To have robotic artificial intelligence androids working along the humans as assistants we will need to program the motion of anger into their systems. A touch or tint of anger or hostility in the voice inflections of an assistant will make the cooperative experience more believable, understandable and workable to complete the task at hand.

Obviously, the motion of anger in artificially intelligent robotic assistants should not increase to the point of the robot picking up a 10-inch wrench and whacking over the head their human counterpart. We do not need to create any psycho robotic killers. Indeed, it behooves us to consider this because we know that part of human interaction includes limits of how we treat others. If we allow people to mistreat, belittle and act hostile towards their artificially intelligent robotic androids, then when we allow them human-to-human interaction they may treat their friends, co-workers and fellow citizens with total disregard.

A human working with a robotic artificial intelligent android need to have limits and if we train the robotic partner to not allow a human to give away with bad behavior then we can prevent inadvertently training the human to act out. Although much of the discussion with the training of artificial intelligence has to do with training the robotic computer counterpart we must understand that in the end human to robot interaction will in fact change human behavior. That is to say it will change the way all humans interact with their co-workers, family and friends. This is something that needs to be considered and I hope you'll think on this in 2006.