ChatGPT-4 in the Turing Test: A Critical Analysis