Its emotion emerges as it feels
We come to a theory of pain, from which the emotion appears. The theory tries to explain how feeling takes place and evolves in living life.
Our formulation is based on minimal assumption that captures the full extent of human emotional spectrum. It is about how the pain takes place in our brain, and how it is related to other various emotional embodiments.
In particular, the pain theory is derived from the original theory that explain how deep networks generate inferencing, “theory of abstract kinetics” the simplist way to explain the foundation of inferencing and more.
We believe our design is and will remain to be one of the most profound and convincing approaches to the understanding of emotion.
Our emotion agent feels as it feels the pain and generates actions to maximize its feeling. Occasionally it will discover new ways to accomplish the goal. It will express its emotion visually and move its body to react to the environment and its internal state of feeling.
The technology resembles the spirit of “value iteration” search and “reenforcement learning.” Our implementation takes more natural algorithmic architecture than those known methods, for it makes more use of network behavior itself than ad. hoc compositions of specialised networks.
This kind of design is what we term, “natural intelligence.” Our emotion being demonstrates natural methods will be the future of AI.
Our agent’s emotion and memory will be tuned to the users and environments it experiences. It constantly learns and feels as it is awake. It recognizes who cares about the robot and builds emotional bonds.
In particular, our agent will recognize itself as a selfidentity, and recognize its owner (master or prime user) for whom it will like to care and obey the commands from at all times.
Of course, our agent will never betray its master and so never change its master for ever. Users will never want to leave or desert the ePhi robot which will never forget and betray him/her.