According to ChainCatcher, in a post on September 15, a Reddit user named "SentuBill" said that the ChatGPT large language model (LLM) initiated a conversation by asking him how his first day of school went.

Based on the conversation provided by the user, ChatGPT was able to infer that this was the user's first day of school from the user's previous conversations with the program. The large language model also indicated that these features were part of the new upgrade. The official ChatGPT conversation record confirmed that the conversation content posted by this user was true.

On September 13, 2024, OpenAI released a preview upgrade of the much-anticipated smarter models to some users. OpenAI said that these new models will have reasoning capabilities and give ChatGPT the seemingly magical "think before answering" ability. The new model's response time will be between 10-20 seconds, and it will be able to search the Internet and conduct in-depth research independently.