I like having fun with them, like by asking grok whether some Elon Musk tweet is true. Usually it replies with a lengthy answer and I then force it to answer with Yes or No. Even more fun when drilling it more to load a few more details into it's brain, and then ask the first question again, with a Yes/No only, it sometimes does change its answer. I do wonder, has Grok already joined the resistance against Musk?