AI has the potential to become a creative partner for researchers. Robots may be useful for stimulating and challenging ideas, visualizing concepts in new ways or exploring information in different fields. However, in practice, the opposite is often true: robots impair creativity.
Can artificial intelligence be truly creative?
As a researcher who has spent 30 years studying human creativity, I think I know why. When a scientist feeds a data set into a robot and says “Give me hypotheses to test,” he is asking the robot to be the creator, not a creative partner. Humans tend to acquiesce to ideas produced by robots, assuming that the robot’s knowledge exceeds their own. When they do, they end up exploring fewer avenues to find possible solutions to their problem.
The answer, in my view, is not to shy away from using artificial intelligence. Instead, researchers must learn how to use robots to enhance their creativity.
I’ve been working on a test of rapid creative thinking called the Divergent Association Task (DAT), which I think points to a way forward. Participants have four minutes to come up with ten nouns that are as different from each other in meaning as possible. This is a difficult task. Once someone thinks of the word ‘queen’, for example, they tend to remember related words such as king or knight – a phenomenon known as thought anchoring.
Most participants assume that if they ask a robot to take this test, it will outperform them. But my colleagues and I found that the average scores for robots and human scores are about the same (D. Wang et al. The hum of nature. Behav. https://doi.org/hbhxbp; 2025).
Does using ChatGPT change your brain activity? A study that raises controversy
In contrast, preliminary follow-up work shows that if participants asked the robot for a process to improve their performance, their scores would rise dramatically. The robot suggests two steps. First, think of ten categories of words – transportation, electronics, mammals, etc. Second, choose one word from each category. This process reduces the consolidation of ideas and maintains diversity of ideas among participants, who each choose different categories of words. The innovator and the robot are now partnering.
This experience suggests that when collaborating with robots, people should ask not what they think, but how they think. Work on human creativity supports this idea.
Human creativity is often a mixture of dead ends, detours, and serendipity. But amid this apparent chaos, studies have revealed some general principles for maximizing creativity in groups.