Despite the hype, from my admittedly limited experience I haven't seen a chatbot that is anywhere near passing the turing test. It can seemingly fool people who want to be fooled but throw some non-sequiturs or anything cryptic and context-dependent at it and it will fail miserably.