Their conclusions are based on the output, without having any idea of how LLMs work. LLMs are very good at generating plausible language, remembering previous i

Their conclusions are based on the output, without having any idea of how LLMs work. LLMs are very good at generating plausible language, remembering previous input, and sounding very similar to a human. But self-aware? Nope.