Has Gemini surpassed ChatGPT? We put the AI models to the test. (arstechnica.com)

In a recent comparative evaluation, Ars Technica tested the performance of Google Gemini 3.2 Fast against OpenAI's ChatGPT 5.2, specifically focusing on the default models accessible to non-subscribers. This test is particularly noteworthy as Apple has chosen Gemini to enhance its Siri voice assistant, underscoring the increasing competition between these AI giants in consumer applications. The evaluation utilized a fresh set of complex prompts, diverging from previous tests conducted in 2023, to better represent real-world user interactions with these technologies. While the assessment lacks rigorous scientific standards, it illuminated striking differences in generative capabilities between the two models. For instance, both Gemini and ChatGPT struggled to produce truly original content when challenged to create original dad jokes, with Gemini's responses closely mirroring existing jokes found online. This highlights not only the ongoing challenges in generative AI regarding originality and creativity but also the differing stylistic approaches of OpenAI and Google to content generation, making it a significant point of consideration for developers and users alike within the AI/ML community.