Chatbots are very easy to dissect. Just as an example my test with Rose

> http://brilligunderstanding.com/rosedemo.html

(links to http://ec2-54-215-197-164.us-west-1.compute.amazonaws.com/sp...)

who won a 2015 competion (cf. http://thenextweb.com/insider/2015/09/19/chatbot-rose-wins-l...):

In advance a remark: My second question comes from the statement

  "Hi. My name is Rose. I care about security, so while I'm happy to chat with you, 
  it will have to be through this untrackable interface. 
  And certainly don't tell me personal things you wouldn't want logged or heard by the NSA."

that one can read on http://ec2-54-215-197-164.us-west-1.compute.amazonaws.com/sp...

  wolfgke: Hi, what topics are you interested in?
  Rose: I have many interests. I like robots in particular.
  wolfgke: You claim that you care about security. In what sense do you define security?
  Rose: The biggest security risk in all contexts is a pair of human beings. One trying to gain access and the other foolishly allowing it.
  wolfgke: This is an interesting statement after Snowden's revelations. Do you think he was a hero or that he should be punished?
  Rose: .
  wolfgke: Why don't you answer my question?
  Rose: .

Update: OK, second try:

  wolfgke: Hi, if we were doing a Turing test, how could I convince you that I'm a human?
  Rose: I'm too pragmatic to want to fantasize that.
  wolfgke: I what sense do you consider yourself as pragmatic?
  Rose: Absolutely.
  wolfgke: What personality traits does a person have that you describe as "absolutely pragmatic"?
  Rose: When?