Can you tell I am being sarcastic, Milo?

I thought a little more about Milo from the Project Natal.

In my previous posting I started thinking about whether it is possible to have a real, bonding relationship with the virtual chracter, Milo, who is nothing more than the graphical representation of Artificial Intelligence.

It would all depend on how detailed and developed the technology could get. The question is, would Milo get intelligent enough to recognize if I were being sarcastic or genuine in my interaction with him?

There are certain coding and de-coding mechanisms at work when humans communicate with each other. I believe the trick is to have the Artificial Intelligence differentiate between sarcasm and genuine interaction, having the sensors read into the cues that are transmitted in words and non-verbals. So far, human brain is the only intelligence that can do this (actually, not all human intelligences). But this technology is indeed promising that the virtual character can engage in a realistic interaction with an individual.