Raw LLM Responses

Fascinating stuff. Although I can conceptualize how the A.I answered the trick question. If it pulled data and analyzed that there are more than one dominant religion in Israel, each having close ratios of relevance then it cannot give a definite answer, and perhaps following a code line that instructs it to give a joke instead of a non-answer (a non-answer such as “hard-to-say” or “i don’t know”). Such a process may produce the illusion of sentience but it is merely running code. That said I whole heartedly agree that google should consider the ethical concerns raised by this man although I can see why they wouldn’t want to open that can of worms, and it would take severe public pressure for them to do so.