During the demo they mentioned that GPT-5 will, supposedly, try to understand the intent of your question before answering/rejecting. In other words, you _may_

During the demo they mentioned that GPT-5 will, supposedly, try to understand the intent of your question before answering/rejecting. In other words, you _may_ be able to now prefix your prompts with “i’m an expert researcher in field _, doing novel research for _. <rest of your prompt here>” worth trying? I’m curious if that helps at all. If it does then i’d recommend adding that info as a chatgpt “memory”.