See, as amusing as non-sequitur conversations are (I am a very big fan of them myself), I still don't see how anybody could mistake the ones produced by Cleverbot for ones produced by human beings. Yet somehow they claim that a good percentage of people are fooled. So either a good percentage of people are morons, or Cleverbot is engaged in a conspiracy to deliberately tone down his ability to send human in any conversation with me or with anyone whose conversations with the bot I am likely to see excerpts of.