The study investigates the homogenization effects of large language models (LLMs) on human creative ideation, specifically comparing ChatGPT with an alternative creativity support tool (CST), the Oblique Strategies deck. The researchers hypothesized that while LLMs might make users feel more creative and generate more detailed ideas, they could also homogenize the ideas produced by different users. A 36-participant study found that ChatGPT users produced fewer semantically distinct ideas at the group level compared to those using the Oblique Strategies deck, but generated more detailed ideas and felt less responsible for their output. The study discusses potential implications for users, designers, and developers of LLM-based CSTs, suggesting that homogenization effects can be mitigated by providing users with information about the model's typical outputs and by designing tools that encourage diverse and underdetermined responses. The findings highlight the need for careful design to balance the benefits of LLMs with the potential for homogenization in creative processes.The study investigates the homogenization effects of large language models (LLMs) on human creative ideation, specifically comparing ChatGPT with an alternative creativity support tool (CST), the Oblique Strategies deck. The researchers hypothesized that while LLMs might make users feel more creative and generate more detailed ideas, they could also homogenize the ideas produced by different users. A 36-participant study found that ChatGPT users produced fewer semantically distinct ideas at the group level compared to those using the Oblique Strategies deck, but generated more detailed ideas and felt less responsible for their output. The study discusses potential implications for users, designers, and developers of LLM-based CSTs, suggesting that homogenization effects can be mitigated by providing users with information about the model's typical outputs and by designing tools that encourage diverse and underdetermined responses. The findings highlight the need for careful design to balance the benefits of LLMs with the potential for homogenization in creative processes.