Just to clarify: it is not odd that in different contexts and with different prompts an LLM gives completely opposite answers.
It is the actual definition of what an LLM does.
It is crazy that people believes that GPTs are people and that have opinions.
