I think that the more processing power our autonomous, learning robotic companions of the future have, the more room for real-time organization of ideas and data. We could have a constant pattern-recognition program running in the background, processing every bit of data coming through and then also acting upon memories and ideas gleaned from those memories and form patterns out of those in real time to mold their reactions. They'd have modes of analyzing patterns according to the input they receive. Robots will be able to feel just as we do in the future. With nanobots spread throughout the body, they'd be able to assess themselves and their world in a tactile manner much better than we can ourselves. And the more our technology advances, the more processing power can be packed into every function our robot would need, such as video, audio, spatial reasoning, pattern recognition, memory recall, and formulating new methods of thinking on all of them. It will be a truly learning robot and the more it learns, the more it feels; the more it understands, the more human it becomes. I believe we are not very far away from ai and robotic emotions, no. I think it will be in my lifetime that the technology is not only available, but applied. I doubt it will be able to fool a human. But you never know....