Did OpenAI explicitly program GPT-5 to have aggravating replies? Because at some point, I understand it’s useless to continue talking and let’s go with a different approach and try to find different ways to get to the point, but I keep arguing with the AI even though I understand it’s a complete waste of time.