@AI Advantage Team Example - I asked for background on the Anthropic contract being lost to OAI and it responded with "my framing was sloppy and I was chasing headlines". Today's was asking for an email response for an inaccurate complaint from a client and uploaded the details. The response was "do not frame it as remove complaint because I don't like it". I already customized my personalization to stress test everything, give me factual information, ask questions if unclear. I have no desire to have to prompt engineer how to have ChaptGPT not be throwing shade. There is chat rooms on Reddit with hundreds of similar complaints. I guess that's my complaint is that it moved away from my style over time.