Bit idea: Ask AI a question then ask it repeatedly how it was incorrect in its last response

This is the only proper use of AI, frustrating the algorithm until it just spews gibberish at you.