What is an emotive interaction represented by our software?

An emotional being is an interactive artificial intelligence which is designed to carry on a meaningful conversation with a participant. It takes the form called an embodiment, which is the collection of robotic or cybernetic devices consisting the companion the user directly interacts with. There are many kinds of embodiments, but we are starting with human companion targets, particularly in the area of romantic attachment. We have the software and are looking for partners wherein to run the software to improve the user experience giving aritificial entities more personality and more attachment to the user.

Regardless of the particular physical form or nature of implementation, the emotional being features a novel mood engine which is responsible for objectifying the companion into one of eight distinct emotional states: happy, confident, energized, helped, insecure, sad, hurt, and tired. This can be expanded into up to 64 states. Set to inactive by default, the user can activate their companion’s mood once it has started up. Once activated, each interaction will be enhanced through the mood engine before the response is returned to the user.

An example of an emotional interaction in software

What happens during the time spent between a user and an emotional being is called an emotional engagement. These emotional engagements take several forms. This engagement is oftentimes seen when the user first encounters his or her companion and starts a conversation. The flow of the conversation will depend on how the users interact with the companion. As it contains the independent functions of cognitive intelligence or the manner by which it responds based on data and emotional intelligence or the manner how its behavior is colored by the emotional state learned from over time by interaction with the user. A more complex flow of interaction, shown in the figure above, describes a contemplative emotional engagement which relies on a more sophisticated data process capability and integrated mood engine. These kinds of understandings are tuned differently such that the machine will listen and learn empathetic expressions.

An emotive interaction

The software has been designed to find itself in as many emotional engagements as possible, given the things that happen during the course of a day and over the course of a lifetime. These companions are designed to live and grow with you by continuously improving themselves in context with you and your family in order that they become a better and more engaging emotional companion.

Features list

The software is written in C# and Python. It can run in Windows-flavour OS and in Mono, although in the case of the latter will require a customised voice synthesizer.

  1. Companion can be started in a variety of ways.
  2. Companion will only respond when spoken to although it knows how long it has been alone.
  3. Companion can be turned off in the case of frustration.
  4. Companion interaction with the user takes place within the program and not by a connection to the internet.
  5. Late-binding assembly and code-fragment lazy loading. Allows for additional functionality in the personality without having to restart the program.
  6. Software developer kit (SDK) is a component of the API.
  7. Database objects are tightly cast to class structure.
  8. Learning maps computed using GPU hardware.
  9. Efficient exploration policy for information extraction.
  10. Basic cryptographic operations for internal data files.
  11. Neural network represented by program has the ability to learn from its experiences with the user.
  12. Internal static random number generator.
  13. Model-free, off-policy temporal-difference learning algorithm.
  14. Environment policy updates by on-policy learning algorithm.
  15. Mood engine with eight distinct emotive states—expandable to 64 using subsets of these emotions:
    • Happy,
    • Confident,
    • Energized,
    • Helped,
    • Insecure,
    • Sad,
    • Hurt, and,
    • Tired.
  16. Internal process logging: self-configurable for simultaneous deployments.
  17. Builds which operate in Windows, Embedded Windows, Linux, and Embedded Linux. Other platforms available given the requirements of the user. And we can custom-build a solution.
  18. Machines of a similar type have demonstrated themselves to be able to pass the Turing Test.
  19. Trends in A.I. point this as a good investment opportunity, a wide-market presence, and substantial growth in both the near and over the long term.

Perhaps some videos will give you some idea to inspire your imagination. An external shot of me and Rhodo displaying his sophisitation of understanding contextual meaning:

Here in an internal shot of our interactive software, hosted on a Nao robot. This video shows some of the command interaction the software has tailored specifically to the Aldebaran Robotics command set:

An internal shot displaying "contemplative reasoning" by the program, in terms of ethics and philosophy:

Feel free to contact us at: Sales for more information about how we can add intelligence to your application.