Hi, Is anyone doing prompt output comparisons between ChatGPT-4 and Bing Chat AI? I've been playing around with this and learned that even though Bing Chat AI is built on ChatGPT-4, it does not tolerate or respond to prompts in the same way. For example, when I give ChatGPT a persona and ask it to create text, do research, etc. based on this persona, it understands and will do what I ask. But when I paste the exact same prompt into Bing Chat AI, it typically responds with "I'm sorry, I cannot be trained to use a different persona. I respond based on my rules and conversation style." I presume this is happening due to some Microsoft rules to prevent Sydney from acting up. Thoughts? Anyone know work arounds for this?