[armchairhacker]

also we have open LLMs including some which allegedly rival GPT3.5.

Open Assistant I specially remember gave some very weird responses and would get “emotional” especially if you asked it creative questions like philisophical ones