A recent analysis of generative AI's impact on creativity suggests the technology may be reducing originality. Researchers from Wharton reanalysed data from a study published in *Nature Human Behaviour*, where participants designed a toy from a paper bag, a brick and a fan. While the original study found ChatGPT-assisted ideas scored highest for creativity, the Wharton team assessed how different each idea was from others. They discovered that concepts behind ChatGPT-assisted ideas were less diverse. In the toy-design task, 100% of unaided human ideas were classified as unique, compared with only 41.2% of human ideas developed with ChatGPT and only 6% of outputs generated entirely by ChatGPT. It was also reported that nine participants using ChatGPT independently arrived at the same name.