Large language models and the reverse turing test