Is that how your llm model reacts when you can't formulate back a reasonable argument?