If you were stuck in a room, with nothing but a computer connected to a chatroom, how would you convince the people in there that you are a human, and not a bot?
This brings up the question artificial intelligence, and the difference between how a robot can be programmed to response to a certain question, and how real humans respond. We humans can write programs that output a certain phrase given an input. So if the question "who are you?" was asked, you could write a program that would reply, "My name is John, who are you?" Based on such a conversation, you would never be able to tell if it were two humans talking, or two bots talking.
This talk arises after reading an interesting post on a website which asks the same question.
It really is an interesting read, although it is long, but just read through the conversation and offer what you think.