We are uncovering better ways of making artificial intelligence dependable, robust, and trusted by engineering AI systems ourselves and helping others to do it. Through this work we have come to appreciate the following principles:


PUSHING AI TO MEET NEW REQUIREMENTS

    RATHER THAN BEING CONTENT WITH WHAT HAS WORKED

STRIVING FOR REUSABILITY AND STANDARD

    OVER SATISFYING JUST A SINGLE USE CASE

SEEKING COLLABORATION WITH OTHERS

    RATHER THAN IMPOSING OWN EXCELLENCE

EXPERIMENTING CONTINUOUSLY  

    RATHER THAN CONSIDERING DEPLOYED AI TO BE FINAL


We believe that these principles allow us to accelerate our work, excel as a community, and ultimately produce trust of those who depend on us.