association Task (DAT), a widely used psychological test that measures divergent creativity, or the flexibility to generate diverse and original ideas from a single prompt.
Created by study co-author Jay Olson, the DAT asks participants, whether human or AI, to list ten words which can be as unrelated in meaning as possible. An example of a highly creative response includes words like “galaxy, fork, freedom, algae, harmonica, quantum, nostalgia, velvet, hurricane, photosynthesis.”
Performance on this task is strongly linked to results on other established creativity tests utilized in writing, idea generation, and artistic problem solving. Although the duty is language-based, it goes well beyond vocabulary. It engages broader cognitive processes involved in creative considering across many domains. The DAT also has practical benefits, because it takes only two to 4 minutes to finish and may be accessed online by most people.
From Word Lists to Real Creative Writing
The researchers then explored whether AI success on this straightforward word association task could extend to more complex and realistic creative activities. To check this, they compared AI systems and human participants on creative writing challenges akin to composing haiku (a brief three-line poetic form), writing movie plot summaries, and producing short stories.
The outcomes followed a well-known pattern. While AI systems sometimes exceeded the performance of average humans, probably the most expert human creators consistently delivered stronger and more original work.
Can AI Creativity Be Adjusted?
These findings raised one other essential query. Is AI creativity fixed, or can it’s shaped? The study shows that creativity in AI may be adjusted by changing technical settings, particularly the model’s temperature. This parameter controls how predictable or adventurous the generated responses are.
At lower temperature settings, AI produces safer and more conventional outputs. At higher temperatures, responses grow to be more varied, less predictable, and more exploratory, allowing the system to maneuver beyond familiar ideas.
The researchers also found that creativity is strongly influenced by how instructions are written. For instance, prompts that encourage models to take into consideration word origins and structure using etymology result in more unexpected associations and better creativity scores. These results emphasize that AI creativity depends heavily on human guidance, making interaction and prompting a central a part of the creative process.
Will AI Replace Human Creators?
The study offers a balanced perspective on fears that artificial intelligence could replace creative professionals. While AI systems can now match or exceed average human creativity on certain tasks, they still have clear limitations and depend on human direction.
“Regardless that AI can now reach human-level creativity on certain tests, we’d like to maneuver beyond this misleading sense of competition,” says Professor Karim Jerbi. “Generative AI has above all grow to be a particularly powerful tool within the service of human creativity: it’ll not replace creators, but profoundly transform how they imagine, explore, and create — for individuals who select to make use of it.”
Relatively than signaling the tip of creative careers, the findings suggest a future where AI serves as a creative assistant. By expanding ideas and opening recent paths for exploration, AI may help amplify human imagination quite than replace it.
“By directly confronting human and machine capabilities, studies like ours push us to rethink what we mean by creativity,” concludes Professor Karim Jerbi.
Concerning the Study
The paper titled “Divergent creativity in humans and enormous language models” was published in Scientific Reports on January 21, 2026. The research brought together scientists from Université de Montréal, Université Concordia, University of Toronto Mississauga, Mila (Quebec AI Institute), and Google DeepMind.
Professor Karim Jerbi led the study, with Antoine Bellemare-Pépin (Université de Montréal) and François Lespinasse (Université Concordia) serving as co-first authors. The research team also included Yoshua Bengio, founding father of Mila and LoiZéro, and a pioneer of deep learning, the technology behind modern AI systems akin to ChatGPT.