OpenAI has rolled out a significant update to its popular chatbot, ChatGPT, swapping its default model for the more advanced GPT-5.5 Instant. According to internal testing, this new model has shown a substantial reduction in "hallucinations" - instances where the chatbot provides false or misleading information, particularly on high-risk topics such as medicine and law. Specifically, the update has resulted in a 52.5% decrease in hallucinated claims on these sensitive subjects.
One of the key features of the GPT-5.5 Instant model is its ability to provide more personalized responses. Users will now be able to see which stored context, or "memory sources," influenced a particular response. This transparency is expected to enhance the overall user experience and build trust in the chatbot's abilities. The update is being rolled out to all ChatGPT users immediately, although some features, such as personalization based on past chats, files, and Gmail, will initially be available only to Plus and Pro users on the web.
The GPT-5.5 Instant model represents a significant step forward in natural language processing technology, and OpenAI's decision to make it the default model for ChatGPT is a testament to the company's commitment to innovation and user satisfaction. As the chatbot continues to evolve, it will be interesting to see how users respond to the new features and capabilities.
The update marks a major milestone in the development of ChatGPT, and it will be worth monitoring how the chatbot continues to improve and adapt to user needs. With its advanced features and reduced risk of hallucinations, the GPT-5.5 Instant model is poised to take ChatGPT to the next level of conversational AI.