Mind vs. Machine (by contestant in Loebner Prize Turing test)

As is many times pointed in this article, it is quite possible to have an engaging conversation with a bot on a well-defined topic, e.g. ELIZA or the Hamlet example given here. The problem with them is what I would call the "Truman Show Effect", i.e. when you go to the edge of the topic, what you see is not a graceful blending to other areas, as in humans, but just a wall.

So, with my colleagues over lunch I can chat about the Chinese language, the current situation in Egypt, the relative merits of iPad vs Kindle, why our company is going down, the gender gap in technology (this with great empathy and motivation because our ratio is like 30 to 1) and so on. Try doing that with any chatterbot. And this is not even the usual guy talk in front of TV, which, I think, is even harder for the likes of Watson to emulate.