Comment by cv5005

Yeah I dont think a single current LLM would fool me in a turing test - I would obiously use all kinds of prompt injection techniques, ask about 'dangerous' or controversial topics, ask about random niche facts in varied fields, etc.