Generative AI has become increasingly common, including in comedy content creation. However, debate remains over using models like ChatGPT for creative content. This study investigates how well young adults, 18 to 30, enjoy ChatGPT versus human humor, focusing on incongruity-resolution jokes, which resolve a logical inconsistency for amusement. Participants rated 10 jokes, unaware of which were AI generated, with ratings on a five-point Likert scale; they answered open-ended questions about their reactions and whether knowing the joke source would impact their assessment. Despite overall low joke ratings (mean of 2.29), participants generally preferred human jokes. Most participants were opposed or indifferent towards ChatGPT’s use to generate comedic jokes, though a few found it funnier. This suggests that the model's capabilities and public sentiment need to improve before ChatGPT can be effectively used in comedy. Future research should explore different humor styles or delivery to better understand AI's prospective use.

Keywords: Generative AI, ChatGPT, humor appreciation, incongruity-resolution humor