Mack Institute predoctoral fellow Lennart Meincke, Wharton Professor Gideon Nave, and Mack Institute Co-Director Christian Terwiesch have published a new paper on AI and creativity in the prestigious journal Nature Human Behaviour. Building on growing research into how artificial intelligence can support idea generation, their study reveals a crucial tradeoff: while ChatGPT can enhance the creativity of individual ideas, it significantly reduces the diversity of ideas within a group of ideas—a key ingredient for effective brainstorming.
In one striking example, participants were asked to invent toys using a brick and a fan. Among those using ChatGPT, 94% of ideas shared overlapping concepts, with nine participants independently naming their toy “Build-a-Breeze Castle.” By contrast, human-generated ideas were entirely unique.
Across five experiments, brainstorming sessions assisted by ChatGPT consistently produced narrower sets of ideas. In 37 of 45 statistical comparisons, this drop in diversity was significant. These findings echo recent research in creative writing and innovation management, highlighting how overreliance on generative AI can limit the breadth of perspectives, even when individual ideas seem original.
“In real-world problem-solving,” the authors write, “the true value of brainstorming stems from the diversity of ideas rather than multiple voices repeating similar thoughts. As original as transforming a tennis racket and a garden hose into a sprinkler may be, successful brainstorming would yield a mosaic of unique perspectives—not just a line-up of sprinklers.”